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/branches/primop-branch-3/NOTES/HISTORY
ViewVC logotype

Diff of /sml/branches/primop-branch-3/NOTES/HISTORY

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

revision 890, Thu Jul 19 20:38:56 2001 UTC revision 1838, Fri Oct 14 21:24:05 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: 2005/10/14
18    Description:
19    Add Zhong Shao's fix for datatype equality functions.
20    
21    ----------------------------------------------------------------------
22    Name: Allen Leung (leunga (at) reservoir (dot) com)
23    Date: 2005/10/14 12:30:00 EST
24    Tag: leunga-20051014-x86-spill
25    Description:
26    
27    Bug fix a bug found by Carl Hauser.
28    There was a typo in the reload code for FCMP in x86SpillInstr.sml.
29    
30    ----------------------------------------------------------------------
31    Name: Allen Leung (leunga (at) reservoir (dot) com)
32    Date: 2005/10/14 11:15:00 EST
33    Tag: leunga-20051014-x86-asm
34    Description:
35    
36    Removed some debugging code in file x86Asm.sml.
37    The function emit_operand was printing out debugging output.
38    
39    ----------------------------------------------------------------------
40    Name: John Reppy
41    Date: 2005/07/27
42    Tag:
43    Description:
44    
45    Fixed ml-lex to recognize "\r" as representing carriage return.
46    
47    ----------------------------------------------------------------------
48    Name: John Reppy
49    Date: 2005/07/27
50    Tag:
51    Description:
52    
53    Fixed ml-yacc to work on files with non-native end-of-line
54    encodings (e.g., Windows text file on a Unix system).
55    
56    ----------------------------------------------------------------------
57    Name: John Reppy
58    Date: 2005/07/20
59    Tag:
60    Description:
61    
62    Added changes from Dominic Evans (oldmanuk (at) gmail (dot) com)
63    to support HPUX 11.
64    
65    ----------------------------------------------------------------------
66    Name: John Reppy
67    Date: 2005/07/06
68    Tag:
69    Description:
70    
71    Changes to the SML/NJ library.  See smlnj-lib/CHANGES for details.
72    
73    ----------------------------------------------------------------------
74    Name: Matthias Blume (blume (at) tti - c (dot) org)
75    Date: 2005/07/06 00:45:00 CDT
76    Tag: blume-20050706-slice-copy
77    Description:
78    
79    Fixed reversed logic for deciding whether to "copy up" or "copy down"
80    in *-array-slice.sml.
81    
82    ----------------------------------------------------------------------
83    Name: Allen Leung (leunga (at) reservoir (dot) com)
84    Date: 2005/05/31 17:00:00 EST
85    Tag: leunga-20050531-cygwin-fault-2
86    Description:
87    
88    A typo in the cygwin code fixed.
89    
90    ----------------------------------------------------------------------
91    Name: Allen Leung (leunga (at) reservoir (dot) com)
92    Date: 2005/05/31 16:47:00 EST
93    Tag: leunga-20050531-cygwin-fault
94    Description:
95    
96    Updated Cygwin's fault/signal handling to match the Windows version.
97    Updated the export list.
98    
99    ----------------------------------------------------------------------
100    Name: Matthias Blume (blume (at) tti - c (dot) org)
101    Date: 2005/05/18 13:35:00 CDT
102    Tag: Release_110_54
103    Description:
104    
105    New working version (110.54).  NEW BOOTFILES!
106    
107    ----------------------------------------------------------------------
108    Name: Matthias Blume (blume (at) tti - c (dot) org)
109    Date: 2005/05/18 11:58:00 CDT
110    Tag: blume-20050518-installer
111    Description:
112    
113    Added support scripts for Mac OS X PackageMaker and modified
114    config/install.sh so that it supports re-dumping a heap image
115    after customization.
116    
117    ----------------------------------------------------------------------
118    Name: Matthias Blume (blume (at) tti - c (dot) org)
119    Date: 2005/05/18 10:55:00 CDT
120    Tag: blume-20050518-realdiv-noovld
121    Description:
122    
123    Un-overloaded / to work around bug in overloading resolution code.
124    
125    ----------------------------------------------------------------------
126    Name: Matthias Blume (blume (at) tti - c (dot) org)
127    Date: 2005/05/16 23:50:00 CDT
128    Tag: blume-20050516-redump-heap
129    Description:
130    
131    Added mechanism for re-creating a heap file for the interactive system
132    after configuration variables have been changed.
133    
134       CM.redump_heap : string -> unit
135    
136    This is much like SMLofNJ.exportML, but starting from the resulting
137    heap does not return to the caller of CM.redump_heap but
138    restarts the interactive system from scratch.  The original call of
139    CM.redump_heap does not return but ends the interactive session.
140    Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
141    
142    Internally, redump_heap winds the dynamic execution context back to
143    the point where the original heap image was created and re-executes
144    the heap image generation code in the boot code.
145    
146    ----------------------------------------------------------------------
147    Name: Matthias Blume (blume (at) tti - c (dot) org)
148    Date: 2005/05/09 21:55:00 CDT
149    Tag: blume-20050509-word64
150    Description:
151    
152    Added a hack to the existing hack known as Word64 to make fromString
153    behave correctly.  I am still not sure whether Word64.scan will work
154    as specified with respect to the interaction of radix and prefix.
155    
156    ----------------------------------------------------------------------
157    Name: Allen Leung (leunga (at) reservoir (dot) com)
158    Date: 2005/05/04 11:50:00 EST
159    Tag: leunga-20050504-checkgc
160    Description:
161    
162    Added a gc protocol checking phase.  This phase is enabled with
163    the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
164    
165    ----------------------------------------------------------------------
166    Name: Matthias Blume (blume (at) tti - c (dot) org)
167    Date: 2005/05/04 10:45:00 CDT
168    Tag: blume-20050504-intinf
169    Description:
170    
171    Fixed a bug in the implementation of div and mod for IntInf.
172    Thanks to Neophytos Michael for reporting the problem.
173    
174    ----------------------------------------------------------------------
175    Name: Matthias Blume (blume (at) tti - c (dot) org)
176    Date: 2005/05/04 10:35:00 CDT
177    Tag: blume-20050504-join
178    Description:
179    
180    Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
181    
182    ----------------------------------------------------------------------
183    Name: Matthias Blume (blume (at) tti - c (dot) org)
184    Date: 2005/02/28 23:40:00 CST
185    Tag: blume-20050228-mVar
186    Description:
187    
188    Fixed serious bug (brown paper bag variety) in new implementation of
189    structure Atom in CML.  (I had accidentally used a mailbox instead of
190    an mvar, leaving the door open for races.)
191    
192    ----------------------------------------------------------------------
193    Name: Matthias Blume (blume (at) tti - c (dot) org)
194    Date: 2005/02/25 15:00:00 CST
195    Tag: Release_110_53
196    Description:
197    
198    New working version (110.53).  NEW BOOTFILES!
199    
200    ----------------------------------------------------------------------
201    Name: Matthias Blume (blume (at) tti - c (dot) org)
202    Date: 2005/02/25 14:50:00 CST
203    Tag: blume-20050225-susp
204    Description:
205    
206    Brought back SMLofNJ.Susp.  The underlying suspension type is the one
207    implemented in Core -- which means that it is the same as the one used
208    by the lazy extension.
209    
210    ----------------------------------------------------------------------
211    Name: Matthias Blume (blume (at) tti - c (dot) org)
212    Date: 2005/02/24 16:50:00 CST
213    Tag: blume-20050224-cml-atom
214    Description:
215    
216    Simpler and at the same time more general implementation of structure
217    Atom in CML.
218    
219    ----------------------------------------------------------------------
220    Name: Matthias Blume (blume (at) tti - c (dot) org)
221    Date: 2005/02/15 17:35:00 CST
222    Tag: blume-20050215-tools
223    Description:
224    
225    Created new "tools" directory under "src" and moved "TraceDebugProf"
226    there.
227    
228    ----------------------------------------------------------------------
229    Name: Matthias Blume (blume (at) tti - c (dot) org)
230    Date: 2005/02/10 17:55:00 CST
231    Tag: blume-20050210-longlong
232    Description:
233    
234    Implemented "long long" arguments and results for NLFFI.  (Only the
235    PPC/MacOS implementation is complete, the other backends still need to
236    be updated.)
237    
238    ----------------------------------------------------------------------
239    Name: Matthias Blume (blume (at) tti - c (dot) org)
240    Date: 2005/01/24 17:40:00 CST
241    Tag: blume-20050124-mlyacc
242    Description:
243    
244    Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
245    problem with certain "as" patterns which previously got rendered
246    using incorrect syntax.
247    
248    ----------------------------------------------------------------------
249    Name: Matthias Blume (blume (at) tti - c (dot) org)
250    Date: 2005/01/18 12:00:00 CST
251    Tag: blume-20050118-profile
252    Description:
253    
254    Made time profiling code (interrupt handler) in runtime system aware
255    of new array representation.
256    
257    ----------------------------------------------------------------------
258    Name: Matthias Blume (blume (at) tti - c (dot) org)
259    Date: 2005/01/14 18:00:00 CST
260    Tag: blume-20050114-heap2exec
261    Description:
262    
263    Implemented new (but still experimental) heap2exec facility.  This is
264    tested under Mac OS X and should work under Linux (will test shortly).
265    It will probably also work on the Sparc (will test some time later).
266    
267      - removed old "HACKED_STANDALONE" hack from runtime
268    
269    To be able to test this, uncomment the request for "heap2asm" in
270    config/targets prior to installation.  (Notice that this is different
271    from "heap2exec" mentioned below.  Not a typo.)
272    
273    To perform an actual test, run the command
274    
275       $ bin/heap2exec heapfile execfile
276    
277    (You can put heap2exec on your shell's path.)
278    
279    For example, run
280    
281      $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
282    
283    This will create a standalone executable called "mly" which you can
284    then invoke directly as a command.
285    
286    ----------------------------------------------------------------------
287    Name: Matthias Blume (blume (at) tti - c (dot) org)
288    Date: 2005/01/07 11:44:00 CST
289    Tag: blume-20050107-mlstring
290    Description:
291    
292    fixed off-by-one error in ML_STRING macro (globals.c)
293    
294    ----------------------------------------------------------------------
295    Name: Matthias Blume (blume (at) tti - c (dot) org)
296    Date: 2004/12/23 18:00:00 CST
297    Tag: blume-20041223-santa
298    Description:
299    
300    Made ml-build script "smarter" (but only very little).
301    
302    ----------------------------------------------------------------------
303    Name: Matthias Blume (blume (at) tti - c (dot) org)
304    Date: 2004/12/21 15:05:00 CST
305    Tag: blume-20041221-longlong
306    Description:
307    
308    * Implemented access to signed and unsigned long long data in NLFFI.
309    (The parameter-passing part of the picture has not complete. But
310    data structure access seems to work.)
311    
312    * Fixed CM's incorrect assumption that the PPC is little-endian.
313      (On the Mac, it is big-endian.  And that's currently our only
314      PPC platform.)
315    
316    ----------------------------------------------------------------------
317    Name: Matthias Blume (blume (at) tti - c (dot) org)
318    Date: 2004/12/21 12:50:00 CST
319    Tag: blume-20041221-memory
320    Description:
321    
322    Some cleanup in the $c/memory.cm library: separated some concerns by
323    moving allocation code and memory access code each into their own
324    files.
325    
326    ----------------------------------------------------------------------
327    Name: Allen Leung (leunga (at) reservoir (dot) com)
328    Date: 2004/12/17 16:12:00 EST
329    Tag: leunga-20041217-cygwin-smlnj-home
330    Description:
331    
332        The Unix I/O library of SML/NJ on cygwin does not understand
333    Windows style pathname, so problems arise when SMLNJ_HOME is set
334    to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
335    to a POSIX pathname on cygwin.
336    
337    ----------------------------------------------------------------------
338    Name: Matthias Blume (blume (at) tti - c (dot) org)
339    Date: 2004/12/16 13:35:00 CST
340    Tag: Release_110_52
341    Description:
342    
343    Last-minute changes incorporated into 110.52.  Release tag moved.
344    
345    The changes:
346       - HashString.hashString' -> HashString.hashSubstring
347       - bug fix in UnivariateStats
348    
349    ----------------------------------------------------------------------
350    Name: Matthias Blume (blume (at) tti - c (dot) org)
351    Date: 2004/12/15 23:40:00 CST
352    Tag: blume-20041215-hashSubstring
353    Description:
354    
355    - HashString.hashString' -> HashString.hashSubstring
356    - corresponding changes in atom.sml
357    - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
358      added some comments
359    
360    ----------------------------------------------------------------------
361    Name: Matthias Blume (blume (at) tti - c (dot) org)
362    Date: 2004/12/15 15:30:00 CST
363    Tag: (Release_110_52)
364    Description:
365    
366    New working version (110.52).  NEW BOOTFILES!
367    
368    ----------------------------------------------------------------------
369    Name: Matthias Blume (blume (at) tti - c (dot) org)
370    Date: 2004/12/15 12:45:00 CST
371    Tag: blume-20041215-spaces
372    Description:
373    
374    More on the space problem (this time for Win32).
375    
376    ----------------------------------------------------------------------
377    Name: Matthias Blume (blume (at) tti - c (dot) org)
378    Date: 2004/12/14 17:30:00 CST
379    Tag: blume-20041214-spaces
380    Description:
381    
382    Hacked some of the scripts (in particular: the installer) to cope with
383    spaces in filenames a bit better.  But beware: the current "solution"
384    is likely still full of bugs and inherently incomplete.  (We need to
385    do away with those shell scripts for a comprehensive solution.)
386    
387    ----------------------------------------------------------------------
388    Name: Matthias Blume (blume (at) tti - c (dot) org)
389    Date: 2004/12/13 14:45:00 CST
390    Tag: blume-20041213-ml-makedepend
391    Description:
392    
393    Fixed bug in code for ml-makedepend.
394    
395    ----------------------------------------------------------------------
396    Name: Matthias Blume (blume (at) tti - c (dot) org)
397    Date: 2004/12/09 16:30:00 CST
398    Tag: blume-20041209-statistics
399    Description:
400    
401    Added two simple but potentially useful statistics modules to SML/NJ Library.
402    (See CHANGES file there.)
403    
404    ----------------------------------------------------------------------
405    Name: Matthias Blume (blume (at) tti - c (dot) org)
406    Date: 2004/12/01 16:50:00 CST
407    Tag: blume-20041201-atom
408    Description:
409    
410    smlnj-lib:
411    
412      Added function HashString.hashString' for substrings.
413      Hand-inlined CharVector.foldl into HashString (for speed).
414      Modified implementation of structure Atom to avoid extracting
415      strings from substrings unless necessary.
416      (Also see CHANGES file for smlnj-lib.)
417    
418    ----------------------------------------------------------------------
419    Name: Matthias Blume (blume (at) tti - c (dot) org)
420    Date: 2004/11/24 22:15:00 CST
421    Tag: blume-20041124-cml
422    Description:
423    
424    Made sure CML compiles when Position = Int64.
425    
426    ----------------------------------------------------------------------
427    Name: Matthias Blume (blume (at) tti - c (dot) org)
428    Date: 2004/11/24 14:30:00 CST
429    Tag: blume-20041124-position
430    Description:
431    
432    The compiler can now be compiled in a mode that makes structure Position
433    equal to Int64.  The default, however, is unchanged (Position = Int31)
434    for the time being.
435    
436    To enable 64-bit positions, use the following procedure:
437    
438    1. Start sml
439    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
440    3. Type
441          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
442    4. Run CMB.make() as usual.
443    
444    This is barely tested.  The only test so far was a little SML program
445    counting the number of characters in an 8-gigabyte file by
446    reading it character-by-character.  That test was successful.
447    
448    In support of 64-bit positions, a number of new functions have been
449    added to the runtime system.
450    
451    ----------------------------------------------------------------------
452    Name: Matthias Blume (blume (at) tti - c (dot) org)
453    Date: 2004/11/23 14:45:00 CST
454    Tag: blume-20041123-useFile
455    Description:
456    
457    Fixed a problem with unhelpful error messages related to problems with
458    .cm- or .sml files that appear as part of the sml command line.
459    
460    ----------------------------------------------------------------------
461    Name: Matthias Blume (blume (at) tti - c (dot) org)
462    Date: 2004/11/18 15:40:00 CST
463    Tag: Release_110_51
464    Description:
465    
466    New working version (110.51).  NEW BOOTFILES!
467    
468    ----------------------------------------------------------------------
469    Name: Matthias Blume (blume (at) tti - c (dot) org)
470    Date: 2004/11/18 15:35:00 CST
471    Tag: <none>
472    Description:
473    
474    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
475    Joemann).
476    
477    ----------------------------------------------------------------------
478    Name: Allen Leung (leunga (at) reservoir (dot) com)
479    Date: 2004/11/17 16:05:21 EST 2004
480    Tag: leunga-20041117-mlrisc-live-kill
481    Description:
482    
483       Added support for MLTree constructs LIVE and KILL
484    to all the architectures.
485    
486    ----------------------------------------------------------------------
487    Name: Matthias Blume (blume (at) tti - c (dot) org)
488    Date: 2004/11/13 00:20:00 CST
489    Tag: blume-20041113-versiontool
490    Description:
491    
492    - Stripped down the versiontool: It now only handles the version number.
493      The date string is generated at bootstrap time (during makeml).
494    
495    - In a previous commit, fixed a minor issue with how polyequal is being
496      translated.  In particular, the code now "looks through" abstractions.
497      This results in slightly fewer polyEqual warnings and hopefully slightly
498      more efficient code.  Important examples for where this matters are
499      the new int64 and word64 types.
500    
501    ----------------------------------------------------------------------
502    Name: Matthias Blume (blume (at) tti - c (dot) org)
503    Date: 2004/11/12 00:30:00 CST
504    Tag: blume-20041112-int64
505    Description:
506    
507    Structure Int64 fully hooked in.  (The implementation is not very
508    efficient, though.)
509    
510    ----------------------------------------------------------------------
511    Name: Matthias Blume (blume (at) tti - c (dot) org)
512    Date: 2004/11/11 17:30:00 CST
513    Tag: blume-20041111-more64
514    Description:
515    
516    All the pieces of Word64 are now there, with the exception of the
517    conversions from and to LargeWord. (Eventually these need to be identities,
518    but for the time being they don't even make sense because LargeWord is
519    32-bit wide.)
520    
521    Also started to add similar support for Int64, but major pieces of that
522    are still missing.
523    
524    ----------------------------------------------------------------------
525    Name: Matthias Blume (blume (at) tti - c (dot) org)
526    Date: 2004/11/11 00:15:00 CST
527    Tag: blume-20041111-word64
528    Description:
529    
530    Structure Word64 is now (almost) complete, word literals and patterns
531    seem to work.  There are a few odd pieces missing.  In particular,
532    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
533    at the moment.
