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

SCM Repository

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

Diff of /sml/trunk/HISTORY

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

revision 572, Thu Mar 9 02:43:06 2000 UTC revision 1856, Wed Nov 2 22:24:39 2005 UTC
# Line 8  Line 8 
8  The form of an entry should be:  The form of an entry should be:
9    
10  Name:  Name:
11  Date:  Date: yyyy/mm/dd
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    
15  ----------------------------------------------------------------------  ----------------------------------------------------------------------
16    Name: Matthias Blume (http://www.tti-c.org/blume/)
17    Date: 2005/11/02 16:25:00 CST
18    Tag: blume-20051102-hex
19    Description:
20    
21    * Runtime system bootstrap code now accepts hex digits in BOOTLIST
22      in either upper case or lower case format.
23    * Pushed changes to names of Pack<N>{Big,Little} structures through
24      CML and eXene.
25    
26    ----------------------------------------------------------------------
27    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
28    Date: 2005/11/02
29    Description:
30    Fixed a Basis incompatibility: Pack<N>{Big,Little} structures should
31    be called PackWord<N>{Big,Little}.
32    
33    ----------------------------------------------------------------------
34    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
35    Date: 2005/10/28
36    Description:
37    Fixed a minor Basis incompatibility: hex digits should be upper case.
38    
39    ----------------------------------------------------------------------
40    Name: Matthias Blume (blume (at) tti - c (dot) org)
41    Date: 2005/10/25 16:00:00 CDT
42    Tag: Release_110_56
43    Description:
44    
45    New working version (110.56).  NEW BOOTFILES!
46    
47    ----------------------------------------------------------------------
48    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
49    Date: 2005/10/25
50    Description:
51    Added interval sets to utility library (signatures INTERVAL_DOMAIN
52    and INTERVAL_SET, and functor IntervalSetFn).
53    
54    ----------------------------------------------------------------------
55    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
56    Date: 2005/10/14
57    Description:
58    Add Zhong Shao's fix for datatype equality functions.
59    
60    ----------------------------------------------------------------------
61    Name: Allen Leung (leunga (at) reservoir (dot) com)
62    Date: 2005/10/14 12:30:00 EST
63    Tag: leunga-20051014-x86-spill
64    Description:
65    
66    Bug fix a bug found by Carl Hauser.
67    There was a typo in the reload code for FCMP in x86SpillInstr.sml.
68    
69    ----------------------------------------------------------------------
70    Name: Allen Leung (leunga (at) reservoir (dot) com)
71    Date: 2005/10/14 11:15:00 EST
72    Tag: leunga-20051014-x86-asm
73    Description:
74    
75    Removed some debugging code in file x86Asm.sml.
76    The function emit_operand was printing out debugging output.
77    
78    ----------------------------------------------------------------------
79    Name: John Reppy
80    Date: 2005/07/27
81    Tag:
82    Description:
83    
84    Fixed ml-lex to recognize "\r" as representing carriage return.
85    
86    ----------------------------------------------------------------------
87    Name: John Reppy
88    Date: 2005/07/27
89    Tag:
90    Description:
91    
92    Fixed ml-yacc to work on files with non-native end-of-line
93    encodings (e.g., Windows text file on a Unix system).
94    
95    ----------------------------------------------------------------------
96    Name: John Reppy
97    Date: 2005/07/20
98    Tag:
99    Description:
100    
101    Added changes from Dominic Evans (oldmanuk (at) gmail (dot) com)
102    to support HPUX 11.
103    
104    ----------------------------------------------------------------------
105    Name: John Reppy
106    Date: 2005/07/06
107    Tag:
108    Description:
109    
110    Changes to the SML/NJ library.  See smlnj-lib/CHANGES for details.
111    
112    ----------------------------------------------------------------------
113    Name: Matthias Blume (blume (at) tti - c (dot) org)
114    Date: 2005/07/06 00:45:00 CDT
115    Tag: blume-20050706-slice-copy
116    Description:
117    
118    Fixed reversed logic for deciding whether to "copy up" or "copy down"
119    in *-array-slice.sml.
120    
121    ----------------------------------------------------------------------
122    Name: Allen Leung (leunga (at) reservoir (dot) com)
123    Date: 2005/05/31 17:00:00 EST
124    Tag: leunga-20050531-cygwin-fault-2
125    Description:
126    
127    A typo in the cygwin code fixed.
128    
129    ----------------------------------------------------------------------
130    Name: Allen Leung (leunga (at) reservoir (dot) com)
131    Date: 2005/05/31 16:47:00 EST
132    Tag: leunga-20050531-cygwin-fault
133    Description:
134    
135    Updated Cygwin's fault/signal handling to match the Windows version.
136    Updated the export list.
137    
138    ----------------------------------------------------------------------
139    Name: Matthias Blume (blume (at) tti - c (dot) org)
140    Date: 2005/05/18 13:35:00 CDT
141    Tag: Release_110_54
142    Description:
143    
144    New working version (110.54).  NEW BOOTFILES!
145    
146    ----------------------------------------------------------------------
147    Name: Matthias Blume (blume (at) tti - c (dot) org)
148    Date: 2005/05/18 11:58:00 CDT
149    Tag: blume-20050518-installer
150    Description:
151    
152    Added support scripts for Mac OS X PackageMaker and modified
153    config/install.sh so that it supports re-dumping a heap image
154    after customization.
155    
156    ----------------------------------------------------------------------
157    Name: Matthias Blume (blume (at) tti - c (dot) org)
158    Date: 2005/05/18 10:55:00 CDT
159    Tag: blume-20050518-realdiv-noovld
160    Description:
161    
162    Un-overloaded / to work around bug in overloading resolution code.
163    
164    ----------------------------------------------------------------------
165    Name: Matthias Blume (blume (at) tti - c (dot) org)
166    Date: 2005/05/16 23:50:00 CDT
167    Tag: blume-20050516-redump-heap
168    Description:
169    
170    Added mechanism for re-creating a heap file for the interactive system
171    after configuration variables have been changed.
172    
173       CM.redump_heap : string -> unit
174    
175    This is much like SMLofNJ.exportML, but starting from the resulting
176    heap does not return to the caller of CM.redump_heap but
177    restarts the interactive system from scratch.  The original call of
178    CM.redump_heap does not return but ends the interactive session.
179    Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
180    
181    Internally, redump_heap winds the dynamic execution context back to
182    the point where the original heap image was created and re-executes
183    the heap image generation code in the boot code.
184    
185    ----------------------------------------------------------------------
186    Name: Matthias Blume (blume (at) tti - c (dot) org)
187    Date: 2005/05/09 21:55:00 CDT
188    Tag: blume-20050509-word64
189    Description:
190    
191    Added a hack to the existing hack known as Word64 to make fromString
192    behave correctly.  I am still not sure whether Word64.scan will work
193    as specified with respect to the interaction of radix and prefix.
194    
195    ----------------------------------------------------------------------
196    Name: Allen Leung (leunga (at) reservoir (dot) com)
197    Date: 2005/05/04 11:50:00 EST
198    Tag: leunga-20050504-checkgc
199    Description:
200    
201    Added a gc protocol checking phase.  This phase is enabled with
202    the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
203    
204    ----------------------------------------------------------------------
205    Name: Matthias Blume (blume (at) tti - c (dot) org)
206    Date: 2005/05/04 10:45:00 CDT
207    Tag: blume-20050504-intinf
208    Description:
209    
210    Fixed a bug in the implementation of div and mod for IntInf.
211    Thanks to Neophytos Michael for reporting the problem.
212    
213    ----------------------------------------------------------------------
214    Name: Matthias Blume (blume (at) tti - c (dot) org)
215    Date: 2005/05/04 10:35:00 CDT
216    Tag: blume-20050504-join
217    Description:
218    
219    Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
220    
221    ----------------------------------------------------------------------
222    Name: Matthias Blume (blume (at) tti - c (dot) org)
223    Date: 2005/02/28 23:40:00 CST
224    Tag: blume-20050228-mVar
225    Description:
226    
227    Fixed serious bug (brown paper bag variety) in new implementation of
228    structure Atom in CML.  (I had accidentally used a mailbox instead of
229    an mvar, leaving the door open for races.)
230    
231    ----------------------------------------------------------------------
232    Name: Matthias Blume (blume (at) tti - c (dot) org)
233    Date: 2005/02/25 15:00:00 CST
234    Tag: Release_110_53
235    Description:
236    
237    New working version (110.53).  NEW BOOTFILES!
238    
239    ----------------------------------------------------------------------
240    Name: Matthias Blume (blume (at) tti - c (dot) org)
241    Date: 2005/02/25 14:50:00 CST
242    Tag: blume-20050225-susp
243    Description:
244    
245    Brought back SMLofNJ.Susp.  The underlying suspension type is the one
246    implemented in Core -- which means that it is the same as the one used
247    by the lazy extension.
248    
249    ----------------------------------------------------------------------
250    Name: Matthias Blume (blume (at) tti - c (dot) org)
251    Date: 2005/02/24 16:50:00 CST
252    Tag: blume-20050224-cml-atom
253    Description:
254    
255    Simpler and at the same time more general implementation of structure
256    Atom in CML.
257    
258    ----------------------------------------------------------------------
259    Name: Matthias Blume (blume (at) tti - c (dot) org)
260    Date: 2005/02/15 17:35:00 CST
261    Tag: blume-20050215-tools
262    Description:
263    
264    Created new "tools" directory under "src" and moved "TraceDebugProf"
265    there.
266    
267    ----------------------------------------------------------------------
268    Name: Matthias Blume (blume (at) tti - c (dot) org)
269    Date: 2005/02/10 17:55:00 CST
270    Tag: blume-20050210-longlong
271    Description:
272    
273    Implemented "long long" arguments and results for NLFFI.  (Only the
274    PPC/MacOS implementation is complete, the other backends still need to
275    be updated.)
276    
277    ----------------------------------------------------------------------
278    Name: Matthias Blume (blume (at) tti - c (dot) org)
279    Date: 2005/01/24 17:40:00 CST
280    Tag: blume-20050124-mlyacc
281    Description:
282    
283    Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
284    problem with certain "as" patterns which previously got rendered
285    using incorrect syntax.
286    
287    ----------------------------------------------------------------------
288    Name: Matthias Blume (blume (at) tti - c (dot) org)
289    Date: 2005/01/18 12:00:00 CST
290    Tag: blume-20050118-profile
291    Description:
292    
293    Made time profiling code (interrupt handler) in runtime system aware
294    of new array representation.
295    
296    ----------------------------------------------------------------------
297    Name: Matthias Blume (blume (at) tti - c (dot) org)
298    Date: 2005/01/14 18:00:00 CST
299    Tag: blume-20050114-heap2exec
300    Description:
301    
302    Implemented new (but still experimental) heap2exec facility.  This is
303    tested under Mac OS X and should work under Linux (will test shortly).
304    It will probably also work on the Sparc (will test some time later).
305    
306      - removed old "HACKED_STANDALONE" hack from runtime
307    
308    To be able to test this, uncomment the request for "heap2asm" in
309    config/targets prior to installation.  (Notice that this is different
310    from "heap2exec" mentioned below.  Not a typo.)
311    
312    To perform an actual test, run the command
313    
314       $ bin/heap2exec heapfile execfile
315    
316    (You can put heap2exec on your shell's path.)
317    
318    For example, run
319    
320      $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
321    
322    This will create a standalone executable called "mly" which you can
323    then invoke directly as a command.
324    
325    ----------------------------------------------------------------------
326    Name: Matthias Blume (blume (at) tti - c (dot) org)
327    Date: 2005/01/07 11:44:00 CST
328    Tag: blume-20050107-mlstring
329    Description:
330    
331    fixed off-by-one error in ML_STRING macro (globals.c)
332    
333    ----------------------------------------------------------------------
334    Name: Matthias Blume (blume (at) tti - c (dot) org)
335    Date: 2004/12/23 18:00:00 CST
336    Tag: blume-20041223-santa
337    Description:
338    
339    Made ml-build script "smarter" (but only very little).
340    
341    ----------------------------------------------------------------------
342    Name: Matthias Blume (blume (at) tti - c (dot) org)
343    Date: 2004/12/21 15:05:00 CST
344    Tag: blume-20041221-longlong
345    Description:
346    
347    * Implemented access to signed and unsigned long long data in NLFFI.
348    (The parameter-passing part of the picture has not complete. But
349    data structure access seems to work.)
350    
351    * Fixed CM's incorrect assumption that the PPC is little-endian.
352      (On the Mac, it is big-endian.  And that's currently our only
353      PPC platform.)
354    
355    ----------------------------------------------------------------------
356    Name: Matthias Blume (blume (at) tti - c (dot) org)
357    Date: 2004/12/21 12:50:00 CST
358    Tag: blume-20041221-memory
359    Description:
360    
361    Some cleanup in the $c/memory.cm library: separated some concerns by
362    moving allocation code and memory access code each into their own
363    files.
364    
365    ----------------------------------------------------------------------
366    Name: Allen Leung (leunga (at) reservoir (dot) com)
367    Date: 2004/12/17 16:12:00 EST
368    Tag: leunga-20041217-cygwin-smlnj-home
369    Description:
370    
371        The Unix I/O library of SML/NJ on cygwin does not understand
372    Windows style pathname, so problems arise when SMLNJ_HOME is set
373    to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
374    to a POSIX pathname on cygwin.
375    
376    ----------------------------------------------------------------------
377    Name: Matthias Blume (blume (at) tti - c (dot) org)
378    Date: 2004/12/16 13:35:00 CST
379    Tag: Release_110_52
380    Description:
381    
382    Last-minute changes incorporated into 110.52.  Release tag moved.
383    
384    The changes:
385       - HashString.hashString' -> HashString.hashSubstring
386       - bug fix in UnivariateStats
387    
388    ----------------------------------------------------------------------
389    Name: Matthias Blume (blume (at) tti - c (dot) org)
390    Date: 2004/12/15 23:40:00 CST
391    Tag: blume-20041215-hashSubstring
392    Description:
393    
394    - HashString.hashString' -> HashString.hashSubstring
395    - corresponding changes in atom.sml
396    - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
397      added some comments
398    
399    ----------------------------------------------------------------------
400    Name: Matthias Blume (blume (at) tti - c (dot) org)
401    Date: 2004/12/15 15:30:00 CST
402    Tag: (Release_110_52)
403    Description:
404    
405    New working version (110.52).  NEW BOOTFILES!
406    
407    ----------------------------------------------------------------------
408    Name: Matthias Blume (blume (at) tti - c (dot) org)
409    Date: 2004/12/15 12:45:00 CST
410    Tag: blume-20041215-spaces
411    Description:
412    
413    More on the space problem (this time for Win32).
414    
415    ----------------------------------------------------------------------
416    Name: Matthias Blume (blume (at) tti - c (dot) org)
417    Date: 2004/12/14 17:30:00 CST
418    Tag: blume-20041214-spaces
419    Description:
420    
421    Hacked some of the scripts (in particular: the installer) to cope with
422    spaces in filenames a bit better.  But beware: the current "solution"
423    is likely still full of bugs and inherently incomplete.  (We need to
424    do away with those shell scripts for a comprehensive solution.)
425    
426    ----------------------------------------------------------------------
427    Name: Matthias Blume (blume (at) tti - c (dot) org)
428    Date: 2004/12/13 14:45:00 CST
429    Tag: blume-20041213-ml-makedepend
430    Description:
431    
432    Fixed bug in code for ml-makedepend.
433    
434    ----------------------------------------------------------------------
435    Name: Matthias Blume (blume (at) tti - c (dot) org)
436    Date: 2004/12/09 16:30:00 CST
437    Tag: blume-20041209-statistics
438    Description:
439    
440    Added two simple but potentially useful statistics modules to SML/NJ Library.
441    (See CHANGES file there.)
442    
443    ----------------------------------------------------------------------
444    Name: Matthias Blume (blume (at) tti - c (dot) org)
445    Date: 2004/12/01 16:50:00 CST
446    Tag: blume-20041201-atom
447    Description:
448    
449    smlnj-lib:
450    
451      Added function HashString.hashString' for substrings.
452      Hand-inlined CharVector.foldl into HashString (for speed).
453      Modified implementation of structure Atom to avoid extracting
454      strings from substrings unless necessary.
455      (Also see CHANGES file for smlnj-lib.)
456    
457    ----------------------------------------------------------------------
458    Name: Matthias Blume (blume (at) tti - c (dot) org)
459    Date: 2004/11/24 22:15:00 CST
460    Tag: blume-20041124-cml
461    Description:
462    
463    Made sure CML compiles when Position = Int64.
464    
465    ----------------------------------------------------------------------
466    Name: Matthias Blume (blume (at) tti - c (dot) org)
467    Date: 2004/11/24 14:30:00 CST
468    Tag: blume-20041124-position
469    Description:
470    
471    The compiler can now be compiled in a mode that makes structure Position
472    equal to Int64.  The default, however, is unchanged (Position = Int31)
473    for the time being.
474    
475    To enable 64-bit positions, use the following procedure:
476    
477    1. Start sml
478    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
479    3. Type
480          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
481    4. Run CMB.make() as usual.
482    
483    This is barely tested.  The only test so far was a little SML program
484    counting the number of characters in an 8-gigabyte file by
485    reading it character-by-character.  That test was successful.
486    
487    In support of 64-bit positions, a number of new functions have been
488    added to the runtime system.
489    
490    ----------------------------------------------------------------------
491    Name: Matthias Blume (blume (at) tti - c (dot) org)
492    Date: 2004/11/23 14:45:00 CST
493    Tag: blume-20041123-useFile
494    Description:
495    
496    Fixed a problem with unhelpful error messages related to problems with
497    .cm- or .sml files that appear as part of the sml command line.
498    
499    ----------------------------------------------------------------------
500    Name: Matthias Blume (blume (at) tti - c (dot) org)
501    Date: 2004/11/18 15:40:00 CST
502    Tag: Release_110_51
503    Description:
504    
505    New working version (110.51).  NEW BOOTFILES!
506    
507    ----------------------------------------------------------------------
508    Name: Matthias Blume (blume (at) tti - c (dot) org)
509    Date: 2004/11/18 15:35:00 CST
510    Tag: <none>
511    Description:
512    
513    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
514    Joemann).
515    
516    ----------------------------------------------------------------------
517    Name: Allen Leung (leunga (at) reservoir (dot) com)
518    Date: 2004/11/17 16:05:21 EST 2004
519    Tag: leunga-20041117-mlrisc-live-kill
520    Description:
521    
522       Added support for MLTree constructs LIVE and KILL
523    to all the architectures.
524    
525    ----------------------------------------------------------------------
526    Name: Matthias Blume (blume (at) tti - c (dot) org)
527    Date: 2004/11/13 00:20:00 CST
528    Tag: blume-20041113-versiontool
529    Description:
530    
531    - Stripped down the versiontool: It now only handles the version number.
532      The date string is generated at bootstrap time (during makeml).
533    
534    - In a previous commit, fixed a minor issue with how polyequal is being
535      translated.  In particular, the code now "looks through" abstractions.
536      This results in slightly fewer polyEqual warnings and hopefully slightly
537      more efficient code.  Important examples for where this matters are
538      the new int64 and word64 types.
539    
540    ----------------------------------------------------------------------
541    Name: Matthias Blume (blume (at) tti - c (dot) org)
542    Date: 2004/11/12 00:30:00 CST
543    Tag: blume-20041112-int64
544    Description:
545    
546    Structure Int64 fully hooked in.  (The implementation is not very
547    efficient, though.)
548    
549    ----------------------------------------------------------------------
550    Name: Matthias Blume (blume (at) tti - c (dot) org)
551    Date: 2004/11/11 17:30:00 CST
552    Tag: blume-20041111-more64
553    Description:
554    
555    All the pieces of Word64 are now there, with the exception of the
556    conversions from and to LargeWord. (Eventually these need to be identities,
557    but for the time being they don't even make sense because LargeWord is
558    32-bit wide.)
559    
560    Also started to add similar support for Int64, but major pieces of that
561    are still missing.
562    
563    ----------------------------------------------------------------------
564    Name: Matthias Blume (blume (at) tti - c (dot) org)
565    Date: 2004/11/11 00:15:00 CST
566    Tag: blume-20041111-word64
567    Description:
568    
569    Structure Word64 is now (almost) complete, word literals and patterns
570    seem to work.  There are a few odd pieces missing.  In particular,
571    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
572    at the moment.
573    
574    Making Word64 official would mean that LargeWord becomes Word64.  But
575    this requires extreme care because most word-word conversions have to
576    go through LargeWord, so making a mistake means loss of efficiency or
577    worse.  Eventually there will be a solution similar to (but actually
578    simpler than) what I did with IntInf.
579    
580    ----------------------------------------------------------------------
581    Name: Matthias Blume (blume (at) tti - c (dot) org)
582    Date: 2004/11/10 18:12:00 CST
583    Tag: blume-20041110-64bit
584    Description:
585    
586    More 64-bit hacking (but still not even half-way there yet).
587    Also, some assorted improvements to the handling of 8-bit words.
588    
589    ----------------------------------------------------------------------
590    Name: Matthias Blume (blume (at) tti - c (dot) org)
591    Date: 2004/11/09 17:50:00 CST
592    Tag: <none>
593    Description:
594    
595    Started adding some infrastructure for supporting 64-bit int- and
596    word-types.  (Still in its very early stages.)
597    
598    ----------------------------------------------------------------------
599    Name: Matthias Blume (blume (at) tti - c (dot) org)
600    Date: 2004/10/28 10:45:00 CDT
601    Tag: Release_110_50
602    Description:
603    
604    New working version (110.50).  NEW BOOTFILES!
605    
606                 =====================
607    
608    Also:
609    
610      - Changend config/srcarchiveurl from a file just
611        containing the URL string into a file containing
612        shell script code.  The code has access to the $VERSION variable.
613      - Made corresponding changes to config/install.sh and config/unpack.
614      - Default contents of config/srcarchiveurl uses $VERSION and
615        normally does not have to be edited to reflect a version change.
616    
617      (As a result, a version change can be done by just editing
618       config/version, the rest is now automatic.)
619    
620    ----------------------------------------------------------------------
621    Name: Matthias Blume (blume (at) tti - c (dot) org)
622    Date: 2004/10/27 17:50:00 CDT
623    Tag: blume-20041027-btrace-msg
624    Description:
625    
626    BackTrace.monitor now also reports the source of the exception that
627    triggered the trace.
628    
629    ----------------------------------------------------------------------
630    Name: Matthias Blume (blume (at) tti - c (dot) org)
631    Date: 2004/10/27 17:20:00 CDT
632    Tag: blume-20041027-x86-c-calls
633    Description:
634    
635    This is the HISTORY entry for two earlier commits, both concerning
636    the x86 c-calls code in MLRISC:
637    
638      - added a missing LOAD in the code that deals with struct arguments
639      - made sure the caller does not add the wrong number of bytes to the
640        stack pointer after a call of a function returning a struct
641        (the callee already pops the implicit argument which points to
642         the space reserved for the result)
643    
644    ----------------------------------------------------------------------
645    Name: Allen Leung (leunga (at) reservoir (dot) com)
646    Date: 2004/10/24 14:00:00 EST
647    Tag: leunga-20041024-x86-gas-fucomip
648    Description:
649    
650       John discovered a bug in the syntax of fucomip.
651       The opcodes FU?COMIP? have been changed to
652    
653           fu?comip? %st(i), %st
654    
655    ----------------------------------------------------------------------
656    Name: Matthias Blume (blume (at) tti - c (dot) org)
657    Date: 2004/10/20 15:06:00 CDT
658    Tag: blume-20041020-standalone-backtrace
659    Description:
660    
661    Added a mechanism for getting back-trace information from standalone
662    programs.  Here is how it works:
663    
664       1. The part of the program from which you want to get backtrace
665          information (usually the whole program) should be wrapped with
666          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
667          main program could be modified from something like
668    
669             fun main (pgm, args) = ...
670    
671         to
672    
673            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
674    
675       2. To be able to access BackTrace.monitor, you have to add
676          library $smlnj-tdp/plugins.cm to the .cm file that contains your
677          main function.
678    
679       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
680          CM might have created in the past for your project).
681    
682       4. Build the system using this command line:
683    
684             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
685                      myprog.cm MyProg.main myprog
686    
687          instead of the usual
688    
689             ml-build myprog.cm MyProg.main myprog
690    
691    I changed a library name:
692    
693       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
694    
695    New libraries:
696    
697       $smlnj-tdp/back-trace.cm
698          -- when loaded causes the back-trace plugin to be installed
699       $smlnj-tdp/coverage.cm
700          -- when loaded causes the coverage plugin to be installed
701    
702    ----------------------------------------------------------------------
703    Name: Matthias Blume (blume (at) tti - c (dot) org)
704    Date: 2004/10/18 16:45:00 CDT
705    Tag: blume-20041018-groupowner
706    Description:
707    
708    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
709    
710    Eliminated group owner specs from .cm files throughout the source tree.
711    
712    ----------------------------------------------------------------------
713    Name: Matthias Blume (blume (at) tti - c (dot) org)
714    Date: 2004/10/15 15:45:00 CDT
715    Tag: blume-20041015-coverage
716    Description:
717    
718    * Test coverage tool added!
719    * Further reorganization of tracing-, debugging-, and profiling support:
720    
721        - moved original BTImp -- now called BackTrace -- into a separate
722          library called $/trace-debug-profile.cm
723        - eliminated all mentions of BTrace from SMLofNJ.Internals
724        - only the instrumentation mechanism is now left in the compiler proper
725        - BackTrace module is a plugin which is NOT plugged in by default
726        - Coverage module is another such plugin
727    
728    To get the benefits of any of these plugin modules, the code in
729    question must be compiled with tdp instrumentation turned on.  This
730    can be done by setting SMLofNJ.Internals.TDP.mode to true.
731    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
732    
733    Plugins are selected at link time.  (Pre-compiled instrumented code
734    can be re-loaded with different plugins in effect.)  When an
735    instrumented module is linked, whatever plugins are at that time
736    enabled will come into effect for that module.
737    
738    To enable the back-trace plugin, load library $/trace-debug-profile.cm
739    and invoke BackTrace.install() (e.g., from the interactive prompt).
740    To enable the coverage plugin, load the same library and invoke
741    Coverage.install().
742    
743    Back-traces are generated automatically on uncaught exceptions and
744    when the code in question explicitly invokes BackTrace.trigger().
745    
746    Coverage (and execution frequency-) information must be queried
747    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
748    
749    ----------------------------------------------------------------------
750    Name: Matthias Blume (blume (at) tti - c (dot) org)
751    Date: 2004/10/14 17:40:00 CDT
752    Tag: blume-20041014-tdp-core
753    Description:
754    
755    Snapshot of a significant overhaul of how the trace/debug/profile support
756    is hooked into the system (specifically: Core and SMLofNJ.Internals).
757    
758    ----------------------------------------------------------------------
759    Name: Matthias Blume (blume (at) tti - c (dot) org)
760    Date: 2004/10/13 16:34:00 CDT
761    Tag: blume-20041013-tdp
762    Description:
763    
764    Some rationalization of names:
765    
766       structure BTrace -> structure TDPInstrument
767       etc.