534    
535    Making Word64 official would mean that LargeWord becomes Word64.  But
536    this requires extreme care because most word-word conversions have to
537    go through LargeWord, so making a mistake means loss of efficiency or
538    worse.  Eventually there will be a solution similar to (but actually
539    simpler than) what I did with IntInf.
540    
541    ----------------------------------------------------------------------
542    Name: Matthias Blume (blume (at) tti - c (dot) org)
543    Date: 2004/11/10 18:12:00 CST
544    Tag: blume-20041110-64bit
545    Description:
546    
547    More 64-bit hacking (but still not even half-way there yet).
548    Also, some assorted improvements to the handling of 8-bit words.
549    
550    ----------------------------------------------------------------------
551    Name: Matthias Blume (blume (at) tti - c (dot) org)
552    Date: 2004/11/09 17:50:00 CST
553    Tag: <none>
554    Description:
555    
556    Started adding some infrastructure for supporting 64-bit int- and
557    word-types.  (Still in its very early stages.)
558    
559    ----------------------------------------------------------------------
560    Name: Matthias Blume (blume (at) tti - c (dot) org)
561    Date: 2004/10/28 10:45:00 CDT
562    Tag: Release_110_50
563    Description:
564    
565    New working version (110.50).  NEW BOOTFILES!
566    
567                 =====================
568    
569    Also:
570    
571      - Changend config/srcarchiveurl from a file just
572        containing the URL string into a file containing
573        shell script code.  The code has access to the $VERSION variable.
574      - Made corresponding changes to config/install.sh and config/unpack.
575      - Default contents of config/srcarchiveurl uses $VERSION and
576        normally does not have to be edited to reflect a version change.
577    
578      (As a result, a version change can be done by just editing
579       config/version, the rest is now automatic.)
580    
581    ----------------------------------------------------------------------
582    Name: Matthias Blume (blume (at) tti - c (dot) org)
583    Date: 2004/10/27 17:50:00 CDT
584    Tag: blume-20041027-btrace-msg
585    Description:
586    
587    BackTrace.monitor now also reports the source of the exception that
588    triggered the trace.
589    
590    ----------------------------------------------------------------------
591    Name: Matthias Blume (blume (at) tti - c (dot) org)
592    Date: 2004/10/27 17:20:00 CDT
593    Tag: blume-20041027-x86-c-calls
594    Description:
595    
596    This is the HISTORY entry for two earlier commits, both concerning
597    the x86 c-calls code in MLRISC:
598    
599      - added a missing LOAD in the code that deals with struct arguments
600      - made sure the caller does not add the wrong number of bytes to the
601        stack pointer after a call of a function returning a struct
602        (the callee already pops the implicit argument which points to
603         the space reserved for the result)
604    
605    ----------------------------------------------------------------------
606    Name: Allen Leung (leunga (at) reservoir (dot) com)
607    Date: 2004/10/24 14:00:00 EST
608    Tag: leunga-20041024-x86-gas-fucomip
609    Description:
610    
611       John discovered a bug in the syntax of fucomip.
612       The opcodes FU?COMIP? have been changed to
613    
614           fu?comip? %st(i), %st
615    
616    ----------------------------------------------------------------------
617    Name: Matthias Blume (blume (at) tti - c (dot) org)
618    Date: 2004/10/20 15:06:00 CDT
619    Tag: blume-20041020-standalone-backtrace
620    Description:
621    
622    Added a mechanism for getting back-trace information from standalone
623    programs.  Here is how it works:
624    
625       1. The part of the program from which you want to get backtrace
626          information (usually the whole program) should be wrapped with
627          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
628          main program could be modified from something like
629    
630             fun main (pgm, args) = ...
631    
632         to
633    
634            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
635    
636       2. To be able to access BackTrace.monitor, you have to add
637          library $smlnj-tdp/plugins.cm to the .cm file that contains your
638          main function.
639    
640       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
641          CM might have created in the past for your project).
642    
643       4. Build the system using this command line:
644    
645             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
646                      myprog.cm MyProg.main myprog
647    
648          instead of the usual
649    
650             ml-build myprog.cm MyProg.main myprog
651    
652    I changed a library name:
653    
654       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
655    
656    New libraries:
657    
658       $smlnj-tdp/back-trace.cm
659          -- when loaded causes the back-trace plugin to be installed
660       $smlnj-tdp/coverage.cm
661          -- when loaded causes the coverage plugin to be installed
662    
663    ----------------------------------------------------------------------
664    Name: Matthias Blume (blume (at) tti - c (dot) org)
665    Date: 2004/10/18 16:45:00 CDT
666    Tag: blume-20041018-groupowner
667    Description:
668    
669    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
670    
671    Eliminated group owner specs from .cm files throughout the source tree.
672    
673    ----------------------------------------------------------------------
674    Name: Matthias Blume (blume (at) tti - c (dot) org)
675    Date: 2004/10/15 15:45:00 CDT
676    Tag: blume-20041015-coverage
677    Description:
678    
679    * Test coverage tool added!
680    * Further reorganization of tracing-, debugging-, and profiling support:
681    
682        - moved original BTImp -- now called BackTrace -- into a separate
683          library called $/trace-debug-profile.cm
684        - eliminated all mentions of BTrace from SMLofNJ.Internals
685        - only the instrumentation mechanism is now left in the compiler proper
686        - BackTrace module is a plugin which is NOT plugged in by default
687        - Coverage module is another such plugin
688    
689    To get the benefits of any of these plugin modules, the code in
690    question must be compiled with tdp instrumentation turned on.  This
691    can be done by setting SMLofNJ.Internals.TDP.mode to true.
692    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
693    
694    Plugins are selected at link time.  (Pre-compiled instrumented code
695    can be re-loaded with different plugins in effect.)  When an
696    instrumented module is linked, whatever plugins are at that time
697    enabled will come into effect for that module.
698    
699    To enable the back-trace plugin, load library $/trace-debug-profile.cm
700    and invoke BackTrace.install() (e.g., from the interactive prompt).
701    To enable the coverage plugin, load the same library and invoke
702    Coverage.install().
703    
704    Back-traces are generated automatically on uncaught exceptions and
705    when the code in question explicitly invokes BackTrace.trigger().
706    
707    Coverage (and execution frequency-) information must be queried
708    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
709    
710    ----------------------------------------------------------------------
711    Name: Matthias Blume (blume (at) tti - c (dot) org)
712    Date: 2004/10/14 17:40:00 CDT
713    Tag: blume-20041014-tdp-core
714    Description:
715    
716    Snapshot of a significant overhaul of how the trace/debug/profile support
717    is hooked into the system (specifically: Core and SMLofNJ.Internals).
718    
719    ----------------------------------------------------------------------
720    Name: Matthias Blume (blume (at) tti - c (dot) org)
721    Date: 2004/10/13 16:34:00 CDT
722    Tag: blume-20041013-tdp
723    Description:
724    
725    Some rationalization of names:
726    
727       structure BTrace -> structure TDPInstrument
728       etc.
729    
730    This is is preparation of using the original back-trace
731    instrumentation for other purposes.  "TDP" stands for
732    Trace/Debug/Profile.
733    
734    The control flag controlling whether instrumentation is on or off is now
735    registered under a different name, so instead of running sml as
736    
737     sml -Cinstrument.btrace-mode=true
738    
739    one has to say
740    
741     sml -Ctdp.instrument=true
742    
743    ----------------------------------------------------------------------
744    Name: Matthias Blume (blume (at) tti - c (dot) org)
745    Date: 2004/10/11 16:37:00 CDT
746    Tag: blume-20041011-regions
747    Description:
748    
749    Made some minor modifications to elabcore.sml to have source regions
750    be propagated more tightly -- resulting in better (i.e., smaller)
751    regions being reported in error- and debug messages.
752    
753    ----------------------------------------------------------------------
754    Name: Matthias Blume (blume (at) tti - c (dot) org)
755    Date: 2004/10/08 22:50:00 CDT
756    Tag: blume-20041008-cmkw
757    Description:
758    
759    Fixed handling of keywords in .cm files: After seeing "is" the lexer
760    treats subsequent occurrences of "group", "library", "source", "is",
761    "*", and "-" as ordinary identifiers rather than keywords.
762    
763    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
764    is supposed to accept a tool argument called "source", but this did
765    not work because of the clash with the keyword.
766    
767    ----------------------------------------------------------------------
768    Name: Matthias Blume (blume (at) tti - c (dot) org)
769    Date: 2004/10/07 16:00:00 CDT
770    Tag: blume-20041007-cleanup
771    Description:
772    
773    Assorted cleanup work:
774    
775      - got rid of intstrmap in favor of using the library's
776        hash table implementation
777      - threw out most of the pathnames stuff, as it was not used anyway
778      - simplified tokentable implementation
779      - fixed some minor spelling errors
780    
781    ----------------------------------------------------------------------
782    Name: Matthias Blume (blume (at) tti - c (dot) org)
783    Date: 2004/10/06 15:15:15 CDT
784    Tag: blume-20041006-handler
785    Description:
786    
787    Cleaned up the absyn to reflect the invariant that HANDLE always
788    carries a FNexp as part of the type definition.  This eliminates some
789    superfluous sanity checks at runtime down the road.
790    
791    Some minor cleanup of the btrace code.
792    
793    ----------------------------------------------------------------------
794    Name: Matthias Blume (blume (at) tti - c (dot) org)
795    Date: 2004/10/01 10:20:30 CDT
796    Tag: blume-20041001-slave
797    Description:
798    
799    Added hack to make slave mode work in the presence of the version
800    tool.  (Still, since the master does two passes over the code for
801    CMB.make, the release number gets bumped twice when slaves are
802    attached.  I don't know if this is worth fixing...)
803    
804    ----------------------------------------------------------------------
805    Name: Matthias Blume (blume (at) tti - c (dot) org)
806    Date: 2004/09/30 10:55:00 CDT
807    Tag: blume-20040930-version
808    Description:
809    
810    * Moved the "version" magic into its own little library under
811      src/system/smlnj/internal.  This avoids expensive reconstruction of
812      a stable src/compiler/core.cm.
813    
814    * At the same time, structure CompilerVersion is now known as
815      structure SMLNJVersion.
816    
817    * Arranged for the version tool to NOT kick in when rebuilding the system
818      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
819      Also, loading the versiontool does not work when rebuilding the system
820      because CM is not properly initialized at that time.
821    
822    ----------------------------------------------------------------------
823    Name: Matthias Blume (blume (at) tti - c (dot) org)
824    Date: 2004/09/29 14:00:00 CDT
825    Tag: blume-20040929-autoversion
826    Description:
827    
828    Implemented some CM magic to have
829        file src/compiler/TopLevel/main/version.sml
830    generated automagically.
831    The version is taken from two files: config/version and config/release.
832    The first is expected to contain a two-part version number such as 110.49.
833    The second should contain a single number, but it may be missing.
834    
835    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
836    time the version tool is loaded (which is the first time you say CMB.make),
837    then the tool will increment the value stored in config/release every
838    time CMB.make is invoked.
839    
840    The binfile format is now insensitive to anything beyond the first
841    two components of a version number, so bumping the release does not render
842    binfiles incompatible.  Auto-bumping can be used to keep track of versions
843    during development without invalidating existing binfiles.
844    
845    In any case, every CMB.make updates the date information in version.sml.
846    (This is the date that is printed in the banner.)
847    
848    ----------------------------------------------------------------------
849    Name: Matthias Blume (blume (at) tti - c (dot) org)
850    Date: 2004/09/28 10:53:00 CDT
851    Tag: blume-20040928-controls
852    Description:
853    
854    Some cleanup of the controls code.
855    
856    ----------------------------------------------------------------------
857    Name: Matthias Blume (blume (at) tti - c (dot) org)
858    Date: 2004/09/27 22:08:00 CDT
859    Tag: blume-20040927-controls
860    Description:
861    
862    Added two pieces of functionality to the Controls interface:
863    
864    1.   val save'restore: 'a control -> unit -> unit
865    
866       grabs the current value of the control in stage 1 and restores it
867       in stage 2.
868    
869    2.   val set' : 'a control * 'a -> unit -> unit
870    
871       stores the given value into the control in stage 2 (i.e., delayed)
872       but does all error checking in stage 1.
873       (This is for string controls that need to do parse their argument --
874        something that might fail.  In some cases, notably in CM, one
875        already knows the intended argument but wants to delay the actual
876        assignment until a time when error recovery would be more difficult.)
877    
878    Changed the handling of controls in tool arguments to classes "sml" and
879    "lazysml":
880       - use Controls.save'restore as a more robust way of restoring the
881         old value (in particular: without having to re-parse the string)
882       - use controls to handle the "overload" keyword in the init group
883         (I believe this change actually fixes a long-standing obscure bug.)
884    
885    ----------------------------------------------------------------------
886    Name: Matthias Blume (blume (at) tti - c (dot) org)
887    Date: 2004/09/27 17:00:00 CDT
888    Tag: blume-20040927-lazysml
889    Description:
890    
891    Added a new tool class called "lazysml" to CM's tool chest.  The only
892    difference to "sml" is that compilation is done with Control.lazysml
893    set to true.  A source of class "lazysml" is automatically recognized
894    by a file name suffix of ".lml".
895    
896    In addition to the above feature, the original class "sml" now also
897    supports a tool argument "lazy" which has the same effect.  As a
898    result, the following three lines are equivalent:
899    
900        foo.sml : lazysml
901        foo.sml : sml (lazy)
902        foo.sml (lazy)
903    
904    The setting goes into effect both during parsing and during
905    compilation.  The original setting is restored right after parsing and
906    after compilation, respectively.
907    
908    In addition to all the above, there is also a general mechanism to set
909    ANY of the "controls" that are available at the command line via
910    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
911    apply as well. (In fact, "lazy" is implemented as a special case of
912    the general mechanism.)
913    
914    The .cm file syntax uses a new keyword tool argument called "with".
915    There are several ways of indicating the desired settings:
916    
917        foo.sml (with:parser.quotations=true)
918        foo.sml (with:(name:parser.quotations value:true))
919        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
920        foo.sml (with:(name1=value1 name2=value2 ...))
921        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
922    
923        etc.
924    
925    Another possible abbreviation is to leave out the =v or value:v part
926    if the name refers to a boolean control (in which case the value is
927    taken to be true).  Thus, one could get lazy sml also by saying:
928    
929        foo.sml (with:parser.lazy-keyword=true)
930        foo.sml (with:parser.lazy-keyword)
931        foo.sml (with:(name:parser.lazy-keyword value:true))
932        foo.sml (with:(name:parser.lazy-keyword))
933    
934    ----------------------------------------------------------------------
935    Name: Matthias Blume (blume (at) tti - c (dot) org)
936    Date: 2004/09/24 16:22:00 CDT
937    Tag: blume-20040924-ppc-long-branch
938    Description:
939    
940    Turned message about "emiting long form of branch" off by default.
941    Added a control flag to turn it back on when desired.
942    
943    ----------------------------------------------------------------------
944    Name: Matthias Blume (blume (at) tti - c (dot) org)
945    Date: 2004/09/24 16:05:00 CDT
946    Tag: blume-20040924-rounding
947    Description:
948    
949    Applied patch for setting rounding modes under Mac OS X.  Thanks to
950    Melissa O'Neill for providing the code!
951    
952    ----------------------------------------------------------------------
953    Name: Matthias Blume (blume (at) tti - c (dot) org)
954    Date: 2004/09/23 17:30:00 CDT
955    Tag: blume-20040923-envvars
956    Description:
957    
958    1. Changed definition of type ControlRegistry.registry_tree to
959       include control_info (i.e., the name of the controlling environment
960       variable).
961    
962    2. Added command-line flags -e and -E to print the names of environment
963       variables that can be used to control internal settings.
964       (This uses the new API mentioned in 1.)
965    
966    ----------------------------------------------------------------------
967    Name: Matthias Blume (blume (at) tti - c (dot) org)
968    Date: 2004/09/13 16:50:00 CDT
969    Tag: Release_110_49
970    Description:
971    
972    New working version (110.49).  NEW BOOTFILES!
973    
974    ----------------------------------------------------------------------
975    Name: Matthias Blume
976    Date: 2004/09/13 16:20:00 CDT
977    Tag: blume-20040913-config-mlrisc
978    Description:
979    
980    Put target "mlrisc" back into the default list.
981    (There is no harm in having it, and some users have expressed their
982    wish to have "mlrisc" included by default.)
983    
984    ----------------------------------------------------------------------
985    Name: John Reppy
986    Date: 2004/09/13
987    Tag: jhr-20040913-signals
988    Description:
989    
990    Fixed the signal masking code to properly nest mask/unmask operations
991    on a per-signal basis.
992    
993    ----------------------------------------------------------------------
994    Name: Matthias Blume (blume (at) tti - c (dot) org)
995    Date: 2004/09/08 13:20:00 CDT
996    Tag: blume-20040908-heap-magic
997    Description:
998    
999    Bumped the heap macig number to 0x09082004 to account for the changed
1000    layout of the ML frame under MacOS X.
1001    
1002    ----------------------------------------------------------------------
1003    Name: Allen Leung (leunga (at) reservoir (dot) com)
1004    Date: 2004/09/03 11:26:00 EST
1005    Tag: leunga-20040903-cygwin-install
1006    Description:
1007    
1008       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
1009    The Cygwin runtime is turned on by setting the environment
1010    variable SMLNJ_CYGWIN_RUNTIME to 1.
1011    
1012    ----------------------------------------------------------------------
1013    Name: Matthias Blume (blume (at) tti - c (dot) org)
1014    Date: 2004/08/31 17:15:00 CDT
1015    Tag: blume-20040831-core
1016    Description:
1017    
1018    Added some exports to src/compiler/core.cm upon request by J. Joemann.
1019    
1020    ----------------------------------------------------------------------
1021    Name: Matthias Blume (blume (at) tti - c (dot) org)
1022    Date: 2004/08/30 17:55:00 CDT
1023    Tag: blume-20040830-installer
1024    Description:
1025    
1026    Upon request by Johannes Joemann:
1027    
1028    - improved ML code of installer to fall back to coping when renaming
1029      fails (i.e., when source and target are on different file systems);
1030      the code compiles but has yet to be tested in anger
1031    - removed mlrisc from list of default targets (config/targets)
1032    
1033    ----------------------------------------------------------------------
1034    Name: Matthias Blume (blume (at) tti - c (dot) org)
1035    Date: 2004/08/27 17:20:00 CDT
1036    Tag: blume-20040827-ptreql
1037    Description:
1038    
1039    Added ptreql primop to structure InlineT (upon request from Larry
1040    Paulson).
1041    
1042    ----------------------------------------------------------------------
1043    Name: Allen Leung (leunga (at) reservoir (dot) com)
1044    Date: 2004/08/15 21:21:00 EST
1045    Tag: leunga-110_48-udgraph
1046    Description:
1047    
1048    Another bug fix from Carl Hauser:
1049    
1050    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
1051    > 48c48
1052    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
1053    > e::L)
1054    > ---
1055    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
1056    > Without this, any deletion of an edge in an undirected graph does severe
1057    > violence to the graph.