768    
769    This is is preparation of using the original back-trace
770    instrumentation for other purposes.  "TDP" stands for
771    Trace/Debug/Profile.
772    
773    The control flag controlling whether instrumentation is on or off is now
774    registered under a different name, so instead of running sml as
775    
776     sml -Cinstrument.btrace-mode=true
777    
778    one has to say
779    
780     sml -Ctdp.instrument=true
781    
782    ----------------------------------------------------------------------
783    Name: Matthias Blume (blume (at) tti - c (dot) org)
784    Date: 2004/10/11 16:37:00 CDT
785    Tag: blume-20041011-regions
786    Description:
787    
788    Made some minor modifications to elabcore.sml to have source regions
789    be propagated more tightly -- resulting in better (i.e., smaller)
790    regions being reported in error- and debug messages.
791    
792    ----------------------------------------------------------------------
793    Name: Matthias Blume (blume (at) tti - c (dot) org)
794    Date: 2004/10/08 22:50:00 CDT
795    Tag: blume-20041008-cmkw
796    Description:
797    
798    Fixed handling of keywords in .cm files: After seeing "is" the lexer
799    treats subsequent occurrences of "group", "library", "source", "is",
800    "*", and "-" as ordinary identifiers rather than keywords.
801    
802    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
803    is supposed to accept a tool argument called "source", but this did
804    not work because of the clash with the keyword.
805    
806    ----------------------------------------------------------------------
807    Name: Matthias Blume (blume (at) tti - c (dot) org)
808    Date: 2004/10/07 16:00:00 CDT
809    Tag: blume-20041007-cleanup
810    Description:
811    
812    Assorted cleanup work:
813    
814      - got rid of intstrmap in favor of using the library's
815        hash table implementation
816      - threw out most of the pathnames stuff, as it was not used anyway
817      - simplified tokentable implementation
818      - fixed some minor spelling errors
819    
820    ----------------------------------------------------------------------
821    Name: Matthias Blume (blume (at) tti - c (dot) org)
822    Date: 2004/10/06 15:15:15 CDT
823    Tag: blume-20041006-handler
824    Description:
825    
826    Cleaned up the absyn to reflect the invariant that HANDLE always
827    carries a FNexp as part of the type definition.  This eliminates some
828    superfluous sanity checks at runtime down the road.
829    
830    Some minor cleanup of the btrace code.
831    
832    ----------------------------------------------------------------------
833    Name: Matthias Blume (blume (at) tti - c (dot) org)
834    Date: 2004/10/01 10:20:30 CDT
835    Tag: blume-20041001-slave
836    Description:
837    
838    Added hack to make slave mode work in the presence of the version
839    tool.  (Still, since the master does two passes over the code for
840    CMB.make, the release number gets bumped twice when slaves are
841    attached.  I don't know if this is worth fixing...)
842    
843    ----------------------------------------------------------------------
844    Name: Matthias Blume (blume (at) tti - c (dot) org)
845    Date: 2004/09/30 10:55:00 CDT
846    Tag: blume-20040930-version
847    Description:
848    
849    * Moved the "version" magic into its own little library under
850      src/system/smlnj/internal.  This avoids expensive reconstruction of
851      a stable src/compiler/core.cm.
852    
853    * At the same time, structure CompilerVersion is now known as
854      structure SMLNJVersion.
855    
856    * Arranged for the version tool to NOT kick in when rebuilding the system
857      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
858      Also, loading the versiontool does not work when rebuilding the system
859      because CM is not properly initialized at that time.
860    
861    ----------------------------------------------------------------------
862    Name: Matthias Blume (blume (at) tti - c (dot) org)
863    Date: 2004/09/29 14:00:00 CDT
864    Tag: blume-20040929-autoversion
865    Description:
866    
867    Implemented some CM magic to have
868        file src/compiler/TopLevel/main/version.sml
869    generated automagically.
870    The version is taken from two files: config/version and config/release.
871    The first is expected to contain a two-part version number such as 110.49.
872    The second should contain a single number, but it may be missing.
873    
874    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
875    time the version tool is loaded (which is the first time you say CMB.make),
876    then the tool will increment the value stored in config/release every
877    time CMB.make is invoked.
878    
879    The binfile format is now insensitive to anything beyond the first
880    two components of a version number, so bumping the release does not render
881    binfiles incompatible.  Auto-bumping can be used to keep track of versions
882    during development without invalidating existing binfiles.
883    
884    In any case, every CMB.make updates the date information in version.sml.
885    (This is the date that is printed in the banner.)
886    
887    ----------------------------------------------------------------------
888    Name: Matthias Blume (blume (at) tti - c (dot) org)
889    Date: 2004/09/28 10:53:00 CDT
890    Tag: blume-20040928-controls
891    Description:
892    
893    Some cleanup of the controls code.
894    
895    ----------------------------------------------------------------------
896    Name: Matthias Blume (blume (at) tti - c (dot) org)
897    Date: 2004/09/27 22:08:00 CDT
898    Tag: blume-20040927-controls
899    Description:
900    
901    Added two pieces of functionality to the Controls interface:
902    
903    1.   val save'restore: 'a control -> unit -> unit
904    
905       grabs the current value of the control in stage 1 and restores it
906       in stage 2.
907    
908    2.   val set' : 'a control * 'a -> unit -> unit
909    
910       stores the given value into the control in stage 2 (i.e., delayed)
911       but does all error checking in stage 1.
912       (This is for string controls that need to do parse their argument --
913        something that might fail.  In some cases, notably in CM, one
914        already knows the intended argument but wants to delay the actual
915        assignment until a time when error recovery would be more difficult.)
916    
917    Changed the handling of controls in tool arguments to classes "sml" and
918    "lazysml":
919       - use Controls.save'restore as a more robust way of restoring the
920         old value (in particular: without having to re-parse the string)
921       - use controls to handle the "overload" keyword in the init group
922         (I believe this change actually fixes a long-standing obscure bug.)
923    
924    ----------------------------------------------------------------------
925    Name: Matthias Blume (blume (at) tti - c (dot) org)
926    Date: 2004/09/27 17:00:00 CDT
927    Tag: blume-20040927-lazysml
928    Description:
929    
930    Added a new tool class called "lazysml" to CM's tool chest.  The only
931    difference to "sml" is that compilation is done with Control.lazysml
932    set to true.  A source of class "lazysml" is automatically recognized
933    by a file name suffix of ".lml".
934    
935    In addition to the above feature, the original class "sml" now also
936    supports a tool argument "lazy" which has the same effect.  As a
937    result, the following three lines are equivalent:
938    
939        foo.sml : lazysml
940        foo.sml : sml (lazy)
941        foo.sml (lazy)
942    
943    The setting goes into effect both during parsing and during
944    compilation.  The original setting is restored right after parsing and
945    after compilation, respectively.
946    
947    In addition to all the above, there is also a general mechanism to set
948    ANY of the "controls" that are available at the command line via
949    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
950    apply as well. (In fact, "lazy" is implemented as a special case of
951    the general mechanism.)
952    
953    The .cm file syntax uses a new keyword tool argument called "with".
954    There are several ways of indicating the desired settings:
955    
956        foo.sml (with:parser.quotations=true)
957        foo.sml (with:(name:parser.quotations value:true))
958        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
959        foo.sml (with:(name1=value1 name2=value2 ...))
960        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
961    
962        etc.
963    
964    Another possible abbreviation is to leave out the =v or value:v part
965    if the name refers to a boolean control (in which case the value is
966    taken to be true).  Thus, one could get lazy sml also by saying:
967    
968        foo.sml (with:parser.lazy-keyword=true)
969        foo.sml (with:parser.lazy-keyword)
970        foo.sml (with:(name:parser.lazy-keyword value:true))
971        foo.sml (with:(name:parser.lazy-keyword))
972    
973    ----------------------------------------------------------------------
974    Name: Matthias Blume (blume (at) tti - c (dot) org)
975    Date: 2004/09/24 16:22:00 CDT
976    Tag: blume-20040924-ppc-long-branch
977    Description:
978    
979    Turned message about "emiting long form of branch" off by default.
980    Added a control flag to turn it back on when desired.
981    
982    ----------------------------------------------------------------------
983    Name: Matthias Blume (blume (at) tti - c (dot) org)
984    Date: 2004/09/24 16:05:00 CDT
985    Tag: blume-20040924-rounding
986    Description:
987    
988    Applied patch for setting rounding modes under Mac OS X.  Thanks to
989    Melissa O'Neill for providing the code!
990    
991    ----------------------------------------------------------------------
992    Name: Matthias Blume (blume (at) tti - c (dot) org)
993    Date: 2004/09/23 17:30:00 CDT
994    Tag: blume-20040923-envvars
995    Description:
996    
997    1. Changed definition of type ControlRegistry.registry_tree to
998       include control_info (i.e., the name of the controlling environment
999       variable).
1000    
1001    2. Added command-line flags -e and -E to print the names of environment
1002       variables that can be used to control internal settings.
1003       (This uses the new API mentioned in 1.)
1004    
1005    ----------------------------------------------------------------------
1006    Name: Matthias Blume (blume (at) tti - c (dot) org)
1007    Date: 2004/09/13 16:50:00 CDT
1008    Tag: Release_110_49
1009    Description:
1010    
1011    New working version (110.49).  NEW BOOTFILES!
1012    
1013    ----------------------------------------------------------------------
1014    Name: Matthias Blume
1015    Date: 2004/09/13 16:20:00 CDT
1016    Tag: blume-20040913-config-mlrisc
1017    Description:
1018    
1019    Put target "mlrisc" back into the default list.
1020    (There is no harm in having it, and some users have expressed their
1021    wish to have "mlrisc" included by default.)
1022    
1023    ----------------------------------------------------------------------
1024    Name: John Reppy
1025    Date: 2004/09/13
1026    Tag: jhr-20040913-signals
1027    Description:
1028    
1029    Fixed the signal masking code to properly nest mask/unmask operations
1030    on a per-signal basis.
1031    
1032    ----------------------------------------------------------------------
1033    Name: Matthias Blume (blume (at) tti - c (dot) org)
1034    Date: 2004/09/08 13:20:00 CDT
1035    Tag: blume-20040908-heap-magic
1036    Description:
1037    
1038    Bumped the heap macig number to 0x09082004 to account for the changed
1039    layout of the ML frame under MacOS X.
1040    
1041    ----------------------------------------------------------------------
1042    Name: Allen Leung (leunga (at) reservoir (dot) com)
1043    Date: 2004/09/03 11:26:00 EST
1044    Tag: leunga-20040903-cygwin-install
1045    Description:
1046    
1047       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
1048    The Cygwin runtime is turned on by setting the environment
1049    variable SMLNJ_CYGWIN_RUNTIME to 1.
1050    
1051    ----------------------------------------------------------------------
1052    Name: Matthias Blume (blume (at) tti - c (dot) org)
1053    Date: 2004/08/31 17:15:00 CDT
1054    Tag: blume-20040831-core
1055    Description:
1056    
1057    Added some exports to src/compiler/core.cm upon request by J. Joemann.
1058    
1059    ----------------------------------------------------------------------
1060    Name: Matthias Blume (blume (at) tti - c (dot) org)
1061    Date: 2004/08/30 17:55:00 CDT
1062    Tag: blume-20040830-installer
1063    Description:
1064    
1065    Upon request by Johannes Joemann:
1066    
1067    - improved ML code of installer to fall back to coping when renaming
1068      fails (i.e., when source and target are on different file systems);
1069      the code compiles but has yet to be tested in anger
1070    - removed mlrisc from list of default targets (config/targets)
1071    
1072    ----------------------------------------------------------------------
1073    Name: Matthias Blume (blume (at) tti - c (dot) org)
1074    Date: 2004/08/27 17:20:00 CDT
1075    Tag: blume-20040827-ptreql
1076    Description:
1077    
1078    Added ptreql primop to structure InlineT (upon request from Larry
1079    Paulson).
1080    
1081    ----------------------------------------------------------------------
1082    Name: Allen Leung (leunga (at) reservoir (dot) com)
1083    Date: 2004/08/15 21:21:00 EST
1084    Tag: leunga-110_48-udgraph
1085    Description:
1086    
1087    Another bug fix from Carl Hauser:
1088    
1089    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
1090    > 48c48
1091    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
1092    > e::L)
1093    > ---
1094    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
1095    > Without this, any deletion of an edge in an undirected graph does severe
1096    > violence to the graph.
1097    
1098    ----------------------------------------------------------------------
1099    Name: Allen Leung (leunga (at) reservoir (dot) com)
1100    Date: 2004/08/10 23:35:00 EST
1101    Tag: leunga-110_48-ppc
1102    Description:
1103    
1104       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
1105    
1106    ----------------------------------------------------------------------
1107    Name: Matthias Blume (blume (at) tti - c (dot) org)
1108    Date: 2004/08/10 12:00:00 CDT
1109    Tag: Release_110_48
1110    Description:
1111    
1112    New working version (110.48).  NEW BOOTFILES!
1113    
1114    ----------------------------------------------------------------------
1115    Name: Allen Leung (leunga (at) reservoir (dot) com)
1116    Date: 2004/08/09 12:21:00 EST
1117    Tag: leunga-110_47-dijsktra
1118    Description:
1119    
1120    Bug fix from Carl Hauser:
1121    
1122    single_source_shortest_paths in dijkstra.sml was observed to get wrong
1123    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1124    
1125    The problem is that following the expression A.update(dist,s,Num.zero)
1126    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1127    
1128    ----------------------------------------------------------------------
1129    Name: Matthias Blume (blume (at) tti - c (dot) org)
1130    Date: 2004/08/06 18:10:00 CDT
1131    Tag: blume-20040806-cmdline
1132    Description:
1133    
1134    Fiddled with handling of command-line options:
1135    
1136      * sml now quits after processing the command line
1137        if -H, -S, -h<n>, or -s<n> appears as the last
1138        command-line argument
1139      * a new option -q terminates the session when encountered on
1140        the command line; subsequent arguments will be ignored
1141      * bug fixes: short (erroneous) arguments are no longer ignored
1142        completely
1143    
1144    ----------------------------------------------------------------------
1145    Name: Allen Leung (leunga (at) reservoir (dot) com)
1146    Date: 2004/08/04 18:17:00 EST
1147    Tag: leunga-110_47-ppc-ibm-asm
1148    Description:
1149    
1150      - Added minimal IBM assembly syntax support for PowerPC.
1151    
1152      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1153        been altered in the runtime.  We need an automatic way to keep the file
1154        in sync.
1155    
1156    ----------------------------------------------------------------------
1157    Name: Matthias Blume (blume (at) tti - c (dot) org)
1158    Date: 2004/08/04 14:00:00 CDT
1159    Tag: Release_110_47
1160    Description:
1161    
1162    New working version (110.47).  NEW BOOTFILES!
1163    
1164    ----------------------------------------------------------------------
1165    Name: Matthias Blume (blume (at) tti - c (dot) org)
1166    Date: 2004/08/03 14:25:00 CDT
1167    Tag: blume-20040803-callingconv
1168    Description:
1169    
1170    Added low-level support for choosing C calling conventions by
1171    twiddling the type of rawccall.  (See
1172    src/compiler/Semant/types/cproto.sml for details.)
1173    
1174    ----------------------------------------------------------------------
1175    Name: Matthias Blume (blume (at) tti - c (dot) org)
1176    Date: 2004/08/02 15:55:00 CDT
1177    Tag: blume-20040802-backout
1178    Description:
1179    
1180    Backed out of change to win32-filesys.c.  The earlier patch to
1181    get_file_time caused CM to produce files with the wrong time stamp.
1182    
1183    ----------------------------------------------------------------------
1184    Name: Matthias Blume (blume (at) tti - c (dot) org)
1185    Date: 2004/08/02 14:45:00 CDT
1186    Tag: blume-20040802-nlffi-win32
1187    Description:
1188    
1189    Added NLFFI support for Win32, adapted from a patch provided by David
1190    Hansel.  This is currently completely untested.  Also, the issue
1191    concerning stdcall vs. ccall is still unresolved.
1192    
1193    ----------------------------------------------------------------------
1194    Name: Matthias Blume (blume (at) tti - c (dot) org)
1195    Date: 2004/07/30 17:55:00 CDT
1196    Tag: blume-20040730-various
1197    Description:
1198    
1199    Gearing up towards 110.47...
1200    
1201    - various minor bugfixes to ml-nlffigen
1202    - a beginning of a manual for nlffi
1203    
1204    - eliminated 'export name=value' in config/install.sh as this does
1205      not work with certain versions of /bin/sh
1206      (Thanks to David King at Motorola for catching this.)
1207    
1208    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1209      - added a test for tm==NULL to gmtime.c and localtime.c
1210      - applied patch for incorrect GetFileTime under win32
1211      - toSeconds -> toMilliseconds in Win32/win32-process.sml
1212    
1213    ----------------------------------------------------------------------
1214    Name: Matthias Blume (blume (at) tti - c (dot) org)
1215    Date: 2004/07/21 18:20:00 CDT
1216    Tag: blume-20040721-nlffigen
1217    Description:
1218    
1219    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1220      for a typedef to an incomplete type, but leave out the "typ" member.
1221      (This is just for consistency.)
1222    - Started to produce what is supposed to become better (i.e., comprehensive)
1223      documentation of what ml-nlffigen does and produces.
1224    
1225    ----------------------------------------------------------------------
1226    Name: Matthias Blume (blume (at) tti - c (dot) org)
1227    Date: 2004/07/14 16:25:00 CDT
1228    Tag: blume-20040714-union
1229    Description:
1230    
1231    Added C_UNION to c-calls/c-types.sml and updated the machinery
1232    (ml-nlffigen, cproto.sml) that conveys C function interface
1233    information to the code generator.
1234    
1235    However, the actual architecture-specific implementation of function
1236    arguments and results that are C unions is still not implemented.
1237    
1238    ----------------------------------------------------------------------
1239    Name: Allen Leung (leunga (at) reservoir (dot) com)
1240    Date: 2004/07/14 14:38:00 EST
1241    Tag: leunga-110_46_1-ppc-lwzu
1242    Description:
1243    
1244        Added these instructions to the PowerPC architecture:
1245        LBZU(X), LHZU(X), LWZU(X),
1246        STWU(X), STFDU, STFSU
1247    
1248        etc...
1249    
1250        Note: I haven't added their instruction encoding into the description.
1251    
1252    ----------------------------------------------------------------------
1253    Name: Allen Leung (leunga (at) reservoir (dot) com)
1254    Date: 2004/07/13 15:04:00 EST
1255    Tag: leunga-110_46_1-ppc-lwarx
1256    Description:
1257    
1258       Added the two instructions LWARX and STWCX to the PowerPC
1259    instruction set.
1260    
1261       A (untested) rewrite of loop-structure.sml.  The old version
1262    is completely broken.
1263    
1264    ----------------------------------------------------------------------
1265    Name: Matthias Blume (blume (at) tti - c (dot) org)
1266    Date: 2004/07/13 13:50:00 CDT
1267    Tag: blume-20040713-nlffi
1268    Description:
1269    
1270    - use paramAlloc to report c-calls with too many arguments
1271      (for PPC version where parameter area is pre-allocated)
1272    - added ccall_maxargspace to machspec (to implement the above)
1273    - made "make" commend in CM's "make" tool configurable
1274    - added option (default: on) for passing the name of the SML/NJ's "bin"
1275      directory to "make"; the call looks like this:
1276    
1277         make <options> SMLNJ_BINDIR=<dir> <target>
1278    
1279      This can be used by the Makefile to, e.g., pick the "right" version
1280      of ml-nlffigen.
1281    - minor code tweaks
1282    
1283    ----------------------------------------------------------------------
1284    Name: Matthias Blume (blume (at) tti - c (dot) org)
1285    Date: 2004/07/12 22:50:00 CDT
1286    Tag: blume-110_46_1-macosx-nlffi
1287    Description:
1288    
1289    NLFFI under Mac OS X now working (sort of).  This is largely untested,
1290    though.
1291    
1292    Note:  1. You have to make a new, clean build of the runtime system.
1293           2. There are new BOOTFILES, you have to use them!
1294              (Doing the bootstrap process yourself would be *very* painful!
1295               If you absolutely have to do it, build the system under
1296               a different architecture and then cross-compile.)
1297    
1298    Version bumped to 110.46.1 to account for runtime data format changes.
1299    
1300    ----------------------------------------------------------------------
1301    Name: Matthias Blume (blume (at) tti - c (dot) org)
1302    Date: 2004/06/18 14:30:00 CDT
1303    Tag: blume-20040618-unix
1304    Description:
1305    
1306    Changed the implementation of structure Unix so that the same stream
1307    is returned every time one of the {text,bin}{In,Out}streamOf functions
1308    is invoked on the same proc.  This is not what the spec currently
1309    says -- although IMO it arguably should.  (See discussion below.)
1310    
1311    ----------------------------------------------------------------------
1312    Name: Matthias Blume (blume (at) tti - c (dot) org)
1313    Date: 2004/06/17 18:15:00 CDT
1314    Tag: Release_110_46
1315    Description:
1316    
1317    New working version (110.46).  NEW BOOTFILES!
1318    
1319    ----------------------------------------------------------------------
1320    Name: Matthias Blume (blume (at) tti - c (dot) org)
1321    Date: 2004/06/17 17:20:00 CDT
1322    Tag: blume-20040617-timer-unix
1323    Description:
1324    
1325    Changed the interface of structures Timer and Unix to match the most
1326    recent Basis spec.
1327    
1328    In the case of Unix there still seems to be an open/weird issue:
1329    
1330       The {text,bin}{In,Out}streamOf functions are supposed to create
1331       fresh streams whenever they are called -- as opposed to have them
1332       return the same stream every time.  This design is supposed to
1333       prevent space leaks caused by proc values hanging on to streams.
1334    
1335       The reap function, on the other hand, is supposed to close the
1336       streams.  This cannot be done without having a handle on the
1337       stream in proc after all...
1338    
1339    I took the liberty to implement the following stopgap solution:
1340    
1341      The proc value hangs on to the most recently created stream(s).
1342      Reap closes those.  If either or both of the two streams hadn't
1343      been created at all yet, then reap will close the corresponding
1344      file descriptors directly.
1345    
1346    PS: I don't understand the original space leak argument anymore.  If
1347    a proc hangs on to the imperative stream, then I/O operations on those
1348    will advance the state of the cached stream and avoid the space leak.
1349    
1350    ----------------------------------------------------------------------
1351    Name: Matthias Blume (blume (at) tti - c (dot) org)
1352    Date: 2004/05/28 16:45:00 CDT
1353    Tag: blume-20040528-basis
1354    Description:
1355    
1356    Added signature PACK_REAL and exported functor PrimIO.
1357    
1358    ----------------------------------------------------------------------
1359    Name: Matthias Blume (blume (at) tti - c (dot) org)
1360    Date: 2004/05/25 16:00:00 CDT
1361    Tag: blume-20040525-group-owner
1362    Description:
1363    
1364    CM now ignores (but still accepts) the "owner" information in group
1365    descriptions.  The owner of a group is its next enclosing
1366    library. Each group must have a unique owner.  (There is a virtual
1367    "toplevel" library that owns groups which are not nested within a real
1368    library.)  Previously, each group had to explicitly declare its owner,
1369    and CM would check that such a declaration is correct.  The new scheme
1370    is to have CM check that for each group there is precisely one owning
1371    library.
1372    
1373    The advantage of the new scheme is that the programmer no longer needs
1374    to maintain the somewhat annoying owner information.  The downside is
1375    that CM cannot enforce the ownership rule across multiple runs of
1376    CM.make.  Fortunately, enclosing the same group in two different
1377    libraries A and B which are not part of the same program does not
1378    cause real problems.
1379    
1380    ----------------------------------------------------------------------
1381    Name: Matthias Blume (blume (at) tti - c (dot) org)
1382    Date: 2004/05/20 16:00:00 CDT
1383    Tag: blume-20040520-win32
1384    Description:
1385    
1386    Made the win32 version work again.  (Strangely, a misplaced comma had
1387    slipped into win32-process.c which prevented the runtime from being
1388    compiled correctly.)
1389    
1390    Also, included a minor addition to ml-build.bat analogous to what was
1391    done in blume-20040519-ml-build.
1392    
1393    ----------------------------------------------------------------------
1394    Name: Matthias Blume (blume (at) tti - c (dot) org)
1395    Date: 2004/05/19 22:10:00 CDT
1396    Tag: blume-20040519-ml-build
1397    Description:
1398    
1399    Arranged for ml-build to clean up after itself a little bit better.
1400    The script generates a temporary SML source file and compiles it using
1401    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1402    gets rid of those at the end, so they don't accumulate under .cm.
1403    
1404    This required a minor change to install.sh because the name of the
1405    metadata directory (default: .cm) is actually configurable at
1406    installation time.
1407    
1408    ----------------------------------------------------------------------
1409    Name: Matthias Blume (blume (at) tti - c (dot) org)
1410    Date: 2004/05/18 15:50:00 CDT
1411    Tag: blume-20040518-mkreader
1412    Description:
1413    
1414    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1415    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1416    
1417    ----------------------------------------------------------------------
1418    Name: Matthias Blume (blume (at) tti - c (dot) org)
1419    Date: 2004/05/11 14:35:00 CDT
1420    Tag: blume-20040511-win32sock
1421    Description:
1422    
1423    Added previously missing support for many socket-related functions
1424    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1425    for the voluminous patch!
1426    
1427    (I have not tested this patch under win32 yet.)
1428    
1429    Here is David's e-mail:
1430    
1431    Hi,
1432    
1433    Attached to this email you find a diff against sml/nj 110.45
1434    that will enable socket support under Windows.
1435    
1436    To apply the patch (using unix or cygwin)
1437    1) gunzip runtime.diff.gz
1438    2) "cd" into "src/runtime" in the source tree of a fresh
1439       110.45 installation.
1440    3) patch -p 1 < [your/path/to]runtime.diff
1441    
1442    The code compiles fine but has NOT yet been extensively tested.
1443    I only ran a few tests for basic socket client functionality
1444    (which worked fine).  Especially the functions that use ioctl
1445    are not tested at all and might not work (see below).
1446    
1447    I implemented this since we want to move to a newer version of sml/nj
1448    but need socket support in order to use it.  This is the first time I
1449    even had a look at the sml/nj source,  so please review my changes
1450    before making this part of the distribution!  Here are a few issues
1451    that I think might be better for someone to solve who is more
1452    familiar with the sml/nj source (and socket programming):
1453    
1454    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1455      exception since I could not figure out what the windows equivalent
1456      of these functions is
1457    
1458    - In sockets-osdep.h there are a some #include statements that are
1459      only used in a few files that include sockets-osdep.h
1460    
1461    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1462      does not process its return value since I don't know how to
1463      report an error upwards.
1464    
1465    - It would probably be good to have a call to WSACleanup() when
1466      the library is unloaded (if there is such a possibility).
1467      Otherwise I think Windows will take care of this automatically
1468      when the process finishes.
1469    
1470    - I used ioctlsocket() as a replacement for ioctl() but I have
1471      no idea if that is actually the proper replacement on Windows.
1472    
1473    - All these issues are marked in the code by "FIXME" comments.
1474    
1475    We use sml/nj extensively in our products and are quite happy
1476    with it.  I hope this contribution will help you.
1477    
1478    Keep up the good work!
1479    
1480    David
1481    
1482    ----------------------------------------------------------------------
1483    Name: Matthias Blume (blume (at) tti - c (dot) org)
1484    Date: 2004/05/11 14:20:00 CDT
1485    Tag: blume-20040511-installml
1486    Description:
1487    
1488    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1489    the patch.)