1058    
1059    ----------------------------------------------------------------------
1060    Name: Allen Leung (leunga (at) reservoir (dot) com)
1061    Date: 2004/08/10 23:35:00 EST
1062    Tag: leunga-110_48-ppc
1063    Description:
1064    
1065       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
1066    
1067    ----------------------------------------------------------------------
1068    Name: Matthias Blume (blume (at) tti - c (dot) org)
1069    Date: 2004/08/10 12:00:00 CDT
1070    Tag: Release_110_48
1071    Description:
1072    
1073    New working version (110.48).  NEW BOOTFILES!
1074    
1075    ----------------------------------------------------------------------
1076    Name: Allen Leung (leunga (at) reservoir (dot) com)
1077    Date: 2004/08/09 12:21:00 EST
1078    Tag: leunga-110_47-dijsktra
1079    Description:
1080    
1081    Bug fix from Carl Hauser:
1082    
1083    single_source_shortest_paths in dijkstra.sml was observed to get wrong
1084    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1085    
1086    The problem is that following the expression A.update(dist,s,Num.zero)
1087    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1088    
1089    ----------------------------------------------------------------------
1090    Name: Matthias Blume (blume (at) tti - c (dot) org)
1091    Date: 2004/08/06 18:10:00 CDT
1092    Tag: blume-20040806-cmdline
1093    Description:
1094    
1095    Fiddled with handling of command-line options:
1096    
1097      * sml now quits after processing the command line
1098        if -H, -S, -h<n>, or -s<n> appears as the last
1099        command-line argument
1100      * a new option -q terminates the session when encountered on
1101        the command line; subsequent arguments will be ignored
1102      * bug fixes: short (erroneous) arguments are no longer ignored
1103        completely
1104    
1105    ----------------------------------------------------------------------
1106    Name: Allen Leung (leunga (at) reservoir (dot) com)
1107    Date: 2004/08/04 18:17:00 EST
1108    Tag: leunga-110_47-ppc-ibm-asm
1109    Description:
1110    
1111      - Added minimal IBM assembly syntax support for PowerPC.
1112    
1113      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1114        been altered in the runtime.  We need an automatic way to keep the file
1115        in sync.
1116    
1117    ----------------------------------------------------------------------
1118    Name: Matthias Blume (blume (at) tti - c (dot) org)
1119    Date: 2004/08/04 14:00:00 CDT
1120    Tag: Release_110_47
1121    Description:
1122    
1123    New working version (110.47).  NEW BOOTFILES!
1124    
1125    ----------------------------------------------------------------------
1126    Name: Matthias Blume (blume (at) tti - c (dot) org)
1127    Date: 2004/08/03 14:25:00 CDT
1128    Tag: blume-20040803-callingconv
1129    Description:
1130    
1131    Added low-level support for choosing C calling conventions by
1132    twiddling the type of rawccall.  (See
1133    src/compiler/Semant/types/cproto.sml for details.)
1134    
1135    ----------------------------------------------------------------------
1136    Name: Matthias Blume (blume (at) tti - c (dot) org)
1137    Date: 2004/08/02 15:55:00 CDT
1138    Tag: blume-20040802-backout
1139    Description:
1140    
1141    Backed out of change to win32-filesys.c.  The earlier patch to
1142    get_file_time caused CM to produce files with the wrong time stamp.
1143    
1144    ----------------------------------------------------------------------
1145    Name: Matthias Blume (blume (at) tti - c (dot) org)
1146    Date: 2004/08/02 14:45:00 CDT
1147    Tag: blume-20040802-nlffi-win32
1148    Description:
1149    
1150    Added NLFFI support for Win32, adapted from a patch provided by David
1151    Hansel.  This is currently completely untested.  Also, the issue
1152    concerning stdcall vs. ccall is still unresolved.
1153    
1154    ----------------------------------------------------------------------
1155    Name: Matthias Blume (blume (at) tti - c (dot) org)
1156    Date: 2004/07/30 17:55:00 CDT
1157    Tag: blume-20040730-various
1158    Description:
1159    
1160    Gearing up towards 110.47...
1161    
1162    - various minor bugfixes to ml-nlffigen
1163    - a beginning of a manual for nlffi
1164    
1165    - eliminated 'export name=value' in config/install.sh as this does
1166      not work with certain versions of /bin/sh
1167      (Thanks to David King at Motorola for catching this.)
1168    
1169    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1170      - added a test for tm==NULL to gmtime.c and localtime.c
1171      - applied patch for incorrect GetFileTime under win32
1172      - toSeconds -> toMilliseconds in Win32/win32-process.sml
1173    
1174    ----------------------------------------------------------------------
1175    Name: Matthias Blume (blume (at) tti - c (dot) org)
1176    Date: 2004/07/21 18:20:00 CDT
1177    Tag: blume-20040721-nlffigen
1178    Description:
1179    
1180    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1181      for a typedef to an incomplete type, but leave out the "typ" member.
1182      (This is just for consistency.)
1183    - Started to produce what is supposed to become better (i.e., comprehensive)
1184      documentation of what ml-nlffigen does and produces.
1185    
1186    ----------------------------------------------------------------------
1187    Name: Matthias Blume (blume (at) tti - c (dot) org)
1188    Date: 2004/07/14 16:25:00 CDT
1189    Tag: blume-20040714-union
1190    Description:
1191    
1192    Added C_UNION to c-calls/c-types.sml and updated the machinery
1193    (ml-nlffigen, cproto.sml) that conveys C function interface
1194    information to the code generator.
1195    
1196    However, the actual architecture-specific implementation of function
1197    arguments and results that are C unions is still not implemented.
1198    
1199    ----------------------------------------------------------------------
1200    Name: Allen Leung (leunga (at) reservoir (dot) com)
1201    Date: 2004/07/14 14:38:00 EST
1202    Tag: leunga-110_46_1-ppc-lwzu
1203    Description:
1204    
1205        Added these instructions to the PowerPC architecture:
1206        LBZU(X), LHZU(X), LWZU(X),
1207        STWU(X), STFDU, STFSU
1208    
1209        etc...
1210    
1211        Note: I haven't added their instruction encoding into the description.
1212    
1213    ----------------------------------------------------------------------
1214    Name: Allen Leung (leunga (at) reservoir (dot) com)
1215    Date: 2004/07/13 15:04:00 EST
1216    Tag: leunga-110_46_1-ppc-lwarx
1217    Description:
1218    
1219       Added the two instructions LWARX and STWCX to the PowerPC
1220    instruction set.
1221    
1222       A (untested) rewrite of loop-structure.sml.  The old version
1223    is completely broken.
1224    
1225    ----------------------------------------------------------------------
1226    Name: Matthias Blume (blume (at) tti - c (dot) org)
1227    Date: 2004/07/13 13:50:00 CDT
1228    Tag: blume-20040713-nlffi
1229    Description:
1230    
1231    - use paramAlloc to report c-calls with too many arguments
1232      (for PPC version where parameter area is pre-allocated)
1233    - added ccall_maxargspace to machspec (to implement the above)
1234    - made "make" commend in CM's "make" tool configurable
1235    - added option (default: on) for passing the name of the SML/NJ's "bin"
1236      directory to "make"; the call looks like this:
1237    
1238         make <options> SMLNJ_BINDIR=<dir> <target>
1239    
1240      This can be used by the Makefile to, e.g., pick the "right" version
1241      of ml-nlffigen.
1242    - minor code tweaks
1243    
1244    ----------------------------------------------------------------------
1245    Name: Matthias Blume (blume (at) tti - c (dot) org)
1246    Date: 2004/07/12 22:50:00 CDT
1247    Tag: blume-110_46_1-macosx-nlffi
1248    Description:
1249    
1250    NLFFI under Mac OS X now working (sort of).  This is largely untested,
1251    though.
1252    
1253    Note:  1. You have to make a new, clean build of the runtime system.
1254           2. There are new BOOTFILES, you have to use them!
1255              (Doing the bootstrap process yourself would be *very* painful!
1256               If you absolutely have to do it, build the system under
1257               a different architecture and then cross-compile.)
1258    
1259    Version bumped to 110.46.1 to account for runtime data format changes.
1260    
1261    ----------------------------------------------------------------------
1262    Name: Matthias Blume (blume (at) tti - c (dot) org)
1263    Date: 2004/06/18 14:30:00 CDT
1264    Tag: blume-20040618-unix
1265    Description:
1266    
1267    Changed the implementation of structure Unix so that the same stream
1268    is returned every time one of the {text,bin}{In,Out}streamOf functions
1269    is invoked on the same proc.  This is not what the spec currently
1270    says -- although IMO it arguably should.  (See discussion below.)
1271    
1272    ----------------------------------------------------------------------
1273    Name: Matthias Blume (blume (at) tti - c (dot) org)
1274    Date: 2004/06/17 18:15:00 CDT
1275    Tag: Release_110_46
1276    Description:
1277    
1278    New working version (110.46).  NEW BOOTFILES!
1279    
1280    ----------------------------------------------------------------------
1281    Name: Matthias Blume (blume (at) tti - c (dot) org)
1282    Date: 2004/06/17 17:20:00 CDT
1283    Tag: blume-20040617-timer-unix
1284    Description:
1285    
1286    Changed the interface of structures Timer and Unix to match the most
1287    recent Basis spec.
1288    
1289    In the case of Unix there still seems to be an open/weird issue:
1290    
1291       The {text,bin}{In,Out}streamOf functions are supposed to create
1292       fresh streams whenever they are called -- as opposed to have them
1293       return the same stream every time.  This design is supposed to
1294       prevent space leaks caused by proc values hanging on to streams.
1295    
1296       The reap function, on the other hand, is supposed to close the
1297       streams.  This cannot be done without having a handle on the
1298       stream in proc after all...
1299    
1300    I took the liberty to implement the following stopgap solution:
1301    
1302      The proc value hangs on to the most recently created stream(s).
1303      Reap closes those.  If either or both of the two streams hadn't
1304      been created at all yet, then reap will close the corresponding
1305      file descriptors directly.
1306    
1307    PS: I don't understand the original space leak argument anymore.  If
1308    a proc hangs on to the imperative stream, then I/O operations on those
1309    will advance the state of the cached stream and avoid the space leak.
1310    
1311    ----------------------------------------------------------------------
1312    Name: Matthias Blume (blume (at) tti - c (dot) org)
1313    Date: 2004/05/28 16:45:00 CDT
1314    Tag: blume-20040528-basis
1315    Description:
1316    
1317    Added signature PACK_REAL and exported functor PrimIO.
1318    
1319    ----------------------------------------------------------------------
1320    Name: Matthias Blume (blume (at) tti - c (dot) org)
1321    Date: 2004/05/25 16:00:00 CDT
1322    Tag: blume-20040525-group-owner
1323    Description:
1324    
1325    CM now ignores (but still accepts) the "owner" information in group
1326    descriptions.  The owner of a group is its next enclosing
1327    library. Each group must have a unique owner.  (There is a virtual
1328    "toplevel" library that owns groups which are not nested within a real
1329    library.)  Previously, each group had to explicitly declare its owner,
1330    and CM would check that such a declaration is correct.  The new scheme
1331    is to have CM check that for each group there is precisely one owning
1332    library.
1333    
1334    The advantage of the new scheme is that the programmer no longer needs
1335    to maintain the somewhat annoying owner information.  The downside is
1336    that CM cannot enforce the ownership rule across multiple runs of
1337    CM.make.  Fortunately, enclosing the same group in two different
1338    libraries A and B which are not part of the same program does not
1339    cause real problems.
1340    
1341    ----------------------------------------------------------------------
1342    Name: Matthias Blume (blume (at) tti - c (dot) org)
1343    Date: 2004/05/20 16:00:00 CDT
1344    Tag: blume-20040520-win32
1345    Description:
1346    
1347    Made the win32 version work again.  (Strangely, a misplaced comma had
1348    slipped into win32-process.c which prevented the runtime from being
1349    compiled correctly.)
1350    
1351    Also, included a minor addition to ml-build.bat analogous to what was
1352    done in blume-20040519-ml-build.
1353    
1354    ----------------------------------------------------------------------
1355    Name: Matthias Blume (blume (at) tti - c (dot) org)
1356    Date: 2004/05/19 22:10:00 CDT
1357    Tag: blume-20040519-ml-build
1358    Description:
1359    
1360    Arranged for ml-build to clean up after itself a little bit better.
1361    The script generates a temporary SML source file and compiles it using
1362    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1363    gets rid of those at the end, so they don't accumulate under .cm.
1364    
1365    This required a minor change to install.sh because the name of the
1366    metadata directory (default: .cm) is actually configurable at
1367    installation time.
1368    
1369    ----------------------------------------------------------------------
1370    Name: Matthias Blume (blume (at) tti - c (dot) org)
1371    Date: 2004/05/18 15:50:00 CDT
1372    Tag: blume-20040518-mkreader
1373    Description:
1374    
1375    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1376    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1377    
1378    ----------------------------------------------------------------------
1379    Name: Matthias Blume (blume (at) tti - c (dot) org)
1380    Date: 2004/05/11 14:35:00 CDT
1381    Tag: blume-20040511-win32sock
1382    Description:
1383    
1384    Added previously missing support for many socket-related functions
1385    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1386    for the voluminous patch!
1387    
1388    (I have not tested this patch under win32 yet.)
1389    
1390    Here is David's e-mail:
1391    
1392    Hi,
1393    
1394    Attached to this email you find a diff against sml/nj 110.45
1395    that will enable socket support under Windows.
1396    
1397    To apply the patch (using unix or cygwin)
1398    1) gunzip runtime.diff.gz
1399    2) "cd" into "src/runtime" in the source tree of a fresh
1400       110.45 installation.
1401    3) patch -p 1 < [your/path/to]runtime.diff
1402    
1403    The code compiles fine but has NOT yet been extensively tested.
1404    I only ran a few tests for basic socket client functionality
1405    (which worked fine).  Especially the functions that use ioctl
1406    are not tested at all and might not work (see below).
1407    
1408    I implemented this since we want to move to a newer version of sml/nj
1409    but need socket support in order to use it.  This is the first time I
1410    even had a look at the sml/nj source,  so please review my changes
1411    before making this part of the distribution!  Here are a few issues
1412    that I think might be better for someone to solve who is more
1413    familiar with the sml/nj source (and socket programming):
1414    
1415    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1416      exception since I could not figure out what the windows equivalent
1417      of these functions is
1418    
1419    - In sockets-osdep.h there are a some #include statements that are
1420      only used in a few files that include sockets-osdep.h
1421    
1422    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1423      does not process its return value since I don't know how to
1424      report an error upwards.
1425    
1426    - It would probably be good to have a call to WSACleanup() when
1427      the library is unloaded (if there is such a possibility).
1428      Otherwise I think Windows will take care of this automatically
1429      when the process finishes.
1430    
1431    - I used ioctlsocket() as a replacement for ioctl() but I have
1432      no idea if that is actually the proper replacement on Windows.
1433    
1434    - All these issues are marked in the code by "FIXME" comments.
1435    
1436    We use sml/nj extensively in our products and are quite happy
1437    with it.  I hope this contribution will help you.
1438    
1439    Keep up the good work!
1440    
1441    David
1442    
1443    ----------------------------------------------------------------------
1444    Name: Matthias Blume (blume (at) tti - c (dot) org)
1445    Date: 2004/05/11 14:20:00 CDT
1446    Tag: blume-20040511-installml
1447    Description:
1448    
1449    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1450    the patch.)
1451    
1452    ----------------------------------------------------------------------
1453    Name: Matthias Blume (blume (at) tti - c (dot) org)
1454    Date: 2004/05/11 14:05:00 CDT
1455    Tag: blume-20040511-nlffi-netbsd
1456    Description:
1457    
1458    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1459    the patch.)
1460    
1461    ----------------------------------------------------------------------
1462    Name: Matthias Blume (blume (at) tti - c (dot) org)
1463    Date: 2004/05/11 12:05:00 CDT
1464    Tag: blume-20040511-exports
1465    Description:
1466    
1467    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1468    export lists in compiler-related .cm-files.
1469    
1470    ----------------------------------------------------------------------
1471    Name: Matthias Blume (blume (at) tti - c (dot) org)
1472    Date: 2004/05/11 11:35:00 CDT
1473    Tag: blume-20040511-allsource
1474    Description:
1475    
1476    The installer now honors the "src-smlnj" target again, although its meaning
1477    has changed from "all sources required for the compiler" to "all sources
1478    the installer knows about".  In other words, if you enable "src-smlnj"
1479    in the "targets" file, then the installer will pull in sources for
1480    everything.  (Notice that this refers to source code only.  Compiled
1481    code is still only installed for modules that were requested explicitly
1482    or which are required for other modules that were requested explicitly.)
1483    
1484    ----------------------------------------------------------------------
1485    Name: Matthias Blume (blume (at) tti - c (dot) org)
1486    Date: 2004/04/23 17:40:00 CDT
1487    Tag: blume-20040423-ieee-scan
1488    Description:
1489    
1490    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1491    in the exponent calculation we get INF or ZERO (depending on the mantissa
1492    and the sign of the exponent).
1493    
1494    ----------------------------------------------------------------------
1495    Name: Matthias Blume (blume (at) tti - c (dot) org)
1496    Date: 2004/04/23 10:40:00 CDT
1497    Tag: blume-20040423-ml-build
1498    Description:
1499    
1500    The ml-build script now terminates with a non-0 status when something
1501    goes wrong.
1502    
1503    ----------------------------------------------------------------------
1504    Name: Matthias Blume (blume (at) tti - c (dot) org)
1505    Date: 2004/04/22 16:35:00 CDT
1506    Tag: blume-20040422-Option
1507    Description:
1508    
1509    Made exception Option to be the same as exception Option.Option
1510    (as it should be).
1511    
1512    ----------------------------------------------------------------------
1513    Name: Allen Leung (leunga (at) reservoir (dot) com)
1514    Date: 2004/03/19 14:40:00 EST
1515    Tag: leunga-20040319-cygwin-nlffi
1516    Description:
1517    
1518    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1519    of SML/NJ.  The problem is that
1520    
1521       lib = dlopen(NULL, ...)
1522       f   = dlsym(lib, "malloc");
1523    
1524    does not work on Windows unless we explicitly export symbols
1525    such as 'malloc' during linking.  We fixed this by explicitly
1526    exporting the required symbols with the magic gcc incantation:
1527    
1528        -Wl,--export-all cygwin.def
1529    
1530    where cygwin.def is a file containing all the symbols that
1531    we wish to export.
1532    
1533    I suspect this is a Windows problem and we'll have to
1534    do the same (somehow with windows compilers) when
1535    we build the native win32 version with the system
1536    calls LoadLibrary/GetProcAddress.