1490    
1491    ----------------------------------------------------------------------
1492    Name: Matthias Blume (blume (at) tti - c (dot) org)
1493    Date: 2004/05/11 14:05:00 CDT
1494    Tag: blume-20040511-nlffi-netbsd
1495    Description:
1496    
1497    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1498    the patch.)
1499    
1500    ----------------------------------------------------------------------
1501    Name: Matthias Blume (blume (at) tti - c (dot) org)
1502    Date: 2004/05/11 12:05:00 CDT
1503    Tag: blume-20040511-exports
1504    Description:
1505    
1506    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1507    export lists in compiler-related .cm-files.
1508    
1509    ----------------------------------------------------------------------
1510    Name: Matthias Blume (blume (at) tti - c (dot) org)
1511    Date: 2004/05/11 11:35:00 CDT
1512    Tag: blume-20040511-allsource
1513    Description:
1514    
1515    The installer now honors the "src-smlnj" target again, although its meaning
1516    has changed from "all sources required for the compiler" to "all sources
1517    the installer knows about".  In other words, if you enable "src-smlnj"
1518    in the "targets" file, then the installer will pull in sources for
1519    everything.  (Notice that this refers to source code only.  Compiled
1520    code is still only installed for modules that were requested explicitly
1521    or which are required for other modules that were requested explicitly.)
1522    
1523    ----------------------------------------------------------------------
1524    Name: Matthias Blume (blume (at) tti - c (dot) org)
1525    Date: 2004/04/23 17:40:00 CDT
1526    Tag: blume-20040423-ieee-scan
1527    Description:
1528    
1529    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1530    in the exponent calculation we get INF or ZERO (depending on the mantissa
1531    and the sign of the exponent).
1532    
1533    ----------------------------------------------------------------------
1534    Name: Matthias Blume (blume (at) tti - c (dot) org)
1535    Date: 2004/04/23 10:40:00 CDT
1536    Tag: blume-20040423-ml-build
1537    Description:
1538    
1539    The ml-build script now terminates with a non-0 status when something
1540    goes wrong.
1541    
1542    ----------------------------------------------------------------------
1543    Name: Matthias Blume (blume (at) tti - c (dot) org)
1544    Date: 2004/04/22 16:35:00 CDT
1545    Tag: blume-20040422-Option
1546    Description:
1547    
1548    Made exception Option to be the same as exception Option.Option
1549    (as it should be).
1550    
1551    ----------------------------------------------------------------------
1552    Name: Allen Leung (leunga (at) reservoir (dot) com)
1553    Date: 2004/03/19 14:40:00 EST
1554    Tag: leunga-20040319-cygwin-nlffi
1555    Description:
1556    
1557    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1558    of SML/NJ.  The problem is that
1559    
1560       lib = dlopen(NULL, ...)
1561       f   = dlsym(lib, "malloc");
1562    
1563    does not work on Windows unless we explicitly export symbols
1564    such as 'malloc' during linking.  We fixed this by explicitly
1565    exporting the required symbols with the magic gcc incantation:
1566    
1567        -Wl,--export-all cygwin.def
1568    
1569    where cygwin.def is a file containing all the symbols that
1570    we wish to export.
1571    
1572    I suspect this is a Windows problem and we'll have to
1573    do the same (somehow with windows compilers) when
1574    we build the native win32 version with the system
1575    calls LoadLibrary/GetProcAddress.
1576    
1577    ----------------------------------------------------------------------
1578    Name: Matthias Blume (blume (at) tti - c (dot) org)
1579    Date: 2004/03/04 16:35:00 CST
1580    Tag: blume-20040304-intinf-fmt
1581    Description:
1582    
1583    Fixed problem with IntInf.fmt (sign would show up on the right instead
1584    of on the left for BIN, OCT, and HEX).
1585    
1586    ----------------------------------------------------------------------
1587    Name: Matthias Blume (blume (at) tti - c (dot) org)
1588    Date: 2004/03/04 11:25:00 CST
1589    Tag: blume-20040304-symlinks
1590    Description:
1591    
1592    Fixed problem with installer script (unix only) where bin/ml-yacc and
1593    friends pointed (via symlinks) to absolute locations instead of just
1594    .run-sml.  This was reported by Vesa A Norrman.
1595    
1596    ----------------------------------------------------------------------
1597    Name: Matthias Blume (blume (at) tti - c (dot) org)
1598    Date: 2004/02/13 14:50:00 CST
1599    Tag: Release_110_45
1600    Description:
1601    
1602    New working version (110.45). New bootfiles.
1603    
1604    ----------------------------------------------------------------------
1605    Name: Matthias Blume (blume (at) tti - c (dot) org)
1606    Date: 2004/01/26 15:15:15 CST
1607    Tag: blume-20040126-toplevel
1608    Description:
1609    
1610    Improved handling of exceptions at the interactive toplevel.
1611    
1612    ----------------------------------------------------------------------
1613    Name: Matthias Blume (blume (at) tti - c (dot) org)
1614    Date: 2004/01/26 11:25:00
1615    Tag: blume-20040126-app
1616    Description:
1617    
1618    Type of top-level "app" corrected.
1619    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1620    
1621    ----------------------------------------------------------------------
1622    Name: Matthias Blume (blume (at) tti - c (dot) org)
1623    Date: 2003/11/18 17:10 CST
1624    Tag: blume-20031118-basis-fiddle
1625    Description:
1626    
1627    - changed Timer interface to what might become the spec
1628    - POSIX_FLAGS -> BIT_FLAGS according to spec
1629    - some other minor discrepancies wrt. spec eliminated
1630    
1631    ----------------------------------------------------------------------
1632    Name: Matthias Blume (blume (at) tti - c (dot) org)
1633    Date: 2003/11/06 12:00:00 CST
1634    Tag: Release_110_44
1635    Description:
1636    
1637    New working version (110.44). New bootfiles.
1638    
1639    ----------------------------------------------------------------------
1640    Name: Matthias Blume (blume (at) tti - c (dot) org)
1641    Date: 2003/11/04 11:50:00 CST
1642    Tag: blume-20031104-move-libraries
1643    Description:
1644    
1645    Eliminated the "dont_move_libraries" directive in config/targets.
1646    (The mechanism was broken and could not be fixed easily.  Moreover,
1647    there does not seem to be any reason not to move all libraries into
1648    lib during installation.  I originally implemented this directive as a
1649    backward-compatibility feature when I first introduced the new CM.
1650    Now that things have been stable for a long time and going back to the
1651    old CM is not an option, there is no reason to keep it around.)
1652    
1653    ----------------------------------------------------------------------
1654    Name: Matthias Blume (blume (at) tti - c (dot) org)
1655    Date: 2003/11/03 16:00:00 CST
1656    Tag: blume-20031103-installdir
1657    Description:
1658    
1659    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1660    Richards for pointing out the problem and providing the solution.)
1661    
1662    ----------------------------------------------------------------------
1663    Name: Matthias Blume (blume (at) tti - c (dot) org)
1664    Date: 2003/10/01 17:05:00 CDT
1665    Tag: blume-20031001-lal-mlrisc
1666    Description:
1667    
1668    MLRISC bug fix from Lal.
1669    
1670    ----------------------------------------------------------------------
1671    Name: Matthias Blume (blume (at) tti - c (dot) org)
1672    Date: 2003/09/30 16:10:00 CDT
1673    Tag: blume-20030930-primio-bat
1674    Description:
1675    
1676    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1677    2. Improved .bat files (for Win32 port) to make things work under Win95.
1678       (thanks to Aaron S. Hawley for this one)
1679    
1680    ----------------------------------------------------------------------
1681    Name: Matthias Blume (blume (at) tti - c (dot) org)
1682    Date: 2003/09/26 16:05:00 CDT
1683    Tag: blume-20030926-wrappriv
1684    Description:
1685    
1686    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1687    
1688    ----------------------------------------------------------------------
1689    Name: Matthias Blume (blume (at) tti - c (dot) org)
1690    Date: 2003/09/26 15:00:00 CDT
1691    Tag: blume-20030926-110_43_3
1692    Description:
1693    
1694    - additional cleanup
1695    - version number bump, NEW BOOTFILES
1696    
1697    ----------------------------------------------------------------------
1698    Name: Matthias Blume (blume (at) tti - c (dot) org)
1699    Date: 2003/09/26 12:00:00 CDT
1700    Tag: blume-20030926-ppautoload
1701    Description:
1702    
1703    I modified the read-eval-print loop so that the autoloader gets
1704    invoked whenever the prettyprinter tries to look up a symbol that
1705    is not currently defined in the toplevel environment but which
1706    appears in CM's autoload registry.  As a result, we see far fewer of
1707    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1708    
1709    In addition to this I tried to clean up some pieces of the Basis
1710    implementation (e.g., Socket, Word8Array) in order to prevent other
1711    instances of these ?.Foo.Bar.xxx names from being printed.
1712    
1713    The mechanism that picks names for types still needs some work, though.
1714    (Right now it seems that if there is a type A.t which is defined to
1715    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1716    "?.B.u" although the perhaps more sensible solution would be to use
1717    "A.t" in this case.  In other words, the prettyprinter should follow
1718    a chain of DEFtycs not farther than there are corresponding toplevel
1719    names in the current environment.)
1720    
1721    ----------------------------------------------------------------------
1722    Name: Matthias Blume (blume (at) tti - c (dot) org)
1723    Date: 2003/09/24 16:31:00 CDT
1724    Tag: blume-20030924-installer
1725    Description:
1726    
1727    Another installer tweak:  All the ML code for the installer is now
1728    compiled during CMB.make and put into a little library called
1729    $smlnj/installer.cm.  The installation then simply invokes
1730    
1731       sml -m $smlnj/installer.cm
1732    
1733    and everything happens automagically.
1734    
1735    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1736    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1737            and BINDIR.
1738    
1739    The new scheme guarantees that the ML code responsible for the installation
1740    is in sync with the APIs of the main system.  Also, the installer is
1741    somewhat faster because the installer script is precompiled.
1742    
1743    ----------------------------------------------------------------------
1744    Name: Matthias Blume (blume (at) tti - c (dot) org)
1745    Date: 2003/09/24 15:35:00 CDT
1746    Tag: blume-20030924-synsock
1747    Description:
1748    
1749    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1750    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1751    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1752    implementations of non-blocking operations from CML's Socket
1753    structure.
1754    
1755    ----------------------------------------------------------------------
1756    Name: Matthias Blume (blume (at) tti - c (dot) org)
1757    Date: 2003/09/24 15:10:05 CDT
1758    Tag: blume-20030924-sockets
1759    Description:
1760    
1761    1. Fixed SOCKET API and implementation to match Basis spec.
1762       This required changing the internal representation of sockets to one
1763       that remembers (for each socket file descriptor) whether it is currently
1764       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1765       call is made only if the state actually needs to change).
1766    
1767    2. OS-specific details of sockets were moved into separate files, thus
1768       making it possible to unify the bulk of the socket implementations
1769       between Unix and Win32.
1770    
1771    3. CML's socket API changed accordingly.
1772       (Note that we need to remove non-blocking functions from this API
1773       since they are redundant in the case of CML!)
1774    
1775    4. CML's socket implementation now makes use of non-blocking functions
1776       provided by Basis, thus removing all OS-dependent code from this part
1777       of CML.
1778    
1779    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1780    
1781    ----------------------------------------------------------------------
1782    Name: Matthias Blume (blume (at) tti - c (dot) org)
1783    Date: 2003/09/22 12:10:00 CDT
1784    Tag: blume-20030922-110_43_2
1785    Description:
1786    
1787    Made a new interim version and bootfiles for developer's bootstrapping
1788    convenience.
1789    
1790       110.43.2 -- NEW BOOTFILES
1791    
1792    ----------------------------------------------------------------------
1793    Name: Matthias Blume (blume (at) tti - c (dot) org)
1794    Date: 2003/09/19 15:55:00 CDT
1795    Tag: blume-20030919-cmdir
1796    Description:
1797    
1798    1. new-install.sh -> install.sh
1799    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1800    3. tweaked installer so that another name instead of .cm can be chosen
1801       at install time (by setting the CM_DIR_ARC environment variable
1802       during installation); once installation is complete, the name is
1803       fixed
1804    
1805    ----------------------------------------------------------------------
1806    Name: Matthias Blume (blume (at) tti - c (dot) org)
1807    Date: 2003/09/18 16:00:00 CDT
1808    Tag: blume-20030918-110_43_1
1809    Description:
1810    
1811    Made a new interim version and bootfiles for developer's bootstrapping
1812    convenience.
1813    
1814       110.43.1 -- NEW BOOTFILES
1815    
1816    ----------------------------------------------------------------------
1817    Name: Matthias Blume (blume (at) tti - c (dot) org)
1818    Date: 2003/09/18 15:20:00 CDT
1819    Tag: blume-20030918-misc
1820    Description:
1821    
1822    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
1823       this seems to be controversial at the moment) and used those in
1824       Posix.ProcEnv.times.
1825    
1826    2. Added Time.{from,to}Nanoseconds to Time.
1827    
1828    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
1829       For example, fromLargeInt never needs to look at more than 3 "big
1830       digits" to get its 53 bits of precision.
1831    
1832    ----------------------------------------------------------------------
1833    Name: Matthias Blume (blume (at) tti - c (dot) org)
1834    Date: 2003/09/17 16:30:00 CDT
1835    Tag: blume-20030917-real32-slices
1836    Description:
1837    
1838    Added an entry to the primitive environment
1839    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
1840    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
1841    
1842    Removed some of the "magic" constants in real64.sml and replaced them
1843    with code that generates these values from their corresponding
1844    integer counterparts.