1537    
1538    ----------------------------------------------------------------------
1539    Name: Matthias Blume (blume (at) tti - c (dot) org)
1540    Date: 2004/03/04 16:35:00 CST
1541    Tag: blume-20040304-intinf-fmt
1542    Description:
1543    
1544    Fixed problem with IntInf.fmt (sign would show up on the right instead
1545    of on the left for BIN, OCT, and HEX).
1546    
1547    ----------------------------------------------------------------------
1548    Name: Matthias Blume (blume (at) tti - c (dot) org)
1549    Date: 2004/03/04 11:25:00 CST
1550    Tag: blume-20040304-symlinks
1551    Description:
1552    
1553    Fixed problem with installer script (unix only) where bin/ml-yacc and
1554    friends pointed (via symlinks) to absolute locations instead of just
1555    .run-sml.  This was reported by Vesa A Norrman.
1556    
1557    ----------------------------------------------------------------------
1558    Name: Matthias Blume (blume (at) tti - c (dot) org)
1559    Date: 2004/02/13 14:50:00 CST
1560    Tag: Release_110_45
1561    Description:
1562    
1563    New working version (110.45). New bootfiles.
1564    
1565    ----------------------------------------------------------------------
1566    Name: Matthias Blume (blume (at) tti - c (dot) org)
1567    Date: 2004/01/26 15:15:15 CST
1568    Tag: blume-20040126-toplevel
1569    Description:
1570    
1571    Improved handling of exceptions at the interactive toplevel.
1572    
1573    ----------------------------------------------------------------------
1574    Name: Matthias Blume (blume (at) tti - c (dot) org)
1575    Date: 2004/01/26 11:25:00
1576    Tag: blume-20040126-app
1577    Description:
1578    
1579    Type of top-level "app" corrected.
1580    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1581    
1582    ----------------------------------------------------------------------
1583    Name: Matthias Blume (blume (at) tti - c (dot) org)
1584    Date: 2003/11/18 17:10 CST
1585    Tag: blume-20031118-basis-fiddle
1586    Description:
1587    
1588    - changed Timer interface to what might become the spec
1589    - POSIX_FLAGS -> BIT_FLAGS according to spec
1590    - some other minor discrepancies wrt. spec eliminated
1591    
1592    ----------------------------------------------------------------------
1593    Name: Matthias Blume (blume (at) tti - c (dot) org)
1594    Date: 2003/11/06 12:00:00 CST
1595    Tag: Release_110_44
1596    Description:
1597    
1598    New working version (110.44). New bootfiles.
1599    
1600    ----------------------------------------------------------------------
1601    Name: Matthias Blume (blume (at) tti - c (dot) org)
1602    Date: 2003/11/04 11:50:00 CST
1603    Tag: blume-20031104-move-libraries
1604    Description:
1605    
1606    Eliminated the "dont_move_libraries" directive in config/targets.
1607    (The mechanism was broken and could not be fixed easily.  Moreover,
1608    there does not seem to be any reason not to move all libraries into
1609    lib during installation.  I originally implemented this directive as a
1610    backward-compatibility feature when I first introduced the new CM.
1611    Now that things have been stable for a long time and going back to the
1612    old CM is not an option, there is no reason to keep it around.)
1613    
1614    ----------------------------------------------------------------------
1615    Name: Matthias Blume (blume (at) tti - c (dot) org)
1616    Date: 2003/11/03 16:00:00 CST
1617    Tag: blume-20031103-installdir
1618    Description:
1619    
1620    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1621    Richards for pointing out the problem and providing the solution.)
1622    
1623    ----------------------------------------------------------------------
1624    Name: Matthias Blume (blume (at) tti - c (dot) org)
1625    Date: 2003/10/01 17:05:00 CDT
1626    Tag: blume-20031001-lal-mlrisc
1627    Description:
1628    
1629    MLRISC bug fix from Lal.
1630    
1631    ----------------------------------------------------------------------
1632    Name: Matthias Blume (blume (at) tti - c (dot) org)
1633    Date: 2003/09/30 16:10:00 CDT
1634    Tag: blume-20030930-primio-bat
1635    Description:
1636    
1637    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1638    2. Improved .bat files (for Win32 port) to make things work under Win95.
1639       (thanks to Aaron S. Hawley for this one)
1640    
1641    ----------------------------------------------------------------------
1642    Name: Matthias Blume (blume (at) tti - c (dot) org)
1643    Date: 2003/09/26 16:05:00 CDT
1644    Tag: blume-20030926-wrappriv
1645    Description:
1646    
1647    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1648    
1649    ----------------------------------------------------------------------
1650    Name: Matthias Blume (blume (at) tti - c (dot) org)
1651    Date: 2003/09/26 15:00:00 CDT
1652    Tag: blume-20030926-110_43_3
1653    Description:
1654    
1655    - additional cleanup
1656    - version number bump, NEW BOOTFILES
1657    
1658    ----------------------------------------------------------------------
1659    Name: Matthias Blume (blume (at) tti - c (dot) org)
1660    Date: 2003/09/26 12:00:00 CDT
1661    Tag: blume-20030926-ppautoload
1662    Description:
1663    
1664    I modified the read-eval-print loop so that the autoloader gets
1665    invoked whenever the prettyprinter tries to look up a symbol that
1666    is not currently defined in the toplevel environment but which
1667    appears in CM's autoload registry.  As a result, we see far fewer of
1668    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1669    
1670    In addition to this I tried to clean up some pieces of the Basis
1671    implementation (e.g., Socket, Word8Array) in order to prevent other
1672    instances of these ?.Foo.Bar.xxx names from being printed.
1673    
1674    The mechanism that picks names for types still needs some work, though.
1675    (Right now it seems that if there is a type A.t which is defined to
1676    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1677    "?.B.u" although the perhaps more sensible solution would be to use
1678    "A.t" in this case.  In other words, the prettyprinter should follow
1679    a chain of DEFtycs not farther than there are corresponding toplevel
1680    names in the current environment.)
1681    
1682    ----------------------------------------------------------------------
1683    Name: Matthias Blume (blume (at) tti - c (dot) org)
1684    Date: 2003/09/24 16:31:00 CDT
1685    Tag: blume-20030924-installer
1686    Description:
1687    
1688    Another installer tweak:  All the ML code for the installer is now
1689    compiled during CMB.make and put into a little library called
1690    $smlnj/installer.cm.  The installation then simply invokes
1691    
1692       sml -m $smlnj/installer.cm
1693    
1694    and everything happens automagically.
1695    
1696    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1697    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1698            and BINDIR.
1699    
1700    The new scheme guarantees that the ML code responsible for the installation
1701    is in sync with the APIs of the main system.  Also, the installer is
1702    somewhat faster because the installer script is precompiled.
1703    
1704    ----------------------------------------------------------------------
1705    Name: Matthias Blume (blume (at) tti - c (dot) org)
1706    Date: 2003/09/24 15:35:00 CDT
1707    Tag: blume-20030924-synsock
1708    Description:
1709    
1710    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1711    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1712    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1713    implementations of non-blocking operations from CML's Socket
1714    structure.
1715    
1716    ----------------------------------------------------------------------
1717    Name: Matthias Blume (blume (at) tti - c (dot) org)
1718    Date: 2003/09/24 15:10:05 CDT
1719    Tag: blume-20030924-sockets
1720    Description:
1721    
1722    1. Fixed SOCKET API and implementation to match Basis spec.
1723       This required changing the internal representation of sockets to one
1724       that remembers (for each socket file descriptor) whether it is currently
1725       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1726       call is made only if the state actually needs to change).
1727    
1728    2. OS-specific details of sockets were moved into separate files, thus
1729       making it possible to unify the bulk of the socket implementations
1730       between Unix and Win32.
1731    
1732    3. CML's socket API changed accordingly.
1733       (Note that we need to remove non-blocking functions from this API
1734       since they are redundant in the case of CML!)
1735    
1736    4. CML's socket implementation now makes use of non-blocking functions
1737       provided by Basis, thus removing all OS-dependent code from this part
1738       of CML.
1739    
1740    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1741    
1742    ----------------------------------------------------------------------
1743    Name: Matthias Blume (blume (at) tti - c (dot) org)
1744    Date: 2003/09/22 12:10:00 CDT
1745    Tag: blume-20030922-110_43_2
1746    Description:
1747    
1748    Made a new interim version and bootfiles for developer's bootstrapping
1749    convenience.
1750    
1751       110.43.2 -- NEW BOOTFILES
1752    
1753    ----------------------------------------------------------------------
1754    Name: Matthias Blume (blume (at) tti - c (dot) org)
1755    Date: 2003/09/19 15:55:00 CDT
1756    Tag: blume-20030919-cmdir
1757    Description:
1758    
1759    1. new-install.sh -> install.sh
1760    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1761    3. tweaked installer so that another name instead of .cm can be chosen
1762       at install time (by setting the CM_DIR_ARC environment variable
1763       during installation); once installation is complete, the name is
1764       fixed
1765    
1766    ----------------------------------------------------------------------
1767    Name: Matthias Blume (blume (at) tti - c (dot) org)
1768    Date: 2003/09/18 16:00:00 CDT
1769    Tag: blume-20030918-110_43_1
1770    Description:
1771    
1772    Made a new interim version and bootfiles for developer's bootstrapping
1773    convenience.
1774    
1775       110.43.1 -- NEW BOOTFILES
1776    
1777    ----------------------------------------------------------------------
1778    Name: Matthias Blume (blume (at) tti - c (dot) org)
1779    Date: 2003/09/18 15:20:00 CDT
1780    Tag: blume-20030918-misc
1781    Description:
1782    
1783    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
1784       this seems to be controversial at the moment) and used those in
1785       Posix.ProcEnv.times.
1786    
1787    2. Added Time.{from,to}Nanoseconds to Time.
1788    
1789    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
1790       For example, fromLargeInt never needs to look at more than 3 "big
1791       digits" to get its 53 bits of precision.
1792    
1793    ----------------------------------------------------------------------
1794    Name: Matthias Blume (blume (at) tti - c (dot) org)
1795    Date: 2003/09/17 16:30:00 CDT
1796    Tag: blume-20030917-real32-slices
1797    Description:
1798    
1799    Added an entry to the primitive environment
1800    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
1801    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
1802    
1803    Removed some of the "magic" constants in real64.sml and replaced them
1804    with code that generates these values from their corresponding
1805    integer counterparts.
1806    
1807    Made all(?) the slice-related changes to the Basis and made everything
1808    compile again...
1809    
1810    ----------------------------------------------------------------------
1811    Name: Matthias Blume (blume (at) tti - c (dot) org)
1812    Date: 2003/09/15 17:45:00 CDT
1813    Tag: blume-20030915-rbase
1814    Description:
1815    
1816    Fixed bug in Real.fromLargeInt.
1817    
1818    ----------------------------------------------------------------------
1819    Name: Matthias Blume (blume (at) tti - c (dot) org)
1820    Date: 2003/09/13 18:11:00 CDT
1821    Tag: blume-20030913-libinstall
1822    Description:
1823    
1824    Minor bugfix in config/libinstall (set anchor with path to
1825    standalone tool after installing it, otherwise libraries that
1826    need ml-lex or ml-yacc won't compile the first time the installer
1827    runs).
1828    
1829    ----------------------------------------------------------------------
1830    Name: Matthias Blume (blume (at) tti - c (dot) org)
1831    Date: 2003/09/12 11:45:00 CDT
1832    Tag: blume-20030912-various
1833    Description:
1834    
1835     - fixed bug in Real.toLargeInt
1836     - fixed bug in Posix.ProcEnv.times
1837     - changed inputLine functions to return an option
1838     - minor installer improvements / bugfixes
1839     - changed default @SMLalloc parameter for x86/celeron to 64k
1840    
1841    ----------------------------------------------------------------------
1842    Name: Matthias Blume (blume (at) tti - c (dot) org)
1843    Date: 2003/09/09 22:00:00 CDT
1844    Tag: Release_110_43
1845    Description:
1846    
1847        New working release 110.43.  New bootfiles.
1848    
1849    ----------------------------------------------------------------------
1850    Name: Matthias Blume (blume (at) tti - c (dot) org)
1851    Date: 2003/09/09 19:20:00 CDT
1852    Tag: blume-20030909-installer
1853    Description:
1854    
1855    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
1856    Modified config/install.bat to take advantage of it.  Also modified
1857    config/install.sh (and called it config/new-install.sh) to take advantage
1858    of it on Unix systems.  (The SML code is (supposed to be) platform-
1859    independent.)
1860    
1861    The installer can now install everything under Win32
1862    as well as under *nix as long as it compiles.
1863    
1864    Other changes:
1865    
1866      - made CML compile again under Win32
1867      - made eXene compile under Win32 (by providing a fake structure UnixSock
1868        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
1869      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
1870        same as type int;  under Win32 it isn't
1871      - fixed some slice-related problems in the win32-specific parts of CML
1872      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
1873        (under Win32, the volume name is case-insensitive, and the
1874        OS.Path code compares volume names for equality)
1875    
1876    ----------------------------------------------------------------------
1877    Name: Matthias Blume (blume (at) tti - c (dot) org)
1878    Date: 2003/09/08 11:55:00 CDT
1879    Tag: blume-20030908-fullpath
1880    Description:
1881    
1882    Made Win32 version of OS.FileSys.fullPath return current directory
1883    when given an empty string.  This is what the spec says, and incidentally,
1884    CM depends on it.  (CM otherwise goes into an infinite loop in certain
1885    cases when presented with the name of a non-existing .cm file.)
1886    
1887    ----------------------------------------------------------------------
1888    Name: Matthias Blume (blume (at) tti - c (dot) org)
1889    Date: 2003/09/04 16:30:00 CDT
1890    Tag: blume-20030905-slices-etc
1891    Description:
1892    
1893      1. Changed interface to vectors and arrays in Basis to match
1894         (draft) Basis spec.
1895      2. Added signatures and implementations of slices according to
1896         Basis spec.
1897      3. Edited source code throughout the system to make it compile again
1898         under 1. and 2.  (In some cases code had to be added to have it
1899         match the new signatures.)
1900      4. MLRISC should be backward-compatible: the copies of the originals
1901         of files that needed to change under 3. were retained, the .cm files
1902         check the compiler version number and use old versions when
1903         appropriate.
1904      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
1905         dirstream -> string option (in accordance with Basis spec).
1906      6. When generating code that counts lines, ml-lex used function
1907         CharVector.foldli, taking advantage of its old interface.
1908         This has been replaced with the corresponding code from
1909         CharVectorSlice.  (html-lex must be re-lexed!)
1910      7. BitArray in smlnj-lib/Util has been extended/modified to match the
1911         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
1912      8. Removed temporary additions (fromInternal, toInternal) from the
1913         (now obsolete) IntInf in smlnj-lib/Util.
1914      9. Cleaned up structure Byte.
1915     10. Added localOffset, scan, and fromString to Date (according to spec).
1916         Cleaned/corrected implementation of Date.
1917         (Still need to check for correctness; implement better canonicalizeDate.)
1918     11. Added "scan" to signature IEEE_REAL.
1919     12. Some improvements to IntInf [in particular: efficiency-hack for
1920         mod and rem when second operand is 2 (for parity checks).]
1921     13. Changed representation of type Time.time, using a single IntInf.int
1922         value counting microseconds.  This considerably simplified the
1923         implementation of structure Time.  We now support negative time
1924         values; scan and fromString handle signs.
1925     14. Functor PrimIO now takes two additional arguments (VectorSlice and
1926         ArraySlice).
1927    
1928    ----------------------------------------------------------------------
1929    Name: Matthias Blume (blume (at) tti - c (dot) org)
1930    Date: 2003/08/28 17:00:00 CDT
1931    Tag: blume-20030828-intinf
1932    Description:
1933    
1934    This is a major update which comes with a version number bump
1935    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
1936    and an implementation of IntInf in the Basis.
1937    
1938    There are a fairly large number of related changes and updates throughout
1939    the system:
1940    
1941    Basis:
1942            - Implemented IntInf.
1943            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
1944            - Added some missing Real64 operations, most notably Real.toLargeInt.
1945            - Added FixedInt as a synonym for Int32.
1946    
1947    compiler:
1948            * Added support for a built-in intinf type.
1949              - literals
1950              - pattern matching
1951              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
1952              - overloading on literals and operations
1953    
1954            This required adding a primitive type intinf, some additional
1955            primops, and implementations for several non-trivial intinf
1956            operations in Core.  (The intinf type is completely abstract
1957            to the compiler; all operations get delegated back to the Core.)
1958    
1959            * Intinf equality is handled by polyequal.  However, the compiler
1960            does not print its usual warning in this case (since polyequal
1961            is the right thing to do there).
1962    
1963            * Improved the organization of structure InlineT.
1964    
1965            * A word about conversion primops:
1966            If conversions involving intinf do not cancel out during
1967            CPS contract, then the compiler must insert calls to Core functions.
1968            Since all core access must be resolved already during the FLINT
1969            translate phase, it would be too late a the time of CPS contract
1970            to add new Core calls.  For this reason, conversion primops
1971            for intinf carry two arguments:  1. the numeric argument that
1972            they are supposed to convert, and 2. the Core function that
1973            can help with this conversion if necessary.  If CPS contract
1974            eliminates a primop, then the associated Core function becomes
1975            dead and goes away.  Intinf conversion primops that do not get
1976            eliminated by CPS contract get rewritten into calls of their
1977            core functions by a separate, new phase.
1978    
1979    interactive system:
1980            - Control.Print.intinfDepth controls max length of intinf constants
1981            being printed. (Analogous to Control.Print.stringDepth.)
1982            - Cleanup in printutil and pputil: got rid of unused stuff and
1983            duplicates; replaced some of the code with code that makes better
1984            use of library functionality.
1985    
1986    CM:
1987            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
1988            are no longer silent.
1989    
1990    CKIT:
1991            Fixed mismatched uses of Int32 and LargeInt.  I always decided
1992            in favor of LargeInt -- which is now the same as IntInf.
1993            CKIT-knowledgable people should check whether this is what's
1994            intended and otherwise change things back to using Int32 or
1995            FixedInt.
1996    
1997    Throughout the code:
1998            Started using IntInf.int literals and built-in operations
1999            (e.g., comparison with 0) where this seems appropriate.
2000    
2001    
2002    ----------------------------------------------------------------------
2003    Name: Dave MacQueen (dbm@cs.uchicago.edu)
2004    Date: 2003/08/13 11:36:00 CDT
2005    Tag: dbm-20030813-mcz-merge1
2006    Description:
2007    
2008    Merging changes from the mcz-branch development branch into trunk.
2009    These changes involve replacement of the emulated old prettyprinter
2010    interface with direct use of the SML/NJ Lib PP library, and fixing
2011    of a couple of bugs (895, 1186) relating to error messages.  A new
2012    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
2013    has been added.
2014    
2015    ----------------------------------------------------------------------
2016    Name: Matthias Blume (blume (at) tti - c (dot) org)
2017    Date: 2003/08/11 15:45:00 CDT
2018    Tag: blume-20030811-windows
2019    Description:
2020    
2021    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
2022    
2023        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
2024    
2025    This patch restores SML/NJ's ability to run under win32.  There are a
2026    number of changes, including fixes for several bugs that had gone
2027    unnoticed until now:
2028    
2029      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
2030        trying to run the win32 version from within cygwin.)