1845    
1846    Made all(?) the slice-related changes to the Basis and made everything
1847    compile again...
1848    
1849    ----------------------------------------------------------------------
1850    Name: Matthias Blume (blume (at) tti - c (dot) org)
1851    Date: 2003/09/15 17:45:00 CDT
1852    Tag: blume-20030915-rbase
1853    Description:
1854    
1855    Fixed bug in Real.fromLargeInt.
1856    
1857    ----------------------------------------------------------------------
1858    Name: Matthias Blume (blume (at) tti - c (dot) org)
1859    Date: 2003/09/13 18:11:00 CDT
1860    Tag: blume-20030913-libinstall
1861    Description:
1862    
1863    Minor bugfix in config/libinstall (set anchor with path to
1864    standalone tool after installing it, otherwise libraries that
1865    need ml-lex or ml-yacc won't compile the first time the installer
1866    runs).
1867    
1868    ----------------------------------------------------------------------
1869    Name: Matthias Blume (blume (at) tti - c (dot) org)
1870    Date: 2003/09/12 11:45:00 CDT
1871    Tag: blume-20030912-various
1872    Description:
1873    
1874     - fixed bug in Real.toLargeInt
1875     - fixed bug in Posix.ProcEnv.times
1876     - changed inputLine functions to return an option
1877     - minor installer improvements / bugfixes
1878     - changed default @SMLalloc parameter for x86/celeron to 64k
1879    
1880    ----------------------------------------------------------------------
1881    Name: Matthias Blume (blume (at) tti - c (dot) org)
1882    Date: 2003/09/09 22:00:00 CDT
1883    Tag: Release_110_43
1884    Description:
1885    
1886        New working release 110.43.  New bootfiles.
1887    
1888    ----------------------------------------------------------------------
1889    Name: Matthias Blume (blume (at) tti - c (dot) org)
1890    Date: 2003/09/09 19:20:00 CDT
1891    Tag: blume-20030909-installer
1892    Description:
1893    
1894    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
1895    Modified config/install.bat to take advantage of it.  Also modified
1896    config/install.sh (and called it config/new-install.sh) to take advantage
1897    of it on Unix systems.  (The SML code is (supposed to be) platform-
1898    independent.)
1899    
1900    The installer can now install everything under Win32
1901    as well as under *nix as long as it compiles.
1902    
1903    Other changes:
1904    
1905      - made CML compile again under Win32
1906      - made eXene compile under Win32 (by providing a fake structure UnixSock
1907        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
1908      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
1909        same as type int;  under Win32 it isn't
1910      - fixed some slice-related problems in the win32-specific parts of CML
1911      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
1912        (under Win32, the volume name is case-insensitive, and the
1913        OS.Path code compares volume names for equality)
1914    
1915    ----------------------------------------------------------------------
1916    Name: Matthias Blume (blume (at) tti - c (dot) org)
1917    Date: 2003/09/08 11:55:00 CDT
1918    Tag: blume-20030908-fullpath
1919    Description:
1920    
1921    Made Win32 version of OS.FileSys.fullPath return current directory
1922    when given an empty string.  This is what the spec says, and incidentally,
1923    CM depends on it.  (CM otherwise goes into an infinite loop in certain
1924    cases when presented with the name of a non-existing .cm file.)
1925    
1926    ----------------------------------------------------------------------
1927    Name: Matthias Blume (blume (at) tti - c (dot) org)
1928    Date: 2003/09/04 16:30:00 CDT
1929    Tag: blume-20030905-slices-etc
1930    Description:
1931    
1932      1. Changed interface to vectors and arrays in Basis to match
1933         (draft) Basis spec.
1934      2. Added signatures and implementations of slices according to
1935         Basis spec.
1936      3. Edited source code throughout the system to make it compile again
1937         under 1. and 2.  (In some cases code had to be added to have it
1938         match the new signatures.)
1939      4. MLRISC should be backward-compatible: the copies of the originals
1940         of files that needed to change under 3. were retained, the .cm files
1941         check the compiler version number and use old versions when
1942         appropriate.
1943      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
1944         dirstream -> string option (in accordance with Basis spec).
1945      6. When generating code that counts lines, ml-lex used function
1946         CharVector.foldli, taking advantage of its old interface.
1947         This has been replaced with the corresponding code from
1948         CharVectorSlice.  (html-lex must be re-lexed!)
1949      7. BitArray in smlnj-lib/Util has been extended/modified to match the
1950         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
1951      8. Removed temporary additions (fromInternal, toInternal) from the
1952         (now obsolete) IntInf in smlnj-lib/Util.
1953      9. Cleaned up structure Byte.
1954     10. Added localOffset, scan, and fromString to Date (according to spec).
1955         Cleaned/corrected implementation of Date.
1956         (Still need to check for correctness; implement better canonicalizeDate.)
1957     11. Added "scan" to signature IEEE_REAL.
1958     12. Some improvements to IntInf [in particular: efficiency-hack for
1959         mod and rem when second operand is 2 (for parity checks).]
1960     13. Changed representation of type Time.time, using a single IntInf.int
1961         value counting microseconds.  This considerably simplified the
1962         implementation of structure Time.  We now support negative time
1963         values; scan and fromString handle signs.
1964     14. Functor PrimIO now takes two additional arguments (VectorSlice and
1965         ArraySlice).
1966    
1967    ----------------------------------------------------------------------
1968    Name: Matthias Blume (blume (at) tti - c (dot) org)
1969    Date: 2003/08/28 17:00:00 CDT
1970    Tag: blume-20030828-intinf
1971    Description:
1972    
1973    This is a major update which comes with a version number bump
1974    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
1975    and an implementation of IntInf in the Basis.
1976    
1977    There are a fairly large number of related changes and updates throughout
1978    the system:
1979    
1980    Basis:
1981            - Implemented IntInf.
1982            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
1983            - Added some missing Real64 operations, most notably Real.toLargeInt.
1984            - Added FixedInt as a synonym for Int32.
1985    
1986    compiler:
1987            * Added support for a built-in intinf type.
1988              - literals
1989              - pattern matching
1990              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
1991              - overloading on literals and operations
1992    
1993            This required adding a primitive type intinf, some additional
1994            primops, and implementations for several non-trivial intinf
1995            operations in Core.  (The intinf type is completely abstract
1996            to the compiler; all operations get delegated back to the Core.)
1997    
1998            * Intinf equality is handled by polyequal.  However, the compiler
1999            does not print its usual warning in this case (since polyequal
2000            is the right thing to do there).
2001    
2002            * Improved the organization of structure InlineT.
2003    
2004            * A word about conversion primops:
2005            If conversions involving intinf do not cancel out during
2006            CPS contract, then the compiler must insert calls to Core functions.
2007            Since all core access must be resolved already during the FLINT
2008            translate phase, it would be too late a the time of CPS contract
2009            to add new Core calls.  For this reason, conversion primops
2010            for intinf carry two arguments:  1. the numeric argument that
2011            they are supposed to convert, and 2. the Core function that
2012            can help with this conversion if necessary.  If CPS contract
2013            eliminates a primop, then the associated Core function becomes
2014            dead and goes away.  Intinf conversion primops that do not get
2015            eliminated by CPS contract get rewritten into calls of their
2016            core functions by a separate, new phase.
2017    
2018    interactive system:
2019            - Control.Print.intinfDepth controls max length of intinf constants
2020            being printed. (Analogous to Control.Print.stringDepth.)
2021            - Cleanup in printutil and pputil: got rid of unused stuff and
2022            duplicates; replaced some of the code with code that makes better
2023            use of library functionality.
2024    
2025    CM:
2026            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
2027            are no longer silent.
2028    
2029    CKIT:
2030            Fixed mismatched uses of Int32 and LargeInt.  I always decided
2031            in favor of LargeInt -- which is now the same as IntInf.
2032            CKIT-knowledgable people should check whether this is what's
2033            intended and otherwise change things back to using Int32 or
2034            FixedInt.
2035    
2036    Throughout the code:
2037            Started using IntInf.int literals and built-in operations
2038            (e.g., comparison with 0) where this seems appropriate.
2039    
2040    
2041    ----------------------------------------------------------------------
2042    Name: Dave MacQueen (dbm@cs.uchicago.edu)
2043    Date: 2003/08/13 11:36:00 CDT
2044    Tag: dbm-20030813-mcz-merge1
2045    Description:
2046    
2047    Merging changes from the mcz-branch development branch into trunk.
2048    These changes involve replacement of the emulated old prettyprinter
2049    interface with direct use of the SML/NJ Lib PP library, and fixing
2050    of a couple of bugs (895, 1186) relating to error messages.  A new
2051    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
2052    has been added.
2053    
2054    ----------------------------------------------------------------------
2055    Name: Matthias Blume (blume (at) tti - c (dot) org)
2056    Date: 2003/08/11 15:45:00 CDT
2057    Tag: blume-20030811-windows
2058    Description:
2059    
2060    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
2061    
2062        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
2063    
2064    This patch restores SML/NJ's ability to run under win32.  There are a
2065    number of changes, including fixes for several bugs that had gone
2066    unnoticed until now:
2067    
2068      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
2069        trying to run the win32 version from within cygwin.)
2070    
2071      - There are a number of simple .bat scripts that substitute for their
2072        corresponding Unix shell-scripts.  (See below.)
2073    
2074      - The internals of ml-build have been modified slightly.  The main
2075        difference is that instead of calling ".link-sml" (or link-sml.bat)
2076        using OS.Process.system, the ML process delegates this task back
2077        to the script.  Otherwise problems arise in mixed environments such
2078        as Cygwin where scripts look and work like Unix scripts, but
2079        where OS.Process.system cannot run them.
2080    
2081      - In CM, the srcpath pickler used native pathname syntax -- which
2082        is incorrect in the case of cross-compilation.  The new pickle format
2083        is independent of platform-specific naming conventions.
2084    
2085      - Path configuration files (such as lib/pathconfig) can now choose
2086        between native and standard syntax.  Placing a line of the form
2087    
2088            standard!
2089    
2090        into the file causes all subsequent paths to be interpreted using
2091        CM standard pathname syntax (= Unix conventions); a line
2092    
2093            native!
2094    
2095        switches back to native style.  This was needed so that
2096        path config files can be written portably, see src/system/pathconfig.
2097    
2098      - Runtime system:
2099    
2100          - win32-filesys.c:  get_file_time and set_file_time now
2101            access modification time, not creation time.
2102    
2103          - I/O code made aware of new array representation.
2104    
2105          - Bug fixes in X86.prim.masm.
2106    
2107       - src/system/makeml made aware of win32. (For use under cygwin
2108            and other Unix-environments for windows.)
2109    
2110       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
2111         which caused BinIO.inputAll to fail consistently.
2112    
2113    .bat scripts:
2114    
2115       Windows .bat scripts assume that SMLNJ_HOME is defined.
2116    
2117       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
2118         applications (sml, ml-yacc, ml-lex).
2119       - ml-build.bat: analogous to ml-build.
2120       - config\install.bat:  Analogous to config/install.sh.  This requires
2121         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2122         (nmake etc. must be on the path, and vcvars32 must have been run.)
2123         Moreover, sources for ml-lex and ml-yacc need to exist under src,
2124         and the bootfile hierarchy must have been unpacked under
2125         sml.boot.x86-win32.
2126         The script is very primitive and does a poor job at error checking.
2127         It only installs the base system, ml-lex, and ml-yacc.  No other
2128         libraries are being installed (i.e., you get only those that
2129         are part of the compiler.)
2130       - link-sml.bat:  analogous to .link-sml, but not currently used
2131    
2132    Unrelated bug fixes:
2133    
2134      - ml-nlffigen now exports structures ST_* corresponding to incomplete
2135        types.
2136      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
2137        otherwise.)
2138    
2139    ----------------------------------------------------------------------
2140    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2141    Date: 2003/06/17
2142    Tag: macqueen-20030617-bug895
2143    Description:
2144    
2145    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2146    Tag will be used for new development branch (mcz-branch) for
2147    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2148    
2149    ----------------------------------------------------------------------
2150    Name: Matthias Blume (blume (at) tti - c (dot) org)
2151    Date: 2003/05/27 16:55:00 CDT
2152    Tag: blume-20030527-polyeq
2153    Description:
2154    
2155    Tried to eliminated most cases of polymorphic equality.
2156    
2157    ----------------------------------------------------------------------
2158    Name: Matthias Blume (blume (at) tti - c (dot) org)
2159    Date: 2003/05/21 17:45:00 CDT
2160    Tag: blume-20030517-complete
2161    Description:
2162    
2163    Two changes:
2164    
2165      1. Added a flag for controlling whether non-exhaustive bindings will
2166         be treated as errors (default is false).
2167      2. Cleaned up the *entire* source tree so that CMB.make goes through
2168         without a single non-exhaustive match- or bind warning.
2169    
2170    ----------------------------------------------------------------------
2171    Name: Matthias Blume (blume (at) tti - c (dot) org)
2172    Date: 2003/05/17 10:20:00 CDT
2173    Tag: blume-20030517-absyn
2174    Description:
2175    
2176    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2177    
2178       This mainly affects the quality of error messages.  However, some
2179       of the code is now more straightforward than before. (Treatment of
2180       the above four constructs in translate.sml is much simpler than
2181       the "macro-expansion" that was going on before.  Plus, the mach-
2182       compiler no longer gets invoked just to be able to compile an
2183       if-expression.)
2184    
2185    2. The ErrorMsg.Error exception is now caught and absorbed by the
2186       interactive loop.
2187    
2188    ----------------------------------------------------------------------
2189    Name: Allen Leung
2190    Date: 2003/05/16 13:05:00 CDT
2191    Tag: leunga-20030516-cygwin-runtime
2192    Description:
2193    
2194        Ported the runtime system to cygwin, which uses the unix
2195    x86-unix bin files.  Missing/buggy features:
2196    
2197       o getnetbyname, getnetbyaddr: these functions seem to be missing in
2198         the Cygwin library.
2199       o Ctrl-C handling may be flaky.
2200       o Windows system calls and Windows I/O are not supported.
2201    
2202       A new set of binfiles is located at:
2203    
2204         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2205    
2206       This is only needed for bootstrapping the cygwin version of smlnj.
2207       Other x86 versions can use the existing binfiles.
2208    
2209    ----------------------------------------------------------------------
2210    Name: Matthias Blume
2211    Date: 2003/04/08 15:42:00 CDT
2212    Tag: blume-20030408-listpair
2213    Description:
2214    
2215    1. Added a target 'mlrisc' to installer.
2216    
2217    2. Added missing elements to structure ListPair.
2218    
2219    ----------------------------------------------------------------------
2220    Name: Allen Leung
2221    Date: 2003/01/07 10:40:00 EST
2222    Tag: leunga-20030107-int-rem
2223    Description:
2224    
2225    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2226    The arguments to the SUBL instruction were swapped.
2227    
2228    ----------------------------------------------------------------------
2229    Name: Matthias Blume
2230    Date: 2002/12/12 16:25:00 EST
2231    Tag: blume-20021212-risc-ra
2232    Description:
2233    
2234    Fixed a serious bug in the rewrite code for FP spilling/reloading that
2235    sent the RA into an infinite loop when floating point registers get
2236    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2237    110.37 and 110.38.)
2238    There was another set of potential problems related to the handling of
2239    MLRISC annotations (but those did not yet cause real problems, apparently).
2240    
2241    ----------------------------------------------------------------------
2242    Name: Matthias Blume
2243    Date: 2002/12/06 22:40:00 EST
2244    Tag: blume-20021206-cm-fileid
2245    Description:
2246    
2247    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2248    This fixes the problem of CM getting confused by files that suddenly
2249    change their identity (e.g., by getting unlinked and recreated by some
2250    text editor such as vi).  There might be a better/cheaper/cleaner way
2251    of doing this, but for now this will have to do.
2252    
2253    ----------------------------------------------------------------------
2254    Name: Matthias Blume
2255    Date: 2002/10/28 09:50:00 EST
2256    Tag: blume-20021028-typecheck
2257    Description:
2258    
2259    Exported structure Typecheck from $smlnj/viscomp/core.cm.
2260    
2261    ----------------------------------------------------------------------
2262    Name: Matthias Blume
2263    Date: 2002/10/17 09:10:00 EDT
2264    Tag: Release_110_42
2265    Description:
2266    
2267    In good old tradition, there has been a slight hiccup so that we have
2268    to patch 110.42 after the fact.   The old release tag has been replaced
2269    (see below).
2270    
2271    The change solves a problem with two competing approaches the
2272    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2273    each other's way.
2274    
2275    This change only affects the runtime system code and the installer script.
2276    (No new bootfiles.)
2277    
2278    ----------------------------------------------------------------------
2279    Name: Matthias Blume
2280    Date: 2002/10/16 12:00:00 EDT
2281    Tag: Release_110_42_removed
2282    Description:
2283    
2284    New working release.  New bootfiles.
2285    
2286    ----------------------------------------------------------------------
2287    Name: Matthias Blume
2288    Date: 2002/10/10 13:10:00 EDT
2289    Tag: blume-20021010-ppc-divs
2290    Description:
2291    
2292    The mltree operator DIVS must be implemented with an overflow check on
2293    the PPC because the hardware indicates divide-by-zero using "overflow" as
2294    well.
2295    
2296    ----------------------------------------------------------------------
2297    Name: Matthias Blume
2298    Date: 2002/07/23 11:45:00
2299    Tag: blume-20020723-smlnj-home
2300    Description:
2301    
2302    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2303    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2304    CM_PATHCONFIG is also set), the path configuration file is assumed
2305    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2306    the entire tree to some other place and everything will "just work".
2307    
2308    (Companion commands such as ml-build and ml-makedepend also sense this
2309    variable.)
2310    
2311    ----------------------------------------------------------------------
2312    Name: Matthias Blume
2313    Date: 2002/07/12 21:19:00 EDT
2314    Tag: blume-20020712-liveness
2315    Description:
2316    
2317    Exported two useful "step" functions from liveness module (MLRISC).
2318    
2319    ----------------------------------------------------------------------
2320    Name: Matthias Blume
2321    Date: 2002/07/05 16:00 EDT
2322    Tag: Release_110_41
2323    Description:
2324    
2325    New working release.  New bootfiles.
2326    
2327    ----------------------------------------------------------------------
2328    Name: Matthias Blume
2329    Date: 2002/07/05 10:25:00 EDT
2330    Tag: blume-20020705-btimp
2331    Description:
2332    
2333    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2334    other clients can set up backtracing support.
2335    
2336    ----------------------------------------------------------------------
2337    Name: Matthias Blume
2338    Date: 2002/06/25 17:23:00 EDT
2339    Tag: blume-20020625-fpmax
2340    Description:
2341    
2342    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2343    case.  (The sense of the isNaN test was reversed -- which made min and
2344    max always return their first argument.)
2345    
2346    ----------------------------------------------------------------------
2347    Name: Matthias Blume
2348    Date: 2002/06/11
2349    Tag: blume-20020611-unixpath
2350    Description:
2351    
2352    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2353    
2354    ----------------------------------------------------------------------
2355    Name: Matthias Blume
2356    Date: 2002/06/10 16:35:00 EDT
2357    Tag: blume-20020610-ieeereal
2358    Description:
2359    
2360    I back-ported my implementation of IEEEReal.fromString from the
2361    idlbasis-devel branch so that we can test it.
2362    
2363    Another small change is that ppDec tries to give more information
2364    than just "<sig>" in the case of functors.  However, this code is
2365    broken in some mysterious way if the functor's body's signature
2366    has not been declared by ascription but gets inferred from the
2367    implementation.  This needs fixing...
2368    
2369    ----------------------------------------------------------------------
2370    Name: Matthias Blume
2371    Date: 2002/05/31
2372    Tag: blume-20020531-btrace-mode
2373    Description:
2374    
2375    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2376    the wayside when I switched over to using Controls everywhere.)
2377    
2378    ----------------------------------------------------------------------
2379    Name: Lal George
2380    Date: 2002/05/23 12:21:40 EDT
2381    Tag: george-20020523-visual-labels
2382    Description:
2383    
2384    Labels are now displayed in the graphical output to make
2385    the fall-through and target blocks obvious.
2386    
2387    ----------------------------------------------------------------------
2388    Name: Matthias Blume
2389    Date: 2002/05/22 11:03:00 EDT
2390    Tag: blume-20020522-shrink
2391    Description:
2392    
2393    John tweaked yesterday's fix for 1131 to handle an out-of-memory
2394    situation that comes up when allocating huge arrays.
2395    
2396    ----------------------------------------------------------------------
2397    Name: Matthias Blume
2398    Date: 2002/05/21 16:00:00 EDT
2399    Tag: Release_110_40
2400    Description:
2401    
2402    New working release (110.40).  New bootfiles.
2403    
2404    [Also: John Reppy fixed GC bug 1131.]
2405    
2406    ----------------------------------------------------------------------
2407    Name: Matthias Blume
2408    Date: 2002/05/21 12:35:00 EDT
2409    Tag: blume-20020521-cmdoc
2410    Description:
2411    
2412    CM documentation update.
2413    
2414    ----------------------------------------------------------------------
2415    Name: Matthias Blume
2416    Date: 2002/05/21 10:55:00 EDT
2417    Tag: blume-20020521-misc
2418    Description:
2419    
2420    - John tweaked runtime to be silent on heap export (except when
2421      GC messages are on).
2422    - I added a few more things (cross-compiling versions of CMB) to
2423      config/preloads (as suggestions).
2424    
2425    ----------------------------------------------------------------------
2426    Name: Matthias Blume
2427    Date: 2002/05/20 22:25:00 EDT
2428    Tag: blume-20020520-controls
2429    Description:
2430    
2431    - Added ControlUtil structure to control-lib.cm.
2432    - Use it throughout.
2433    - Used Controls facility to define MLRISC controls (as opposed to
2434      registering MLRISC control ref cells with Controls after the fact)
2435    - Fixed messed-up controls priorities.
2436    
2437    * Removed again all the stuff from config/preloads that one wouldn't
2438      be able to preload at the time the initial heap image is built.
2439      (Many libraries, e.g., CML, do not exist yet at this time.  The
2440       only libraries that can be preloaded via config/preloads are those
2441       that come bundled with the bootfiles.)
2442    
2443    ----------------------------------------------------------------------
2444    Name: Matthias Blume
2445    Date: 2002/05/20 10:59:00 EDT
2446    Tag: blume-20020520-preloads
2447    Description:
2448    
2449    Added a lot of commented-out suggestions for things to be included
2450    in config/preloads.
2451    
2452    ----------------------------------------------------------------------
2453    Name: Allen Leung
2454    Date: 2002/05/18 14:20:00 EDT
2455    Tag: leunga-20020518-mdl
2456    Description:
2457    
2458    o Made the mdl tool stuff compile and run again.
2459    o I've disabled all the stuff that depends on RTL specifications; they
2460      are all badly broken anyway.
2461    
2462    ----------------------------------------------------------------------
2463    Name: Matthias Blume
2464    Date: 2002/05/17 16:49:00 EDT
2465    Tag: blume-20020517-controls
2466    Description:
2467    
2468    0. John Reppy made several modifications to the SML/NJ library.
2469       In particular, there is a shiny new controls-lib.cm.
2470    
2471    1. Pushed new controls interface through compiler so that everything
2472       compiles again.
2473    
2474    2. Added FormatComb and FORMAT_COMB to the CML version of the
2475       SML/NJ library (so that CML compiles again).
2476    
2477    3. Modified init scripts because XXX_DEFAULT environment variables
2478       are no longer with us.  (Boot-time initialization is now done
2479       using the same environment variables that are also used for
2480       startup-time initialization of controls.)
2481    
2482    ----------------------------------------------------------------------
2483    Name: Lal George
2484    Date: 2002/05/15 09:20:10 EDT
2485    Tag: george-20020515-pseudo-op-decls
2486    Description:
2487    
2488            All pseudo-ops emitted before the first segment declaration
2489    such as TEXT, DATA, and BSS directives are assumed to be global
2490    declarations and are emitted first in the assembly file. This is
2491    useful in a number of situations where one has pseudo-ops that are not
2492    specific to any segment, and also works around the constraint that one
2493    cannot have client pseudo-ops in the TEXT segment.
2494    
2495            Because no segment is associated with these declarations it is
2496    an error to allocate any space or objects before the first segment
2497    directive and an exception will be raised. However, we cannot make
2498    this check for client pseudo-ops.
2499    
2500            These top level declarations are a field in the CFG graph_info.
2501    In theory you can continue to add to this field after the CFG has been
2502    built -- provided you know what you are doing;-)
2503    
2504    ----------------------------------------------------------------------
2505    Name: Matthias Blume
2506    Date: 2002/05/13 16:40:00 EDT
2507    Tag: blume-20020513-pp-etc
2508    Description:
2509    
2510    A few minor bugfixes:
2511    
2512      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2513        (Bogus printouts for redefined bindings still occur.  Compiler
2514        bug should no longer occur now.  We need to redo the prettyprinter
2515        from scratch.)
2516    
2517      - CM pathname printer now also adds escape sequences for ( and )
2518    
2519      - commend and docu fixes for ml-nlffi
2520    
2521    ----------------------------------------------------------------------
2522    Name: Matthias Blume
2523    Date: 2002/05/10 16:40:00 EDT
2524    Tag: blume-20020510-erg-textio
2525    Description:
2526    
2527    Applied the following bugfix provided by Emden Gansner:
2528    
2529        Output is corrupted when outputSubstr is used rather than output.
2530    
2531        The problem occurs when a substring
2532    
2533            ss = (s, dataStart, dataLen)
2534    
2535        where dataStart > 0, fills a stream buffer with avail bytes left.
2536        avail bytes of s, starting at index dataStart, are copied into the
2537        buffer, the buffer is flushed, and then the remaining dataLen-avail
2538        bytes of ss are copied into the beginning of the buffer. Instead of
2539        starting this copy at index dataStart+avail in s, the current code
2540        starts the copy at index avail.
2541    
2542        Fix:
2543        In text-io-fn.sml, change line 695 from
2544             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2545        to
2546             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2547    
2548    ----------------------------------------------------------------------
2549    Name: Matthias Blume
2550    Date: 2002/04/12 13:55:00 EDT
2551    Tag: blume-20020412-assyntax
2552    Description:
2553    
2554    1. Grabbed newer assyntax.h from the XFree86 project.
2555    2. Fiddled with how to compile X86.prim.asm without warnings.
2556    3. (Very) Minor cleanup in CM.
2557    
2558    ----------------------------------------------------------------------
2559    Name: Matthias Blume
2560    Date: 2002/04/01 (no joke!) 17:07:00 EST
2561    Tag: blume-20020401-x86div
2562    Description:
2563    
2564    Added full support for div/mod/rem/quot on the x86, using the machine
2565    instruction's two results (without clumsily recomputing the remainder)
2566    directly where appropriate.
2567    
2568    Some more extensive power-of-two support was added to the x86 instruction
2569    selector (avoiding expensive divs, mods, and muls where they can be
2570    replaced with cheaper shifts and masks).  However, this sort of thing
2571    ought to be done earlier, e.g., within the CPS optimizer so that
2572    all architectures benefit from it.
2573    
2574    The compiler compiles to a fixed point, but changes might be somewhat
2575    fragile nevertheless.  Please, report any strange things that you might
2576    see wrt. div/mod/quot/rem...