2031    
2032      - There are a number of simple .bat scripts that substitute for their
2033        corresponding Unix shell-scripts.  (See below.)
2034    
2035      - The internals of ml-build have been modified slightly.  The main
2036        difference is that instead of calling ".link-sml" (or link-sml.bat)
2037        using OS.Process.system, the ML process delegates this task back
2038        to the script.  Otherwise problems arise in mixed environments such
2039        as Cygwin where scripts look and work like Unix scripts, but
2040        where OS.Process.system cannot run them.
2041    
2042      - In CM, the srcpath pickler used native pathname syntax -- which
2043        is incorrect in the case of cross-compilation.  The new pickle format
2044        is independent of platform-specific naming conventions.
2045    
2046      - Path configuration files (such as lib/pathconfig) can now choose
2047        between native and standard syntax.  Placing a line of the form
2048    
2049            standard!
2050    
2051        into the file causes all subsequent paths to be interpreted using
2052        CM standard pathname syntax (= Unix conventions); a line
2053    
2054            native!
2055    
2056        switches back to native style.  This was needed so that
2057        path config files can be written portably, see src/system/pathconfig.
2058    
2059      - Runtime system:
2060    
2061          - win32-filesys.c:  get_file_time and set_file_time now
2062            access modification time, not creation time.
2063    
2064          - I/O code made aware of new array representation.
2065    
2066          - Bug fixes in X86.prim.masm.
2067    
2068       - src/system/makeml made aware of win32. (For use under cygwin
2069            and other Unix-environments for windows.)
2070    
2071       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
2072         which caused BinIO.inputAll to fail consistently.
2073    
2074    .bat scripts:
2075    
2076       Windows .bat scripts assume that SMLNJ_HOME is defined.
2077    
2078       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
2079         applications (sml, ml-yacc, ml-lex).
2080       - ml-build.bat: analogous to ml-build.
2081       - config\install.bat:  Analogous to config/install.sh.  This requires
2082         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2083         (nmake etc. must be on the path, and vcvars32 must have been run.)
2084         Moreover, sources for ml-lex and ml-yacc need to exist under src,
2085         and the bootfile hierarchy must have been unpacked under
2086         sml.boot.x86-win32.
2087         The script is very primitive and does a poor job at error checking.
2088         It only installs the base system, ml-lex, and ml-yacc.  No other
2089         libraries are being installed (i.e., you get only those that
2090         are part of the compiler.)
2091       - link-sml.bat:  analogous to .link-sml, but not currently used
2092    
2093    Unrelated bug fixes:
2094    
2095      - ml-nlffigen now exports structures ST_* corresponding to incomplete
2096        types.
2097      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
2098        otherwise.)
2099    
2100    ----------------------------------------------------------------------
2101    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2102    Date: 2003/06/17
2103    Tag: macqueen-20030617-bug895
2104    Description:
2105    
2106    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2107    Tag will be used for new development branch (mcz-branch) for
2108    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2109    
2110    ----------------------------------------------------------------------
2111    Name: Matthias Blume (blume (at) tti - c (dot) org)
2112    Date: 2003/05/27 16:55:00 CDT
2113    Tag: blume-20030527-polyeq
2114    Description:
2115    
2116    Tried to eliminated most cases of polymorphic equality.
2117    
2118    ----------------------------------------------------------------------
2119    Name: Matthias Blume (blume (at) tti - c (dot) org)
2120    Date: 2003/05/21 17:45:00 CDT
2121    Tag: blume-20030517-complete
2122    Description:
2123    
2124    Two changes:
2125    
2126      1. Added a flag for controlling whether non-exhaustive bindings will
2127         be treated as errors (default is false).
2128      2. Cleaned up the *entire* source tree so that CMB.make goes through
2129         without a single non-exhaustive match- or bind warning.
2130    
2131    ----------------------------------------------------------------------
2132    Name: Matthias Blume (blume (at) tti - c (dot) org)
2133    Date: 2003/05/17 10:20:00 CDT
2134    Tag: blume-20030517-absyn
2135    Description:
2136    
2137    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2138    
2139       This mainly affects the quality of error messages.  However, some
2140       of the code is now more straightforward than before. (Treatment of
2141       the above four constructs in translate.sml is much simpler than
2142       the "macro-expansion" that was going on before.  Plus, the mach-
2143       compiler no longer gets invoked just to be able to compile an
2144       if-expression.)
2145    
2146    2. The ErrorMsg.Error exception is now caught and absorbed by the
2147       interactive loop.
2148    
2149    ----------------------------------------------------------------------
2150    Name: Allen Leung
2151    Date: 2003/05/16 13:05:00 CDT
2152    Tag: leunga-20030516-cygwin-runtime
2153    Description:
2154    
2155        Ported the runtime system to cygwin, which uses the unix
2156    x86-unix bin files.  Missing/buggy features:
2157    
2158       o getnetbyname, getnetbyaddr: these functions seem to be missing in
2159         the Cygwin library.
2160       o Ctrl-C handling may be flaky.
2161       o Windows system calls and Windows I/O are not supported.
2162    
2163       A new set of binfiles is located at:
2164    
2165         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2166    
2167       This is only needed for bootstrapping the cygwin version of smlnj.
2168       Other x86 versions can use the existing binfiles.
2169    
2170    ----------------------------------------------------------------------
2171    Name: Matthias Blume
2172    Date: 2003/04/08 15:42:00 CDT
2173    Tag: blume-20030408-listpair
2174    Description:
2175    
2176    1. Added a target 'mlrisc' to installer.
2177    
2178    2. Added missing elements to structure ListPair.
2179    
2180    ----------------------------------------------------------------------
2181    Name: Allen Leung
2182    Date: 2003/01/07 10:40:00 EST
2183    Tag: leunga-20030107-int-rem
2184    Description:
2185    
2186    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2187    The arguments to the SUBL instruction were swapped.
2188    
2189    ----------------------------------------------------------------------
2190    Name: Matthias Blume
2191    Date: 2002/12/12 16:25:00 EST
2192    Tag: blume-20021212-risc-ra
2193    Description:
2194    
2195    Fixed a serious bug in the rewrite code for FP spilling/reloading that
2196    sent the RA into an infinite loop when floating point registers get
2197    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2198    110.37 and 110.38.)
2199    There was another set of potential problems related to the handling of
2200    MLRISC annotations (but those did not yet cause real problems, apparently).
2201    
2202    ----------------------------------------------------------------------
2203    Name: Matthias Blume
2204    Date: 2002/12/06 22:40:00 EST
2205    Tag: blume-20021206-cm-fileid
2206    Description:
2207    
2208    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2209    This fixes the problem of CM getting confused by files that suddenly
2210    change their identity (e.g., by getting unlinked and recreated by some
2211    text editor such as vi).  There might be a better/cheaper/cleaner way
2212    of doing this, but for now this will have to do.
2213    
2214    ----------------------------------------------------------------------
2215    Name: Matthias Blume
2216    Date: 2002/10/28 09:50:00 EST
2217    Tag: blume-20021028-typecheck
2218    Description:
2219    
2220    Exported structure Typecheck from $smlnj/viscomp/core.cm.
2221    
2222    ----------------------------------------------------------------------
2223    Name: Matthias Blume
2224    Date: 2002/10/17 09:10:00 EDT
2225    Tag: Release_110_42
2226    Description:
2227    
2228    In good old tradition, there has been a slight hiccup so that we have
2229    to patch 110.42 after the fact.   The old release tag has been replaced
2230    (see below).
2231    
2232    The change solves a problem with two competing approaches the
2233    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2234    each other's way.
2235    
2236    This change only affects the runtime system code and the installer script.
2237    (No new bootfiles.)
2238    
2239    ----------------------------------------------------------------------
2240    Name: Matthias Blume
2241    Date: 2002/10/16 12:00:00 EDT
2242    Tag: Release_110_42_removed
2243    Description:
2244    
2245    New working release.  New bootfiles.
2246    
2247    ----------------------------------------------------------------------
2248    Name: Matthias Blume
2249    Date: 2002/10/10 13:10:00 EDT
2250    Tag: blume-20021010-ppc-divs
2251    Description:
2252    
2253    The mltree operator DIVS must be implemented with an overflow check on
2254    the PPC because the hardware indicates divide-by-zero using "overflow" as
2255    well.
2256    
2257    ----------------------------------------------------------------------
2258    Name: Matthias Blume
2259    Date: 2002/07/23 11:45:00
2260    Tag: blume-20020723-smlnj-home
2261    Description:
2262    
2263    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2264    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2265    CM_PATHCONFIG is also set), the path configuration file is assumed
2266    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2267    the entire tree to some other place and everything will "just work".
2268    
2269    (Companion commands such as ml-build and ml-makedepend also sense this
2270    variable.)
2271    
2272    ----------------------------------------------------------------------
2273    Name: Matthias Blume
2274    Date: 2002/07/12 21:19:00 EDT
2275    Tag: blume-20020712-liveness
2276    Description:
2277    
2278    Exported two useful "step" functions from liveness module (MLRISC).
2279    
2280    ----------------------------------------------------------------------
2281    Name: Matthias Blume
2282    Date: 2002/07/05 16:00 EDT
2283    Tag: Release_110_41
2284    Description:
2285    
2286    New working release.  New bootfiles.
2287    
2288    ----------------------------------------------------------------------
2289    Name: Matthias Blume
2290    Date: 2002/07/05 10:25:00 EDT
2291    Tag: blume-20020705-btimp
2292    Description:
2293    
2294    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2295    other clients can set up backtracing support.
2296    
2297    ----------------------------------------------------------------------
2298    Name: Matthias Blume
2299    Date: 2002/06/25 17:23:00 EDT
2300    Tag: blume-20020625-fpmax
2301    Description:
2302    
2303    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2304    case.  (The sense of the isNaN test was reversed -- which made min and
2305    max always return their first argument.)
2306    
2307    ----------------------------------------------------------------------
2308    Name: Matthias Blume
2309    Date: 2002/06/11
2310    Tag: blume-20020611-unixpath
2311    Description:
2312    
2313    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2314    
2315    ----------------------------------------------------------------------
2316    Name: Matthias Blume
2317    Date: 2002/06/10 16:35:00 EDT
2318    Tag: blume-20020610-ieeereal
2319    Description:
2320    
2321    I back-ported my implementation of IEEEReal.fromString from the
2322    idlbasis-devel branch so that we can test it.
2323    
2324    Another small change is that ppDec tries to give more information
2325    than just "<sig>" in the case of functors.  However, this code is
2326    broken in some mysterious way if the functor's body's signature
2327    has not been declared by ascription but gets inferred from the
2328    implementation.  This needs fixing...
2329    
2330    ----------------------------------------------------------------------
2331    Name: Matthias Blume
2332    Date: 2002/05/31
2333    Tag: blume-20020531-btrace-mode
2334    Description:
2335    
2336    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2337    the wayside when I switched over to using Controls everywhere.)
2338    
2339    ----------------------------------------------------------------------
2340    Name: Lal George
2341    Date: 2002/05/23 12:21:40 EDT
2342    Tag: george-20020523-visual-labels
2343    Description:
2344    
2345    Labels are now displayed in the graphical output to make
2346    the fall-through and target blocks obvious.
2347    
2348    ----------------------------------------------------------------------
2349    Name: Matthias Blume
2350    Date: 2002/05/22 11:03:00 EDT
2351    Tag: blume-20020522-shrink
2352    Description:
2353    
2354    John tweaked yesterday's fix for 1131 to handle an out-of-memory
2355    situation that comes up when allocating huge arrays.
2356    
2357    ----------------------------------------------------------------------
2358    Name: Matthias Blume
2359    Date: 2002/05/21 16:00:00 EDT
2360    Tag: Release_110_40
2361    Description:
2362    
2363    New working release (110.40).  New bootfiles.
2364    
2365    [Also: John Reppy fixed GC bug 1131.]
2366    
2367    ----------------------------------------------------------------------
2368    Name: Matthias Blume
2369    Date: 2002/05/21 12:35:00 EDT
2370    Tag: blume-20020521-cmdoc
2371    Description:
2372    
2373    CM documentation update.
2374    
2375    ----------------------------------------------------------------------
2376    Name: Matthias Blume
2377    Date: 2002/05/21 10:55:00 EDT
2378    Tag: blume-20020521-misc
2379    Description:
2380    
2381    - John tweaked runtime to be silent on heap export (except when
2382      GC messages are on).
2383    - I added a few more things (cross-compiling versions of CMB) to
2384      config/preloads (as suggestions).
2385    
2386    ----------------------------------------------------------------------
2387    Name: Matthias Blume
2388    Date: 2002/05/20 22:25:00 EDT
2389    Tag: blume-20020520-controls
2390    Description:
2391    
2392    - Added ControlUtil structure to control-lib.cm.
2393    - Use it throughout.
2394    - Used Controls facility to define MLRISC controls (as opposed to
2395      registering MLRISC control ref cells with Controls after the fact)
2396    - Fixed messed-up controls priorities.
2397    
2398    * Removed again all the stuff from config/preloads that one wouldn't
2399      be able to preload at the time the initial heap image is built.
2400      (Many libraries, e.g., CML, do not exist yet at this time.  The
2401       only libraries that can be preloaded via config/preloads are those
2402       that come bundled with the bootfiles.)
2403    
2404    ----------------------------------------------------------------------
2405    Name: Matthias Blume
2406    Date: 2002/05/20 10:59:00 EDT
2407    Tag: blume-20020520-preloads
2408    Description:
2409    
2410    Added a lot of commented-out suggestions for things to be included
2411    in config/preloads.
2412    
2413    ----------------------------------------------------------------------
2414    Name: Allen Leung
2415    Date: 2002/05/18 14:20:00 EDT
2416    Tag: leunga-20020518-mdl
2417    Description:
2418    
2419    o Made the mdl tool stuff compile and run again.
2420    o I've disabled all the stuff that depends on RTL specifications; they
2421      are all badly broken anyway.
2422    
2423    ----------------------------------------------------------------------
2424    Name: Matthias Blume
2425    Date: 2002/05/17 16:49:00 EDT
2426    Tag: blume-20020517-controls
2427    Description:
2428    
2429    0. John Reppy made several modifications to the SML/NJ library.
2430       In particular, there is a shiny new controls-lib.cm.
2431    
2432    1. Pushed new controls interface through compiler so that everything
2433       compiles again.
2434    
2435    2. Added FormatComb and FORMAT_COMB to the CML version of the
2436       SML/NJ library (so that CML compiles again).
2437    
2438    3. Modified init scripts because XXX_DEFAULT environment variables
2439       are no longer with us.  (Boot-time initialization is now done
2440       using the same environment variables that are also used for
2441       startup-time initialization of controls.)
2442    
2443    ----------------------------------------------------------------------
2444    Name: Lal George
2445    Date: 2002/05/15 09:20:10 EDT
2446    Tag: george-20020515-pseudo-op-decls
2447    Description:
2448    
2449            All pseudo-ops emitted before the first segment declaration
2450    such as TEXT, DATA, and BSS directives are assumed to be global
2451    declarations and are emitted first in the assembly file. This is
2452    useful in a number of situations where one has pseudo-ops that are not
2453    specific to any segment, and also works around the constraint that one
2454    cannot have client pseudo-ops in the TEXT segment.
2455    
2456            Because no segment is associated with these declarations it is
2457    an error to allocate any space or objects before the first segment
2458    directive and an exception will be raised. However, we cannot make
2459    this check for client pseudo-ops.
2460    
2461            These top level declarations are a field in the CFG graph_info.
2462    In theory you can continue to add to this field after the CFG has been
2463    built -- provided you know what you are doing;-)
2464    
2465    ----------------------------------------------------------------------
2466    Name: Matthias Blume
2467    Date: 2002/05/13 16:40:00 EDT
2468    Tag: blume-20020513-pp-etc
2469    Description:
2470    
2471    A few minor bugfixes:
2472    
2473      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2474        (Bogus printouts for redefined bindings still occur.  Compiler
2475        bug should no longer occur now.  We need to redo the prettyprinter
2476        from scratch.)
2477    
2478      - CM pathname printer now also adds escape sequences for ( and )
2479    
2480      - commend and docu fixes for ml-nlffi
2481    
2482    ----------------------------------------------------------------------
2483    Name: Matthias Blume
2484    Date: 2002/05/10 16:40:00 EDT
2485    Tag: blume-20020510-erg-textio
2486    Description:
2487    
2488    Applied the following bugfix provided by Emden Gansner:
2489    
2490        Output is corrupted when outputSubstr is used rather than output.
2491    
2492        The problem occurs when a substring
2493    
2494            ss = (s, dataStart, dataLen)
2495    
2496        where dataStart > 0, fills a stream buffer with avail bytes left.
2497        avail bytes of s, starting at index dataStart, are copied into the
2498        buffer, the buffer is flushed, and then the remaining dataLen-avail
2499        bytes of ss are copied into the beginning of the buffer. Instead of
2500        starting this copy at index dataStart+avail in s, the current code
2501        starts the copy at index avail.
2502    
2503        Fix:
2504        In text-io-fn.sml, change line 695 from
2505             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2506        to
2507             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2508    
2509    ----------------------------------------------------------------------
2510    Name: Matthias Blume
2511    Date: 2002/04/12 13:55:00 EDT
2512    Tag: blume-20020412-assyntax
2513    Description:
2514    
2515    1. Grabbed newer assyntax.h from the XFree86 project.
2516    2. Fiddled with how to compile X86.prim.asm without warnings.
2517    3. (Very) Minor cleanup in CM.
2518    
2519    ----------------------------------------------------------------------
2520    Name: Matthias Blume
2521    Date: 2002/04/01 (no joke!) 17:07:00 EST
2522    Tag: blume-20020401-x86div
2523    Description:
2524    
2525    Added full support for div/mod/rem/quot on the x86, using the machine
2526    instruction's two results (without clumsily recomputing the remainder)
2527    directly where appropriate.
2528    
2529    Some more extensive power-of-two support was added to the x86 instruction
2530    selector (avoiding expensive divs, mods, and muls where they can be
2531    replaced with cheaper shifts and masks).  However, this sort of thing
2532    ought to be done earlier, e.g., within the CPS optimizer so that
2533    all architectures benefit from it.
2534    
2535    The compiler compiles to a fixed point, but changes might be somewhat
2536    fragile nevertheless.  Please, report any strange things that you might
2537    see wrt. div/mod/quot/rem...
2538    
2539    ----------------------------------------------------------------------
2540    Name: Matthias Blume
2541    Date: 2002/03/29 17:22:00
2542    Tag: blume-20020329-div
2543    Description:
2544    
2545    Fixed my broken div/mod logic.  Unfortunately, this means that the
2546    inline code for div/mod now has one more comparison than before.
2547    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2548    The problem was with quotient = 0, because that alone does not tell
2549    us which way the rounding went.  One then has to look at whether
2550    remainder and divisor have the same sign...  :(
2551    
2552    Anyway, I replaced the bootfiles with fresh ones...
2553    
2554    ----------------------------------------------------------------------
2555    Name: Matthias Blume
2556    Date: 2002/03/29 14:10:00 EST
2557    Tag: blume-20020329-inlprims
2558    Description:
2559    
2560    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2561    
2562    Primops have changed. This means that the bin/boot-file formats have
2563    changed as well.
2564    
2565    To make sure that there is no confusion, I made a new version.
2566    
2567    
2568    CHANGES:
2569    
2570    * removed REMT from mltree (remainder should never overflow).
2571    
2572    * added primops to deal with divisions of all flavors to the frontend
2573    
2574    * handled these primops all the way through so they map to their respective
2575      MLRISC support
2576    
2577    * used these primops in the implementation of Int, Int32, Word, Word32
2578    
2579    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2580    
2581    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2582    
2583    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2584      floating point
2585    
2586    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2587      and Real (but Real.abs maps to a separate floating-point-only primop)
2588    
2589    
2590    TODO items:
2591    
2592    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2593      instructions because the machine instruction encoder cannot handle
2594      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2595      REM in the first place, and REM is supported by the x86 machine coder.)
2596    
2597    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2598      selection phase.  (The two can be streamlined because the hardware
2599      delivers both quotient and remainder at the same time anyway.)
2600    
2601    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2602      (Currently the behavior is inconsistent both across architectures and
2603      wrt. the draft Basis spec.)
2604    
2605    * Word8 should eventually be handled natively, too.
2606    
2607    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2608      as if there currently is no execution path that could trigger it in
2609      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2610      hold for things like multiplication and division.)
2611    
2612    ----------------------------------------------------------------------
2613    Name: Matthias Blume
2614    Date: 2002/03/27 16:27:00 EST
2615    Tag: blume-20020327-mlrisc-divisions
2616    Description:
2617    
2618    Added support for all four division operations (ML's div, mod, quot,
2619    and rem) to MLRISC.  In the course of doing so, I also rationalized
2620    the naming (no more annoying switch-around of DIV and QUOT), by
2621    parameterizing the operation by div_rounding_mode (which can be either
2622    DIV_TO_ZERO or DIV_TO_NEGINF).
2623    
2624    The generic MLTreeGen functor takes care of compiling all four
2625    operations down to only round-to-zero div.
2626    
2627    Missing pieces:
2628    
2629      * Doing something smarter than relying on MLTreeGen on architectures
2630        like, e.g., the x86 where hardware division delivers both quotient and
2631        remainder at the same time.  With this, the implementation of the
2632        round-to-neginf operations could be further streamlined.
2633    
2634      * Remove inlining support for div/mod/rem from the frontend and replace it
2635        with primops that get carried through to the backend.  Do this for all
2636        int and word types.
2637    
2638    ----------------------------------------------------------------------
2639    Name: Matthias Blume
2640    Date: 2002/03/25 17:25:00 EST
2641    Tag: blume-20020325-divmod
2642    Description:
2643    
2644    I improved (hopefully without breaking them) the implementation of Int.div,
2645    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2646    advantage of the following observations:
2647    
2648      Let  q = x quot y      r = x rem y
2649           d = x div  y      m = x mod y
2650    
2651    where "quot" is the round-to-zero version of integer division that
2652    hardware usually provides.  Then we have:
2653    
2654         r = x - q * y        where neither the * nor the - will overflow
2655         d = if q >= 0 orelse x = q * y then q else q - 1
2656                              where neither the * nor the - will overflow
2657         m = if q >= 0 orelse r = 0 then r else r + y
2658                              where the + will not overflow
2659    
2660    This results in substantial simplification of the generated code.
2661    The following table shows the number of CFG nodes and edges generated
2662    for
2663            fun f (x, y) = x OPER y
2664            (* with OPER \in div, mod, quot, rem *)
2665    
2666    
2667        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2668        --------------------------------------------------------
2669         div |         24 |         39 |         12 |         16
2670         mod |         41 |         71 |         12 |         16
2671        quot |          8 |         10 |          8 |         10
2672         rem |         10 |         14 |          8 |         10
2673    
2674    
2675    ----------------------------------------------------------------------
2676    Name: Matthias Blume
2677    Date: 2002/03/25 22:06:00 EST
2678    Tag: blume-20020325-cprotobug
2679    Description:
2680    
2681    Fixed a bug in cproto (c prototype decoder).
2682    
2683    ----------------------------------------------------------------------
2684    Name: Matthias Blume
2685    Date: 2002/03/25 16:00:00 EST
2686    Tag: blume-20020325-raw-primops
2687    Description:
2688    
2689    I did some cleanup to Allen's new primop code and
2690    replaced yesterday's bootfiles with new ones.
2691    (But they are stored in the same place.)
2692    
2693    ----------------------------------------------------------------------
2694    Name: Matthias Blume
2695    Date: 2002/03/24 22:40:00 EST
2696    Tag: blume-20020324-bootfiles
2697    Description:
2698    
2699    Made the bootfiles that Allen asked for.
2700    
2701    ----------------------------------------------------------------------
2702    Name: Allen Leung
2703    Date: 2002/03/23 15:50:00 EST
2704    Tag: leunga-20020323-flint-cps-rcc-primops
2705    Description:
2706    
2707      1. Changes to FLINT primops:
2708    
2709        (* make a call to a C-function;
2710         * The primop carries C function prototype information and specifies
2711         * which of its (ML-) arguments are floating point. C prototype
2712         * information is for use by the backend, ML information is for
2713         * use by the CPS converter. *)
2714      | RAW_CCALL of { c_proto: CTypes.c_proto,
2715                       ml_args: ccall_type list,
2716                       ml_res_opt: ccall_type option,
2717                       reentrant : bool
2718                     } option
2719       (* Allocate uninitialized storage on the heap.
2720        * The record is meant to hold short-lived C objects, i.e., they
2721        * are not ML pointers.  With the tag, the representation is
2722        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2723        *)
2724      | RAW_RECORD of {tag:bool,sz:int}
2725      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2726    
2727      2.  These CPS primops are now overloaded:
2728    
2729           rawload of {kind:numkind}
2730           rawstore of {kind:numkind}
2731    
2732          The one argument form is:
2733    
2734             rawload {kind} address
2735    
2736          The two argument form is:
2737    
2738             rawload {kind} [ml object, byte-offset]
2739    
2740      3. RAW_CCALL/RCC now takes two extra arguments:
2741    
2742         a. The first is whether the C call is reentrant, i.e., whether
2743            ML state should be saved and restored.
2744         b. The second argument is a string argument specifying the name of
2745            library and the C function.
2746    
2747         These things are currently not handled in the code generator, yet.
2748    
2749      4. In CProto,
2750    
2751         An encoding type of "bool" means "ml object" and is mapped into
2752         C prototype of PTR.  Note that "bool" is different than "string",
2753         even though "string" is also mapped into PTR, because "bool"
2754         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2755    
2756      5. Pickler/unpicker
2757    
2758         Changed to handle RAW_RECORD and newest RAW_CCALL
2759    
2760      6. MLRiscGen,
2761    
2762         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2763         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2764            in the file CodeGen/cpscompile/cps-c-calls.sml
2765    
2766      7. Added the conditional move operator
2767    
2768             condmove of branch
2769    
2770         to cps.  Generation of this is still buggy so it is currently
2771         disabled.
2772    
2773    ----------------------------------------------------------------------
2774    Name: Lal George
2775    Date: 2002/03/22 14:18:25 EST
2776    Tag: george-20020322-cps-branch-prob
2777    Description:
2778    
2779    Implemented the Ball-Larus branch prediction-heuristics, and
2780    incorporated graphical viewers for control flow graphs.
2781    
2782    Ball-Larus Heuristics:
2783    ---------------------
2784    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
2785    
2786    By design it uses the Dempster-Shafer theory for combining
2787    probabilities.  For example, in the function:
2788    
2789        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
2790    
2791    the ball-larus heuristics predicts that the n=0 is unlikely
2792    (OH-heuristic), and the 'then' branch is unlikely because of the
2793    RH-heuristic -- giving the 'then' branch an even lower combined
2794    probability using the Dempster-Shafer theory.
2795    
2796    Finally, John Reppy's loop analysis in MLRISC, further lowers the
2797    probability of the 'then' branch because of the loop in the else
2798    branch.
2799    
2800    
2801    Graphical Viewing:
2802    ------------------
2803    I merely plugged in Allen's graphical viewers into the compiler. The
2804    additional code is not much. At the top level, saying:
2805    
2806            Control.MLRISC.getFlag "cfg-graphical-view" := true;
2807    
2808    will display the graphical view of the control flow graph just before
2809    back-patching.  daVinci must be in your path for this to work. If
2810    daVinci is not available, then the default viewer can be changed
2811    using:
2812    
2813            Control.MLRISC.getString "viewer"
2814    
2815    which can be set to "dot" or "vcg" for the corresponding viewers. Of
2816    course, these viewers must be in your path.
2817    
2818    The above will display the compilation unit at the level of clusters,
2819    many of which are small, boring, and un-interesting. Also setting:
2820    
2821            Control.MLRISC.getInt "cfg-graphical-view_size"
2822    
2823    will display clusters that are larger than the value set by the above.
2824    
2825    
2826    ----------------------------------------------------------------------
2827    Name: Matthias Blume
2828    Date: 2002/03/21 22:20:00 EST
2829    Tag: blume-20020321-kmp-bugfix
2830    Description:
2831    
2832    Changed the interface to the KMP routine in PreString and fixed
2833    a minor bug in one place where it was used.
2834    
2835    ----------------------------------------------------------------------
2836    Name: Allen Leung
2837    Date: 2002/03/21 20:30:00 EST
2838    Tag: leunga-20020321-cfg
2839    Description:
2840    
2841      Fixed a potential problem in cfg edge splitting.
2842    
2843    ----------------------------------------------------------------------
2844    Name: Allen Leung
2845    Date: 2002/03/21 17:15:00 EST
2846    Tag: leunga-20020321-x86-fp-cfg
2847    Description:
2848    
2849      1. Recoded the buggy parts of x86-fp.
2850    
2851         a. All the block reordering code has been removed.
2852            We now depend on the block placement phases to do this work.
2853    
2854         b. Critical edge splitting code has been simplified and moved into the
2855            CFG modules, as where they belong.
2856    
2857         Both of these were quite buggy and complex.  The code is now much, much
2858         simpler.
2859    
2860      2. X86 backend.
2861    
2862         a. Added instructions for 64-bit support.  Instruction selection for
2863            64-bit has not been committed, however, since that
2864            requires changes to MLTREE which haven't been approved by
2865            Lal and John.
2866    
2867         b. Added support for FUCOMI and FUCOMIP when generating code for
2868            PentiumPro and above.  We only generate these instructions in
2869            the fast-fp mode.
2870    
2871         c. Added cases for JP and JNP in X86FreqProps.
2872    
2873      3. CFG
2874    
2875         CFG now has a bunch of methods for edge splitting and merging.
2876    
2877      4. Machine description.
2878    
2879         John's simplification of MLTREE_BASIS.fcond broke a few machine
2880         description things:
2881    
2882         rtl-build.{sig,sml} and hppa.mdl fixed.
2883    
2884         NOTE: the machine description stuff in the repository is still broken.
2885               Again, I can't put my fixes in because that involves
2886               changes to MLTREE.
2887    
2888    ----------------------------------------------------------------------
2889    Name: Matthias Blume
2890    Date: 2002/03/20 15:55:00 EST
2891    Tag: blume-20020320-kmp
2892    Description:
2893    
2894    Implemented Knuth-Morris-Pratt string matching in PreString and used
2895    it for String.isSubstring, Substring.isSubstring, and
2896    Substring.position.
2897    
2898    (Might need some stress-testing.  Simple examples worked fine.)
2899    
2900    ----------------------------------------------------------------------
2901    Name: Matthias Blume
2902    Date: 2002/03/19 16:37:00 EST
2903    Tag: blume-20020319-witnesses
2904    Description:
2905    
2906    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
2907    
2908    This implements a generic mechanism for changing constness qualifiers
2909    anywhere within big C types without resorting to outright "casts".
2910    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
2911    modify the constness at the outermost level.)
2912    The implementation of "convert" is based on the idea of "witness"
2913    values -- values that are not used by the operation but whose types
2914    "testify" to their applicability.  On the implementation side, "convert"
2915    is simply a projection (returning its second curried argument).  With
2916    cross-module inlining, it should not result in any machine code being
2917    generated.
2918    
2919    ----------------------------------------------------------------------
2920    Name: Matthias Blume
2921    Date: 2002/03/15 16:40:00 EST
2922    Tag: blume-20020315-basis
2923    Description:
2924    
2925    Provided (preliminary?) implementations for
2926    
2927      {String,Substring}.{concatWith,isSuffix,isSubstring}
2928    
2929    and
2930    
2931      Substring.full
2932    
2933    Those are in the Basis spec but they were missing in SML/NJ.
2934    
2935    ----------------------------------------------------------------------
2936    Name: Matthias Blume
2937    Date: 2002/03/14 21:30:00 EST
2938    Tag: blume-20020314-controls
2939    Description:
2940    
2941    Controls:
2942    ---------
2943    
2944    1. Factored out the recently-added Controls : CONTROLS stuff and put
2945       it into its own library $/controls-lib.cm.  The source tree for
2946       this is under src/smlnj-lib/Controls.
2947    
2948    2. Changed the names of types and functions in this interface, so they
2949       make a bit more "sense":
2950    
2951          module -> registry
2952          'a registry -> 'a group
2953    
2954    3. The interface now deals in ref cells only.  The getter/setter interface
2955       is (mostly) gone.
2956    
2957    4. Added a function that lets one register an already-existing ref cell.
2958    
2959    5. Made the corresponding modifications to the rest of the code so that
2960       everything compiles again.
2961    
2962    6. Changed the implementation of Controls.MLRISC back to something closer
2963       to the original.  In particular, this module (and therefore MLRISC)
2964       does not depend on Controls.  There now is some link-time code in
2965       int-sys.sml that registers the MLRISC controls with the Controls
2966       module.
2967    
2968    CM:
2969    ---
2970    
2971      * One can now specify the lambda-split aggressiveness in init.cmi.
2972    
2973    ----------------------------------------------------------------------
2974    Name: Allen Leung
2975    Date: 2002/03/13 17:30:00 EST
2976    Tag: leunga-20020313-x86-fp-unary
2977    Description:
2978    
2979    Bug fix for:
2980    
2981    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
2982    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
2983    > - fun f(x,(y,z)) = Real.~ y;
2984    > [autoloading]
2985    > [autoloading done]
2986    >       fchsl   (%eax), 184(%esp)
2987    > Error: MLRisc bug: X86MCEmitter.emitInstr
2988    >
2989    > uncaught exception Error
2990    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
2991    
2992    The problem was that the code generator did not generate any fp registers
2993    in this case, and the ra didn't know that it needed to run the X86FP phase to
2994    translate the pseudo fp instruction.   This only happened with unary fp
2995    operators in certain situations.
2996    
2997    ----------------------------------------------------------------------
2998    Name: Matthias Blume
2999    Date: 2002/03/13 14:00:00 EST
3000    Tag: blume-20020313-overload-etc
3001    Description:
3002    
3003    1. Added _overload as a synonym for overload for backward compatibility.
3004       (Control.overloadKW must be true for either version to be accepted.)
3005    
3006    2. Fixed bug in install script that caused more things to be installed
3007       than what was requested in config/targets.
3008    
3009    3. Made CM aware of the (_)overload construct so that autoloading
3010       works.
3011    
3012    ----------------------------------------------------------------------
3013    Name: Matthias Blume
3014    Date: 2002/03/12 22:03:00 EST
3015    Tag: blume-20020312-url
3016    Description:
3017    
3018    Forgot to update BOOT and srcarchiveurl.
3019    
3020    ----------------------------------------------------------------------
3021    Name: Matthias Blume
3022    Date: 2002/03/12 17:30:00 EST
3023    Tag: blume-20020312-version110392
3024    Description:
3025    
3026    Yet another version number bump (because of small changes to the
3027    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
3028    
3029    Changes:
3030    
3031      The new pid generation scheme described a few weeks ago was overly
3032      complicated.  I implemented a new mechanism that is simpler and
3033      provides a bit more "stability":  Once CM has seen a compilation
3034      unit, it keeps its identity constant (as long as you do not delete
3035      those crucial CM/GUID/* files).  This means that when you change
3036      an interface, compile, then go back to the old interface, and
3037      compile again, you arrive at the original pid.
3038    
3039      There now also is a mechanism that instructs CM to use the plain
3040      environment hash as a module's pid (effectively making its GUID
3041      the empty string).  For this, "noguid" must be specified as an
3042      option to the .sml file in question within its .cm file.
3043      This is most useful for code that is being generated by tools such
3044      as ml-nlffigen (because during development programmers tend to
3045      erase the tool's entire output directory tree including CM's cached
3046      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
3047      revert to the old, broken behavior of SML/NJ, but in specific cases
3048      where there is no danger of interface confusion, its use is ok
3049      (I think).
3050    
3051      ml-nlffigen by default generates "noguid" annotations.  They can be
3052      turned off by specifying -guid in its command line.
3053    
3054    ----------------------------------------------------------------------
3055    Name: Lal George
3056    Date: 2002/03/12 12 14:42:36 EST
3057    Tag: george-20020312-frequency-computation
3058    Description:
3059    
3060    Integrated jump chaining and static block frequency into the
3061    compiler. More details and numbers later.
3062    
3063    ----------------------------------------------------------------------
3064    Name: Lal George
3065    Date: 2002/03/11 11 22:38:53 EST
3066    Tag: george-20020311-jump-chain-elim
3067    Description:
3068    
3069    Tested the jump chain elimination on all architectures (except the
3070    hppa).  This is on by default right now and is profitable for the
3071    alpha and x86, however, it may not be profitable for the sparc and ppc
3072    when compiling the compiler.
3073    
3074    The gc test will typically jump to a label at the end of the cluster,
3075    where there is another jump to an external cluster containing the actual
3076    code to invoke gc. This is to allow factoring of common gc invocation
3077    sequences. That is to say, we generate:
3078    
3079            f:
3080               testgc
3081               ja   L1      % jump if above to L1
3082    
3083            L1:
3084               jmp L2
3085    
3086    
3087    After jump chain elimination the 'ja L1' instructions is converted to
3088    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
3089    up being implemented in their long form (if L2 is far away) using:
3090    
3091            jbe     L3      % jump if below or equal to L3
3092            jmp     L2
3093         L3:
3094            ...
3095    
3096    
3097    For large compilation units L2  may be far away.
3098    
3099    
3100    ----------------------------------------------------------------------
3101    Name: Matthias Blume
3102    Date: 2002/03/11 13:30:00 EST
3103    Tag: blume-20020311-mltreeeval
3104    Description:
3105    
3106    A functor parameter was missing.
3107    
3108    ----------------------------------------------------------------------
3109    Name: Allen Leung
3110    Date: 2002/03/11 10:30:00 EST
3111    Tag: leunga-20020311-runtime-string0
3112    Description:
3113    
3114       The representation of the empty string now points to a
3115    legal null terminated C string instead of unit.  It is now possible
3116    to convert an ML string into C string with InlineT.CharVector.getData.
3117    This compiles into one single machine instruction.
3118    
3119    ----------------------------------------------------------------------
3120    Name: Allen Leung
3121    Date: 2002/03/10 23:55:00 EST
3122    Tag: leunga-20020310-x86-call
3123    Description:
3124    
3125       Added machine generation for CALL instruction (relative displacement mode)
3126    
3127    ----------------------------------------------------------------------
3128    Name: Matthias Blume
3129    Date: 2002/03/08 16:05:00
3130    Tag: blume-20020308-entrypoints
3131    Description:
3132    
3133    Version number bumped to 110.39.1.  NEW BOOTFILES!
3134    
3135    Entrypoints: non-zero offset into a code object where execution should begin.
3136    
3137    - Added the notion of an entrypoint to CodeObj.
3138    - Added reading/writing of entrypoint info to Binfile.
3139    - Made runtime system bootloader aware of entrypoints.
3140    - Use the address of the label of the first function given to mlriscGen
3141      as the entrypoint.  This address is currently always 0, but it will
3142      not be 0 once we turn on block placement.
3143    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3144      with entry points) from mlriscGen.
3145    
3146    ----------------------------------------------------------------------
3147    Name: Allen Leung
3148    Date: 2002/03/07 20:45:00 EST
3149    Tag: leunga-20020307-x86-cmov
3150    Description:
3151    
3152       Bug fixes for CMOVcc on x86.
3153    
3154       1. Added machine code generation for CMOVcc
3155       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3156       3. CMOVcc cannot have an immediate operand as argument.
3157    
3158    ----------------------------------------------------------------------
3159    Name: Matthias Blume
3160    Date: 2002/03/07 16:15:00 EST
3161    Tag: blume-20020307-controls
3162    Description:
3163    
3164    This is a very large but mostly boring patch which makes (almost)
3165    every tuneable compiler knob (i.e., pretty much everything under
3166    Control.* plus a few other things) configurable via both the command
3167    line and environment variables in the style CM did its configuration
3168    until now.
3169    
3170    Try starting sml with '-h' (or, if you are brave, '-H')
3171    
3172    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3173    implements the underlying generic mechanism.
3174    
3175    The interface to some of the existing such facilities has changed somewhat.
3176    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3177    (The getFoo interface is still there for backward-compatibility, but its
3178    use is deprecated.)
3179    
3180    The ml-build script passes -Cxxx=yyy command-line arguments through so
3181    that one can now twiddle the compiler settings when using this "batch"
3182    compiler.
3183    
3184    TODO items:
3185    
3186    We should go through and throw out all controls that are no longer
3187    connected to anything.  Moreover, we should go through and provide
3188    meaningful (and correct!) documentation strings for those controls
3189    that still are connected.
3190    
3191    Currently, multiple calls to Controls.new are accepted (only the first
3192    has any effect).  Eventually we should make sure that every control
3193    is being made (via Controls.new) exactly once.  Future access can then
3194    be done using Controls.acc.
3195    
3196    Finally, it would probably be a good idea to use the getter-setter
3197    interface to controls rather than ref cells.  For the time being, both
3198    styles are provided by the Controls module, but getter-setter pairs are
3199    better if thread-safety is of any concern because they can be wrapped.
3200    
3201    *****************************************
3202    
3203    One bug fix: The function blockPlacement in three of the MLRISC
3204    backpatch files used to be hard-wired to one of two possibilities at
3205    link time (according to the value of the placementFlag).  But (I
3206    think) it should rather sense the flag every time.
3207    
3208    *****************************************
3209    
3210    Other assorted changes (by other people who did not supply a HISTORY entry):
3211    
3212    1. the cross-module inliner now works much better (Monnier)
3213    2. representation of weights, frequencies, and probabilities in MLRISC
3214       changed in preparation of using those for weighted block placement
3215       (Reppy, George)
3216    
3217    ----------------------------------------------------------------------
3218    Name: Lal George
3219    Date: 2002/03/07 14:44:24 EST 2002
3220    Tag: george-20020307-weighted-block-placement
3221    
3222    Tested the weighted block placement optimization on all architectures
3223    (except the hppa) using AMPL to generate the block and edge frequencies.
3224    Changes were required in the machine properties to correctly
3225    categorize trap instructions. There is an MLRISC flag
3226    "weighted-block-placement" that can be used to enable weighted block
3227    placement, but this will be ineffective without block/edge
3228    frequencies (coming soon).
3229    
3230    
3231    ----------------------------------------------------------------------
3232    Name: Lal George
3233    Date: 2002/03/05 17:24:48 EST
3234    Tag: george-20020305-linkage-cluster
3235    
3236    In order to support the block placement optimization, a new cluster
3237    is generated as the very first cluster (called the linkage cluster).
3238    It contains a single jump to the 'real' entry point for the compilation
3239    unit. Block placement has no effect on the linkage cluster itself, but
3240    all the other clusters  have full freedom in the manner in which they
3241    reorder blocks or functions.
3242    
3243    On the x86 the typical linkage code that is generated is:
3244       ----------------------
3245            .align 2
3246       L0:
3247            addl    $L1-L0, 72(%esp)
3248            jmp     L1
3249    
3250    
3251            .align  2
3252       L1:
3253       ----------------------
3254    
3255    72(%esp) is the memory location for the stdlink register. This
3256    must contain the address of the CPS function being called. In the
3257    above example, it contains the address of  L0; before
3258    calling L1 (the real entry point for the compilation unit), it
3259    must contain the address for L1, and hence
3260    
3261            addl $L1-L0, 72(%esp)
3262    
3263    I have tested this on all architectures except the hppa.The increase
3264    in code size is of course negligible
3265    
3266    ----------------------------------------------------------------------
3267    Name: Allen Leung
3268    Date: 2002/03/03 13:20:00 EST
3269    Tag: leunga-20020303-mlrisc-tools
3270    
3271      Added #[ ... ] expressions to mlrisc tools
3272    
3273    ----------------------------------------------------------------------
3274    Name: Matthias Blume
3275    Date: 2002/02/27 12:29:00 EST
3276    Tag: blume-20020227-cdebug
3277    Description:
3278    
3279    - made types in structure C and C_Debug to be equal
3280    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3281    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3282    
3283    ----------------------------------------------------------------------
3284    Name: Matthias Blume
3285    Date: 2002/02/26 12:00:00 EST
3286    Tag: blume-20020226-ffi
3287    Description:
3288    
3289    1. Fixed a minor bug in CM's "noweb" tool:
3290       If numbering is turned off, then truly don't number (i.e., do not
3291       supply the -L option to noweb).  The previous behavior was to supply
3292       -L'' -- which caused noweb to use the "default" line numbering scheme.
3293       Thanks to Chris Richards for pointing this out (and supplying the fix).
3294    
3295    2. Once again, I reworked some aspects of the FFI:
3296    
3297       A. The incomplete/complete type business:
3298    
3299       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3300         gone!
3301       - ML types representing an incomplete type are now *equal* to
3302         ML types representing their corresponding complete types (just like
3303         in C).  This is still safe because ml-nlffigen will not generate
3304         RTTI for incomplete types, nor will it generate functions that
3305         require access to such RTTI.   But when ML code generated from both
3306         incomplete and complete versions of the C type meet, the ML types
3307         are trivially interoperable.
3308    
3309         NOTE:  These changes restore the full generality of the translation
3310         (which was previously lost when I eliminated functorization)!
3311    
3312       B. Enum types:
3313    
3314       - Structure C now has a type constructor "enum" that is similar to
3315         how the "su" constructor works.  However, "enum" is not a phantom
3316         type because each "T enum" has values (and is isomorphic to
3317         MLRep.Signed.int).
3318       - There are generic access operations for enum objects (using
3319         MLRep.Signed.int).
3320       - ml-nlffigen will generate a structure E_foo for each "enum foo".
3321         * The structure contains the definition of type "mlrep" (the ML-side
3322         representation type of the enum).  Normally, mlrep is the same
3323         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3324         then mlrep will be defined as a datatype -- thus facilitating
3325         pattern matching on mlrep values.
3326         ("-ec" will be suppressed if there are duplicate values in an
3327          enumeration.)
3328         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3329         will be generated for each C enum constant xxx.
3330         * Conversion functions m2i and i2m convert between mlrep and
3331         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
3332         * Coversion functions c and ml convert between mlrep and "tag enum".
3333         * Access functions (get/set) fetch and store mlrep values.
3334       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3335         enumerations are merged into one single enumeration represented by
3336         structure E_'.
3337    
3338    ----------------------------------------------------------------------
3339    Name: Allen Leung
3340    Date: 2002/02/25 04:45:00 EST
3341    Tag: leunga-20020225-cps-spill
3342    
3343    This is a new implementation of the CPS spill phase.
3344    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3345    In case of problems, replace it with the old file spill.sml
3346    
3347    The current compiler runs into some serious performance problems when
3348    constructing a large record.  This can happen when we try to compile a
3349    structure with many items.  Even a very simple structure like the following
3350    makes the compiler slow down.
3351    
3352        structure Foo = struct
3353           val x_1 = 0w1 : Word32.int
3354           val x_2 = 0w2 : Word32.int
3355           val x_3 = 0w3 : Word32.int
3356           ...
3357           val x_N = 0wN : Word32.int
3358        end
3359    
3360    The following table shows the compile time, from N=1000 to N=4000,
3361    with the old compiler:
3362    
3363    N
3364    1000   CPS 100 spill                           0.04u  0.00s  0.00g
3365           MLRISC ra                               0.06u  0.00s  0.05g
3366              (spills = 0 reloads = 0)
3367           TOTAL                                   0.63u  0.07s  0.21g
3368    
3369    1100   CPS 100 spill                           8.25u  0.32s  0.64g
3370           MLRISC ra                               5.68u  0.59s  3.93g
3371              (spills = 0 reloads = 0)
3372           TOTAL                                   14.71u  0.99s  4.81g
3373    
3374    1500   CPS 100 spill                           58.55u  2.34s  1.74g
3375           MLRISC ra                               5.54u  0.65s  3.91g
3376              (spills = 543 reloads = 1082)
3377           TOTAL                                   65.40u  3.13s  6.00g
3378    
3379    2000   CPS 100 spill                           126.69u  4.84s  3.08g
3380           MLRISC ra                               0.80u  0.10s  0.55g
3381              (spills = 42 reloads = 84)
3382           TOTAL                                   129.42u  5.10s  4.13g
3383    
3384    3000   CPS 100 spill                           675.59u  19.03s  11.64g
3385           MLRISC ra                               2.69u  0.27s  1.38g
3386              (spills = 62 reloads = 124)
3387           TOTAL                                   682.48u  19.61s  13.99g
3388    
3389    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
3390           MLRISC ra                               4.96u  0.27s  2.72g
3391              (spills = 85 reloads = 170)
3392           TOTAL                                   2375.26u  57.21s  48.00g
3393    
3394    As you can see the old cps spill module suffers from some serious
3395    performance problem.  But since I cannot decipher the old code fully,
3396    instead of patching the problems up, I'm reimplementing it
3397    with a different algorithm.  The new code is more modular,
3398    smaller when compiled, and substantially faster
3399    (O(n log n) time and O(n) space).  Timing of the new spill module:
3400    
3401    4000  CPS 100 spill                           0.02u  0.00s  0.00g
3402          MLRISC ra                               0.25u  0.02s  0.15g
3403             (spills=1 reloads=3)
3404          TOTAL                                   7.74u  0.34s  1.62g
3405    
3406    Implementation details:
3407    
3408    As far as I can tell, the purpose of the CPS spill module is to make sure the
3409    number of live variables at any program point (the bandwidth)
3410    does not exceed a certain limit, which is determined by the
3411    size of the spill area.
3412    
3413    When the bandwidth is too large, we decrease the register pressure by
3414    packing live variables into spill records.  How we achieve this is
3415    completely different than what we did in the old code.
3416    
3417    First, there is something about the MLRiscGen code generator
3418    that we should be aware of:
3419    
3420    o MLRiscGen performs code motion!
3421    
3422       In particular, it will move floating point computations and
3423       address computations involving only the heap pointer to
3424       their use sites (if there is only a single use).
3425       What this means is that if we have a CPS record construction
3426       statement
3427    
3428           RECORD(k,vl,w,e)
3429    
3430       we should never count the new record address w as live if w
3431       has only one use (which is often the case).
3432    
3433       We should do something similar to floating point, but the transformation
3434       there is much more complex, so I won't deal with that.
3435    
3436    Secondly, there are now two new cps primops at our disposal:
3437    
3438     1. rawrecord of record_kind option
3439        This pure operator allocates some uninitialized storage from the heap.
3440        There are two forms:
3441    
3442         rawrecord NONE [INT n]  allocates a tagless record of length n
3443         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
3444                                     and initializes the tag.
3445    
3446     2. rawupdate of cty
3447          rawupdate cty (v,i,x)
3448          Assigns to x to the ith component of record v.
3449          The storelist is not updated.
3450    
3451    We use these new primops for both spilling and increment record construction.
3452    
3453     1. Spilling.
3454    
3455        This is implemented with a linear scan algorithm (but generalized
3456        to trees).  The algorithm will create a single spill record at the
3457        beginning of the cps function and use rawupdate to spill to it,
3458        and SELECT or SELp to reload from it.  So both spills and reloads
3459        are fine-grain operations.  In contrast, in the old algorithm
3460        "spills" have to be bundled together in records.
3461    
3462        Ideally, we should sink the spill record construction to where
3463        it is needed.  We can even split the spill record into multiple ones
3464        at the places where they are needed.  But CPS is not a good
3465        representation for global code motion, so I'll keep it simple and
3466        am not attempting this.
3467    
3468     2. Incremental record construction (aka record splitting).
3469    
3470        Long records with many component values which are simulatenously live
3471        (recall that single use record addresses are not considered to
3472         be live) are constructed with rawrecord and rawupdate.
3473        We allocate space on the heap with rawrecord first, then gradually
3474        fill it in with rawupdate.  This is the technique suggested to me
3475        by Matthias.
3476    
3477        Some restrictions on when this is applicable:
3478        1. It is not a VECTOR record.  The code generator currently does not handle
3479           this case. VECTOR record uses double indirection like arrays.
3480        2. All the record component values are defined in the same "basic block"
3481           as the record constructor.  This is to prevent speculative
3482           record construction.
3483    
3484    ----------------------------------------------------------------------
3485    Name: Allen Leung
3486    Date: 2002/02/22 01:02:00 EST
3487    Tag: leunga-20020222-mlrisc-tools
3488    
3489    Minor bug fixes in the parser and rewriter
3490    
3491    ----------------------------------------------------------------------
3492    Name: Allen Leung
3493    Date: 2002/02/21 20:20:00 EST
3494    Tag: leunga-20020221-peephole
3495    
3496    Regenerated the peephole files.  Some contained typos in the specification
3497    and some didn't compile because of pretty printing bugs in the old version
3498    of 'nowhere'.
3499    
3500    ----------------------------------------------------------------------
3501    Name: Allen Leung
3502    Date: 2002/02/19 20:20:00 EST
3503    Tag: leunga-20020219-mlrisc-tools
3504    Description:
3505    
3506       Minor bug fixes to the mlrisc-tools library:
3507    
3508       1.  Fixed up parsing colon suffixed keywords
3509       2.  Added the ability to shut the error messages up
3510       3.  Reimplemented the pretty printer and fixed up/improved
3511           the pretty printing of handle and -> types.
3512       4.  Fixed up generation of literal symbols in the nowhere tool.
3513       5.  Added some SML keywords to to sml.sty
3514    
3515    ----------------------------------------------------------------------
3516    Name: Matthias Blume
3517    Date: 2002/02/19 16:20:00 EST
3518    Tag: blume-20020219-cmffi
3519    Description:
3520    
3521    A wild mix of changes, some minor, some major:
3522    
3523    * All C FFI-related libraries are now anchored under $c:
3524        $/c.cm      --> $c/c.cm
3525        $/c-int.cm  --> $c/internals/c-int.cm
3526        $/memory.cm --> $c/memory/memory.cm
3527    
3528    * "make" tool (in CM) now treats its argument pathname slightly
3529      differently:
3530        1. If the native expansion is an absolute name, then before invoking
3531           the "make" command on it, CM will apply OS.Path.mkRelative
3532           (with relativeTo = OS.FileSys.getDir()) to it.
3533        2. The argument will be passed through to subsequent phases of CM
3534           processing without "going native".  In particular, if the argument
3535           was an anchored path, then "make" will not lose track of that anchor.
3536    
3537    * Compiler backends now "know" their respective C calling conventions
3538      instead of having to be told about it by ml-nlffigen.  This relieves
3539      ml-nlffigen from one of its burdens.
3540    
3541    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
3542    
3543    * Export C_DEBUG and C_Debug from $c/c.cm.
3544    
3545    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
3546      subtyping relationship between incomplete pointers and their complete
3547      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
3548      with the convention of instantiating 'o with ('t, 'c) obj whenever
3549      the pointer target type is complete.  In the incomplete case, 'o
3550      will be instantiated with some "'c iobj" -- a type obtained by
3551      using one of the functors PointerToIncompleteType or PointerToCompleteType.
3552    
3553      Operations that work on both incomplete and complete pointer types are
3554      typed as taking an 'o ptr while operations that require the target to
3555      be known are typed as taking some ('t, 'c) obj ptr.
3556    
3557      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
3558      where void is an eqtype without any values.  This makes it possible
3559      to work on voidptr values using functions meant to operate on light
3560      incomplete pointers.
3561    
3562    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
3563      been vastly simplified.
3564    
3565    ----------------------------------------------------------------------
3566    Name: Matthias Blume
3567    Date: 2002/02/19 10:48:00 EST
3568    Tag: blume-20020219-pqfix
3569    Description:
3570    
3571    Applied Chris Okasaki's bug fix for priority queues.
3572    
3573    ----------------------------------------------------------------------
3574    Name: Matthias Blume
3575    Date: 2002/02/15 17:05:00
3576    Tag: Release_110_39
3577    Description:
3578    
3579    Last-minute retagging is becoming a tradition... :-(
3580    
3581    This is the working release 110.39.
3582    
3583    ----------------------------------------------------------------------
3584    Name: Matthias Blume
3585    Date: 2002/02/15 16:00:00 EST
3586    Tag: Release_110_39-orig
3587    Description:
3588    
3589    Working release 110.39.  New bootfiles.
3590    
3591    (Update: There was a small bug in the installer so it wouldn't work
3592    with all shells.  So I retagged. -Matthias)
3593    
3594    ----------------------------------------------------------------------
3595    Name: Matthias Blume
3596    Date: 2002/02/15 14:17:00 EST
3597    Tag: blume-20020215-showbindings
3598    Description:
3599    
3600    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
3601    the latter can be useful for exploring what bindings are available at
3602    the interactive prompt.  (The first function returns only the list
3603    of symbols that are really bound, the second prints those but also the
3604    ones that CM's autoloading mechanism knows about.)
3605    
3606    ----------------------------------------------------------------------
3607    Name: Matthias Blume
3608    Date: 2002/02/15 12:08:00 EST
3609    Tag: blume-20020215-iptrs
3610    Description:
3611    
3612    Two improvements to ml-nlffigen:
3613    
3614      1. Write files only if they do not exist or if their current contents
3615         do not coincide with what's being written.  (That is, avoid messing
3616         with the time stamps unless absolutely necessary.)
3617    
3618      2. Implement a "repository" mechanism for generated files related
3619         to "incomplete pointer types".   See the README file for details.
3620    
3621    ----------------------------------------------------------------------
3622    Name: Matthias Blume
3623    Date: 2002/02/14 11:50:00 EST
3624    Tag: blume-20020214-quote
3625    Description:
3626    
3627    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
3628    because of the new and improved tag generation scheme.  (Thanks to Allen
3629    Leung for pointing it out.)
3630    
3631    ----------------------------------------------------------------------
3632    Name: Lal George
3633    Date: 2002/02/14 09:55:27 EST 2002
3634    Tag: george-20020214-isabelle-bug
3635    Description:
3636    
3637    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
3638    of Isabelle on the x86.
3639    
3640    From Allen:
3641    -----------
3642     I've found the problem:
3643    
3644         in ra-core.sml, I use the counter "blocked" to keep track of the
3645         true number of elements in the freeze queue.  When the counter goes
3646         to zero, I skip examining the queue.  But I've messed up the
3647         bookkeeping in combine():
3648    
3649             else ();
3650             case !ucol of
3651               PSEUDO => (if !cntv > 0 then
3652                     (if !cntu > 0 then blocked := !blocked - 1 else ();
3653                                        ^^^^^^^^^^^^^^^^^^^^^^^
3654                      moveu := mergeMoveList(!movev, !moveu)
3655                     )
3656                  else ();
3657    
3658         combine() is called to coalesce two nodes u and v.
3659         I think I was thinking that if the move counts of u and v are both
3660         greater than zero then after they are coalesced then one node is
3661         removed from the freeze queue.  Apparently I was thinking that
3662         both u and v are of low degree, but that's clearly not necessarily true.
3663    
3664    
3665    02/12/2002:
3666        Here's the patch.  HOL now compiles.
3667    
3668        I don't know how this impact on performance (compile
3669        time or runtime).  This bug caused the RA (especially on the x86)
3670        to go thru the potential spill phase when there are still nodes on the
3671        freeze queue.
3672    
3673    
3674    
3675    
3676    ----------------------------------------------------------------------
3677    Name: Matthias Blume
3678    Date: 2002/02/13 22:40:00 EST
3679    Tag: blume-20020213-fptr-rtti
3680    Description:
3681    
3682    Fixed a bug in ml-nlffigen that was introduced with one of the previous
3683    updates.
3684    
3685    ----------------------------------------------------------------------
3686    Name: Matthias Blume
3687    Date: 2002/02/13 16:41:00 EST
3688    Tag: blume-20020213-cmlpq
3689    Description:
3690    
3691    Added new priority queue export symbols (which have just been added to
3692    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
3693    not compile and the installer would choke.)
3694    
3695    ----------------------------------------------------------------------
3696    Name: Matthias Blume
3697    Date: 2002/02/13 16:15:00 EST
3698    Tag: blume-20020213-various
3699    Description:
3700    
3701    1. More tweaks to ml-nlffigen:
3702    
3703       - better internal datastructures (resulting in slight speedup)
3704       - "-match" option requires exact match
3705       - "localized" gensym counters (untagged structs/unions nested within
3706         other structs/unions or within typedefs get a fresh counter; their
3707         tag will be prefixed by a concatenation of their parents' tags)
3708       - bug fixes (related to calculation of transitive closure of types
3709         to be included in the output)
3710    
3711    2. Minor Basis updates:
3712    
3713       - added implementations for List.collate and Option.app
3714    
3715    ----------------------------------------------------------------------
3716    Name: Matthias Blume
3717    Date: 2002/02/11 15:55:00 EST
3718    Tag: blume-20020211-gensym
3719    Description:
3720    
3721    Added a "-gensym" option to command line of ml-nlffigen.  This can be
3722    used to specify a "stem" -- a string that is inserted in all "gensym'd"
3723    names (ML structure names that correspond to unnamed C structs, unions,
3724    and enums), so that separate runs of ml-nlffigen do not clash.
3725    
3726    ----------------------------------------------------------------------
3727    Name: Matthias Blume
3728    Date: 2002/02/11 12:05:00 EST
3729    Tag: blume-20020211-gensml
3730    Description:
3731    
3732    A quick fix for a problem with GenSML (in the pgraph-util library):
3733    Make generation of toplevel "local" optional.  (Strictly speaking,
3734    signature definitions within "local" are not legal SML.)
3735    
3736    Other than that: updates to INSTALL and cm/TODO.
3737    
3738    ----------------------------------------------------------------------
3739    Name: Matthias Blume
3740    Date: 2002/02/08 15:00:00 EST
3741    Tag: blume-20020208-uniquepid
3742    Description:
3743    
3744    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
3745    
3746    1. The installer (config/install.sh) has gotten smarter:
3747    
3748         - Configuration options are a bit easier to specify now
3749           (in config/targets).
3750         - Bug in recognizing .tar.bz2 files fixed.
3751         - Installer automatically resolves dependencies between
3752           configuration options (e.g., if you ask for eXene, you will
3753           also get cml -- regardless whether you asked for it or not).
3754         - Installer can run in "quieter mode" by setting the environment
3755           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
3756           "completely silent", though.
3757         - Build HashCons library as part of smlnj-lib.
3758    
3759    2. A new scheme for assigning persistent identifiers to compilation
3760       units (and, by extension, to types etc.) has been put into place.
3761       This fixes a long-standing bug where types and even dynamic values
3762       can get internally confused, thereby compromising type safety
3763       (abstraction) and dynamic correctness.  See
3764    
3765         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
3766    
3767       for an example of how things could go wrong until now.
3768    
3769       The downside of the new scheme is that pids are not quite as
3770       persistent as they used to be: CM will generate a fresh pid
3771       for every compilation unit that it thinks it sees for the first
3772       time.  That means that if you compile starting from a clean, fresh
3773       source tree at two different times, you end up with different
3774       binaries.
3775    
3776       Cutoff recompilation, however, has not been compromised because
3777       CM keeps pid information in special caches between runs.
3778    
3779    ----------------------------------------------------------------------
3780    Name: Lal George
3781    Date: 2002/02/07 15:34:13 EST 2002
3782    Tag: <none>
3783    Description:
3784    
3785    Compilers that generate assembly code may produce  global labels
3786    whose value is resolved at link time. The various peephole optimization
3787    modules did not take this in account.
3788    
3789    TODO. The Labels.addrOf function should really return an option
3790    type so that clients are forced to deal with this issue, rather
3791    than an exception being raised.
3792    
3793    ----------------------------------------------------------------------
3794    Name: Lal George
3795    Date: 2002/02/06 13:55:02 EST
3796    Tag: george-20020206-ra-breakup
3797    Description:
3798    
3799    1. A bug fix from Allen.
3800    
3801        A typo causes extra fstp %st(0)'s to be generated at compensation
3802        edges, which might cause stack underflow traps at runtime.  This
3803        occurs in fft where there are extraneous fstps right before the 'into'
3804        trap instruction (in this case they are harmless since none of the
3805        integers overflow.)
3806    
3807    2. Pulled out various utility modules that were embedded in the modules
3808       of the register allocator. I need these modules for other purposes, but
3809       they are not complete enough to put into a library (just yet).
3810    ----------------------------------------------------------------------
3811    Name: Matthias Blume
3812    Date: 2002/01/31 16:05:00 EST
3813    Tag: blume-20020131-sparc-ccalls
3814    Description:
3815    
3816    1. C-calls on Sparc needlessly allocated a huge chunk (96 bytes)
3817       of extra stack space by mistake.  Fixed.
3818    
3819    2. Bug in logic of handling of command-line options in ml-nlffigen fixed.
3820    
3821    ----------------------------------------------------------------------
3822    Name: Allen Leung
3823    Date: 2002/01/30
3824    Tag: leunga-20020130-nowhere-bug-fix
3825    Description:
3826    
3827       MLRISC bug fixes:
3828       1. Fixed a bindings computation bug in the 'nowhere' program generator tool.
3829       2. MachineInt.fromString was negating its value.
3830    
3831    ----------------------------------------------------------------------
3832    Name: Matthias Blume
3833    Date: 2002/01/29
3834    Tag: blume-20020129-INSTALL
3835    Description:
3836    
3837    - Added somewhat detailed installation instructions (file INSTALL).
3838    - Fixed curl-detection bug in config/install.sh.
3839    - It is now possible to select the URL getter using the URLGETTER
3840      environment variable:
3841    
3842          not set / "unknown"      --> automatic detection (script tries wget,
3843                                       curl, and lynx)
3844          "wget" / "curl" / "lynx" --> use the specified program (script "knows"
3845                                       how to properly invoke them)
3846          other                    --> use $URLGETTER directly, it must take
3847                                       precisely two command-line arguments
3848                                       (source URL and destination file name)
3849    
3850    ----------------------------------------------------------------------
3851    Name: Matthias Blume
3852    Date: 2002/01/28
3853    Tag: blume-20020128-sparc-ccalls
3854    Description:
3855    
3856    - Fixed problem with calculation of "used" registers in sparc-c-calls.
3857    - Make use of the allocParam argument in sparc-c-calls.
3858    
3859    ----------------------------------------------------------------------
3860    Name: Matthias Blume
3861    Date: 2002/01/28
3862    Tag: blume-20020128-allocParam
3863    Description:
3864    
3865    John Reppy:  Changes c-calls API to accept client-callback for
3866    allocating extra stack space.
3867    me: Corresponding changes to mlriscGen (using a dummy argument that
3868        does not change the current behavior).
3869    
3870    ----------------------------------------------------------------------
3871    Name: Matthias Blume
3872    Date: 2002/01/28 12:00:00
3873    Tag: Release_110_38
3874    Description:
3875    
3876    This time for real!!!
3877    
3878    ----------------------------------------------------------------------
3879    Name: Matthias Blume
3880    Date: 2002/01/28 10:56:00 EST
3881    Tag: blume-20020128-retraction
3882    Description:
3883    
3884    0. Retracted earlier 110.38.  (The Release_110_38 tag has been replaced
3885       with blume-Release_110_38-retracted.)
3886    
3887    1. Fixed a problem with incorrect rounding modes in real64.sml.
3888       (Thanks to Andrew Mccreight <andrew.mccreight@yale.edu>.)
3889    
3890    2. A bug in ml-nlffigen related to the handling of unnamed structs, unions,
3891       and enums fixed.  The naming of corresponding ML identifiers should
3892       now be consistent again.
3893    
3894    ----------------------------------------------------------------------
3895    Name: Allen Leung
3896    Date: 2002/01/27
3897    Tag: leunga-20020127-nowhere
3898    Description:
3899    
3900       Added a target called nowhere in the configuration scripts.
3901       Enabling this will build the MLRISC 'nowhere' tool (for translating
3902       programs with where-clauses into legal SML code) during installation.
3903    
3904    ----------------------------------------------------------------------
3905    Name: Matthias Blume
3906    Date: 2002/01/25 21:27:00 EST
3907    Tag: blume-Release_110_38-retracted
3908    Description:
3909    
3910    Call it a (working) release!  Version is 110.38. Bootfiles are ready.
3911    
3912    README will be added later.
3913    
3914    !!! NOTE:  Re-tagged as blume-Release_110_38-retracted. Original tag
3915    (Release_110_38) removed.  Reason: Last-minute bug fixes.
3916    
3917    ----------------------------------------------------------------------
3918    Name: Matthias Blume
3919    Date: 2002/01/25
3920    Tag: blume-20020125-ffi
3921    Description:
3922    
3923    A large number of tweaks and improvements to ml-nlffi-lib and
3924    ml-nlffigen:
3925    
3926       - ML represenation types have been streamlined
3927       - getter and setter functions work with concrete values, not abstract
3928         ones where possible
3929       - ml-nlffigen command line more flexible (see README file there)
3930       - some bugs have been fixed (hopefully)
3931    
3932    ----------------------------------------------------------------------
3933    Name: Lal George
3934    Date: 2002/01/24
3935    Tag: george-20020124-risc-ra-interface
3936    Description:
3937    
3938       There is a dramatic simplification in the interface to the
3939       register allocator for RISC architectures as a result of making
3940       parallel copy instructions explicit.
3941    
3942    ----------------------------------------------------------------------
3943    Name: Matthias Blume
3944    Date: 2002/01/22
3945    Tag: blume-20020122-x86-ccalls
3946    Description:
3947    
3948    Bug fix for c-calls on x86 (having to do with how char- and
3949    short-arguments are being handled).
3950    
3951    ----------------------------------------------------------------------
3952    Name: Matthias Blume
3953    Date: 2002/01/21
3954    Tag: blume-20020121-ff
3955    Description:
3956    
3957    Another day of fiddling with the FFI...
3958    
3959    1. Bug fix/workaround:  CKIT does not complain about negative array
3960       dimensions, so ml-nlffigen has to guard itself against this possibility.
3961       (Otherwise a negative dimension would send it into an infinite loop.)
3962    
3963    2. Some of the abstract types (light objects, light pointers, most "base"
3964       types) in structure C are now eqtypes.
3965    
3966    3. Added constructors and test functions for NULL function pointers.
3967    
3968    ----------------------------------------------------------------------
3969    Name: Matthias Blume
3970    Date: 2002/01/18
3971    Tag: blume-20020118-ready-for-new-release
3972    Description:
3973    
3974    Made config/srcarchiveurl point to a new place.  (Will provide boot
3975    files shortly.)
3976    
3977    Maybe we christen this to be 110.38?
3978    
3979    ----------------------------------------------------------------------
3980    Name: Matthias Blume
3981    Date: 2002/01/18
3982    Tag: blume-20020118-more-ffifiddle
3983    Description:
3984    
3985    Today's FFI fiddling:
3986    
3987      - Provided a structure CGetSet with "convenient" versions of C.Get.* and
3988        C.Set.* that use concrete (MLRep.*) arguments and results instead
3989        of abstract ones.
3990    
3991      - Provided word-style bit operations etc. for "int" representation
3992        types in MLRep.S<Foo>Bitops where <Foo> ranges over Char, Int, Short,
3993        and Long.
3994    
3995    ----------------------------------------------------------------------
3996    Name: Matthias Blume
3997    Date: 2002/01/18
3998    Tag: blume-20020118-use-x86-fp
3999    Description:
4000    
4001    Now that x86-fast-fp seems to be working, I turned it back on again
4002    by default.  (Seems to work fine now, even with the FFI.)
4003    
4004    Other than that, I added some documentation about the FFI to
4005    src/ml-nlffigen/README and updated the FFI test examples in
4006    src/ml-nlffi-lib/Tests/*.
4007    
4008    ----------------------------------------------------------------------
4009    Name: Allen Leung
4010    Date: 2002/01/17
4011    Tag: leunga-20020117-x86-fast-fp-call
4012    Description:
4013    
4014       1. Fixed a problem with handling return fp values when x86's fast fp
4015          mode is turned on.
4016    
4017       2. Minor pretty printing fix for cellset.  Print %st(0) as %st(0) instead
4018          of %f32.
4019    
4020       3. Added a constructor INT32lit to the ast of MLRISC tools.
4021    
4022    ----------------------------------------------------------------------
4023    Name: Matthias Blume
4024    Date: 2002/01/16
4025    Tag: blume-20020116-ffifiddle
4026    Description:
4027    
4028    More fiddling with the FFI interface:
4029    
4030     - Make constness 'c instead of rw wherever possible.  This eliminates
4031       the need for certain explicit coercions.  (However, due to ML's
4032       value polymorphism, there will still be many cases where explicit
4033       coercions are necessary.  Phantom types are not the whole answer
4034       to modeling a subtyping relationship in ML.)
4035    
4036     - ro/rw coersions for pointers added.  (Avoids the detour through */&.)
4037    
4038     - "printf" test example added to src/ml-nlffi-lib/Tests.  (Demonstrates
4039       clumsy workaround for varargs problem.)
4040    
4041    ----------------------------------------------------------------------
4042    Name: Lal George
4043    Date: 2002/01/15
4044    Tag: <none>
4045    Description:
4046    
4047    1. Since COPY instructions are no longer native to the architecture,
4048       a generic functor can be used to implement the expandCopies function.
4049    
4050    2. Allowed EXPORT and IMPORT pseudo-op declarations to appear inside a
4051       TEXT segment.
4052    
4053    ----------------------------------------------------------------------
4054    Name: Matthias Blume
4055    Date: 2002/01/15
4056    Tag: blume-20020115-ffiupdates
4057    Description:
4058    
4059    1. Fix for bug resulting in single-precision float values being returned
4060       incorrectly from FFI calls.
4061    
4062    2. Small modifications to C FFI API:
4063    
4064        - memory-allocation routines return straight objects (no options)
4065          and raise an exception in out-of-memory situations
4066        - unsafe extensions to cast between function pointers and pointers
4067          from/to ints
4068        - added structure C_Debug as an alternative to structure C where
4069          pointer-dereferencing (|*| and |*!) always check for null-pointers
4070        - added open_lib' to DynLinkage;  open_lib' works like open_lib
4071          but also takes a (possibly empty) list of existing library handles
4072          that the current library depends on
4073    
4074    ----------------------------------------------------------------------
4075    Name: Matthias Blume
4076    Date: 2002/01/10
4077    Tag: blume-20020110-newffigen
4078    Description:
4079    
4080    1. Updates to portable graph code.
4081    
4082    2. Major update to ml-nlffigen and ml-nlffi-lib.  Things are much
4083       more scalable now so that even huge interfaces such as the one
4084       for GTK compile in finite time and space. :-)
4085       See src/ml-nlffigen/README for details on what's new.
4086    
4087    ----------------------------------------------------------------------
4088    Name: Lal George
4089    Date: 2001/01/09 14:31:35 EST 2002
4090    Tag: george-20011206-rm-native-copy
4091    Description:
4092    
4093            Removed the native COPY and FCOPY instructions
4094            from all the architectures and replaced it with the
4095            explicit COPY instruction from the previous commit.
4096    
4097            It is now possible to simplify many of the optimizations
4098            modules that manipulate copies. This has not been
4099            done in this change.
4100    
4101