2577    
2578    ----------------------------------------------------------------------
2579    Name: Matthias Blume
2580    Date: 2002/03/29 17:22:00
2581    Tag: blume-20020329-div
2582    Description:
2583    
2584    Fixed my broken div/mod logic.  Unfortunately, this means that the
2585    inline code for div/mod now has one more comparison than before.
2586    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2587    The problem was with quotient = 0, because that alone does not tell
2588    us which way the rounding went.  One then has to look at whether
2589    remainder and divisor have the same sign...  :(
2590    
2591    Anyway, I replaced the bootfiles with fresh ones...
2592    
2593    ----------------------------------------------------------------------
2594    Name: Matthias Blume
2595    Date: 2002/03/29 14:10:00 EST
2596    Tag: blume-20020329-inlprims
2597    Description:
2598    
2599    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2600    
2601    Primops have changed. This means that the bin/boot-file formats have
2602    changed as well.
2603    
2604    To make sure that there is no confusion, I made a new version.
2605    
2606    
2607    CHANGES:
2608    
2609    * removed REMT from mltree (remainder should never overflow).
2610    
2611    * added primops to deal with divisions of all flavors to the frontend
2612    
2613    * handled these primops all the way through so they map to their respective
2614      MLRISC support
2615    
2616    * used these primops in the implementation of Int, Int32, Word, Word32
2617    
2618    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2619    
2620    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2621    
2622    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2623      floating point
2624    
2625    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2626      and Real (but Real.abs maps to a separate floating-point-only primop)
2627    
2628    
2629    TODO items:
2630    
2631    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2632      instructions because the machine instruction encoder cannot handle
2633      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2634      REM in the first place, and REM is supported by the x86 machine coder.)
2635    
2636    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2637      selection phase.  (The two can be streamlined because the hardware
2638      delivers both quotient and remainder at the same time anyway.)
2639    
2640    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2641      (Currently the behavior is inconsistent both across architectures and
2642      wrt. the draft Basis spec.)
2643    
2644    * Word8 should eventually be handled natively, too.
2645    
2646    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2647      as if there currently is no execution path that could trigger it in
2648      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2649      hold for things like multiplication and division.)
2650    
2651    ----------------------------------------------------------------------
2652    Name: Matthias Blume
2653    Date: 2002/03/27 16:27:00 EST
2654    Tag: blume-20020327-mlrisc-divisions
2655    Description:
2656    
2657    Added support for all four division operations (ML's div, mod, quot,
2658    and rem) to MLRISC.  In the course of doing so, I also rationalized
2659    the naming (no more annoying switch-around of DIV and QUOT), by
2660    parameterizing the operation by div_rounding_mode (which can be either
2661    DIV_TO_ZERO or DIV_TO_NEGINF).
2662    
2663    The generic MLTreeGen functor takes care of compiling all four
2664    operations down to only round-to-zero div.
2665    
2666    Missing pieces:
2667    
2668      * Doing something smarter than relying on MLTreeGen on architectures
2669        like, e.g., the x86 where hardware division delivers both quotient and
2670        remainder at the same time.  With this, the implementation of the
2671        round-to-neginf operations could be further streamlined.
2672    
2673      * Remove inlining support for div/mod/rem from the frontend and replace it
2674        with primops that get carried through to the backend.  Do this for all
2675        int and word types.
2676    
2677    ----------------------------------------------------------------------
2678    Name: Matthias Blume
2679    Date: 2002/03/25 17:25:00 EST
2680    Tag: blume-20020325-divmod
2681    Description:
2682    
2683    I improved (hopefully without breaking them) the implementation of Int.div,
2684    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2685    advantage of the following observations:
2686    
2687      Let  q = x quot y      r = x rem y
2688           d = x div  y      m = x mod y
2689    
2690    where "quot" is the round-to-zero version of integer division that
2691    hardware usually provides.  Then we have:
2692    
2693         r = x - q * y        where neither the * nor the - will overflow
2694         d = if q >= 0 orelse x = q * y then q else q - 1
2695                              where neither the * nor the - will overflow
2696         m = if q >= 0 orelse r = 0 then r else r + y
2697                              where the + will not overflow
2698    
2699    This results in substantial simplification of the generated code.
2700    The following table shows the number of CFG nodes and edges generated
2701    for
2702            fun f (x, y) = x OPER y
2703            (* with OPER \in div, mod, quot, rem *)
2704    
2705    
2706        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2707        --------------------------------------------------------
2708         div |         24 |         39 |         12 |         16
2709         mod |         41 |         71 |         12 |         16
2710        quot |          8 |         10 |          8 |         10
2711         rem |         10 |         14 |          8 |         10
2712    
2713    
2714    ----------------------------------------------------------------------
2715    Name: Matthias Blume
2716    Date: 2002/03/25 22:06:00 EST
2717    Tag: blume-20020325-cprotobug
2718    Description:
2719    
2720    Fixed a bug in cproto (c prototype decoder).
2721    
2722    ----------------------------------------------------------------------
2723    Name: Matthias Blume
2724    Date: 2002/03/25 16:00:00 EST
2725    Tag: blume-20020325-raw-primops
2726    Description:
2727    
2728    I did some cleanup to Allen's new primop code and
2729    replaced yesterday's bootfiles with new ones.
2730    (But they are stored in the same place.)
2731    
2732    ----------------------------------------------------------------------
2733    Name: Matthias Blume
2734    Date: 2002/03/24 22:40:00 EST
2735    Tag: blume-20020324-bootfiles
2736    Description:
2737    
2738    Made the bootfiles that Allen asked for.
2739    
2740    ----------------------------------------------------------------------
2741    Name: Allen Leung
2742    Date: 2002/03/23 15:50:00 EST
2743    Tag: leunga-20020323-flint-cps-rcc-primops
2744    Description:
2745    
2746      1. Changes to FLINT primops:
2747    
2748        (* make a call to a C-function;
2749         * The primop carries C function prototype information and specifies
2750         * which of its (ML-) arguments are floating point. C prototype
2751         * information is for use by the backend, ML information is for
2752         * use by the CPS converter. *)
2753      | RAW_CCALL of { c_proto: CTypes.c_proto,
2754                       ml_args: ccall_type list,
2755                       ml_res_opt: ccall_type option,
2756                       reentrant : bool
2757                     } option
2758       (* Allocate uninitialized storage on the heap.
2759        * The record is meant to hold short-lived C objects, i.e., they
2760        * are not ML pointers.  With the tag, the representation is
2761        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2762        *)
2763      | RAW_RECORD of {tag:bool,sz:int}
2764      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2765    
2766      2.  These CPS primops are now overloaded:
2767    
2768           rawload of {kind:numkind}
2769           rawstore of {kind:numkind}
2770    
2771          The one argument form is:
2772    
2773             rawload {kind} address
2774    
2775          The two argument form is:
2776    
2777             rawload {kind} [ml object, byte-offset]
2778    
2779      3. RAW_CCALL/RCC now takes two extra arguments:
2780    
2781         a. The first is whether the C call is reentrant, i.e., whether
2782            ML state should be saved and restored.
2783         b. The second argument is a string argument specifying the name of
2784            library and the C function.
2785    
2786         These things are currently not handled in the code generator, yet.
2787    
2788      4. In CProto,
2789    
2790         An encoding type of "bool" means "ml object" and is mapped into
2791         C prototype of PTR.  Note that "bool" is different than "string",
2792         even though "string" is also mapped into PTR, because "bool"
2793         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2794    
2795      5. Pickler/unpicker
2796    
2797         Changed to handle RAW_RECORD and newest RAW_CCALL
2798    
2799      6. MLRiscGen,
2800    
2801         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2802         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2803            in the file CodeGen/cpscompile/cps-c-calls.sml
2804    
2805      7. Added the conditional move operator
2806    
2807             condmove of branch
2808    
2809         to cps.  Generation of this is still buggy so it is currently
2810         disabled.
2811    
2812    ----------------------------------------------------------------------
2813    Name: Lal George
2814    Date: 2002/03/22 14:18:25 EST
2815    Tag: george-20020322-cps-branch-prob
2816    Description:
2817    
2818    Implemented the Ball-Larus branch prediction-heuristics, and
2819    incorporated graphical viewers for control flow graphs.
2820    
2821    Ball-Larus Heuristics:
2822    ---------------------
2823    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
2824    
2825    By design it uses the Dempster-Shafer theory for combining
2826    probabilities.  For example, in the function:
2827    
2828        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
2829    
2830    the ball-larus heuristics predicts that the n=0 is unlikely
2831    (OH-heuristic), and the 'then' branch is unlikely because of the
2832    RH-heuristic -- giving the 'then' branch an even lower combined
2833    probability using the Dempster-Shafer theory.
2834    
2835    Finally, John Reppy's loop analysis in MLRISC, further lowers the
2836    probability of the 'then' branch because of the loop in the else
2837    branch.
2838    
2839    
2840    Graphical Viewing:
2841    ------------------
2842    I merely plugged in Allen's graphical viewers into the compiler. The
2843    additional code is not much. At the top level, saying:
2844    
2845            Control.MLRISC.getFlag "cfg-graphical-view" := true;
2846    
2847    will display the graphical view of the control flow graph just before
2848    back-patching.  daVinci must be in your path for this to work. If
2849    daVinci is not available, then the default viewer can be changed
2850    using:
2851    
2852            Control.MLRISC.getString "viewer"
2853    
2854    which can be set to "dot" or "vcg" for the corresponding viewers. Of
2855    course, these viewers must be in your path.
2856    
2857    The above will display the compilation unit at the level of clusters,
2858    many of which are small, boring, and un-interesting. Also setting:
2859    
2860            Control.MLRISC.getInt "cfg-graphical-view_size"
2861    
2862    will display clusters that are larger than the value set by the above.
2863    
2864    
2865    ----------------------------------------------------------------------
2866    Name: Matthias Blume
2867    Date: 2002/03/21 22:20:00 EST
2868    Tag: blume-20020321-kmp-bugfix
2869    Description:
2870    
2871    Changed the interface to the KMP routine in PreString and fixed
2872    a minor bug in one place where it was used.
2873    
2874    ----------------------------------------------------------------------
2875    Name: Allen Leung
2876    Date: 2002/03/21 20:30:00 EST
2877    Tag: leunga-20020321-cfg
2878    Description:
2879    
2880      Fixed a potential problem in cfg edge splitting.
2881    
2882    ----------------------------------------------------------------------
2883    Name: Allen Leung
2884    Date: 2002/03/21 17:15:00 EST
2885    Tag: leunga-20020321-x86-fp-cfg
2886    Description:
2887    
2888      1. Recoded the buggy parts of x86-fp.
2889    
2890         a. All the block reordering code has been removed.
2891            We now depend on the block placement phases to do this work.
2892    
2893         b. Critical edge splitting code has been simplified and moved into the
2894            CFG modules, as where they belong.
2895    
2896         Both of these were quite buggy and complex.  The code is now much, much
2897         simpler.
2898    
2899      2. X86 backend.
2900    
2901         a. Added instructions for 64-bit support.  Instruction selection for
2902            64-bit has not been committed, however, since that
2903            requires changes to MLTREE which haven't been approved by
2904            Lal and John.
2905    
2906         b. Added support for FUCOMI and FUCOMIP when generating code for
2907            PentiumPro and above.  We only generate these instructions in
2908            the fast-fp mode.
2909    
2910         c. Added cases for JP and JNP in X86FreqProps.
2911    
2912      3. CFG
2913    
2914         CFG now has a bunch of methods for edge splitting and merging.
2915    
2916      4. Machine description.
2917    
2918         John's simplification of MLTREE_BASIS.fcond broke a few machine
2919         description things:
2920    
2921         rtl-build.{sig,sml} and hppa.mdl fixed.
2922    
2923         NOTE: the machine description stuff in the repository is still broken.
2924               Again, I can't put my fixes in because that involves
2925               changes to MLTREE.
2926    
2927    ----------------------------------------------------------------------
2928    Name: Matthias Blume
2929    Date: 2002/03/20 15:55:00 EST
2930    Tag: blume-20020320-kmp
2931    Description:
2932    
2933    Implemented Knuth-Morris-Pratt string matching in PreString and used
2934    it for String.isSubstring, Substring.isSubstring, and
2935    Substring.position.
2936    
2937    (Might need some stress-testing.  Simple examples worked fine.)
2938    
2939    ----------------------------------------------------------------------
2940    Name: Matthias Blume
2941    Date: 2002/03/19 16:37:00 EST
2942    Tag: blume-20020319-witnesses
2943    Description:
2944    
2945    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
2946    
2947    This implements a generic mechanism for changing constness qualifiers
2948    anywhere within big C types without resorting to outright "casts".
2949    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
2950    modify the constness at the outermost level.)
2951    The implementation of "convert" is based on the idea of "witness"
2952    values -- values that are not used by the operation but whose types
2953    "testify" to their applicability.  On the implementation side, "convert"
2954    is simply a projection (returning its second curried argument).  With
2955    cross-module inlining, it should not result in any machine code being
2956    generated.
2957    
2958    ----------------------------------------------------------------------
2959    Name: Matthias Blume
2960    Date: 2002/03/15 16:40:00 EST
2961    Tag: blume-20020315-basis
2962    Description:
2963    
2964    Provided (preliminary?) implementations for
2965    
2966      {String,Substring}.{concatWith,isSuffix,isSubstring}
2967    
2968    and
2969    
2970      Substring.full
2971    
2972    Those are in the Basis spec but they were missing in SML/NJ.
2973    
2974    ----------------------------------------------------------------------
2975    Name: Matthias Blume
2976    Date: 2002/03/14 21:30:00 EST
2977    Tag: blume-20020314-controls
2978    Description:
2979    
2980    Controls:
2981    ---------
2982    
2983    1. Factored out the recently-added Controls : CONTROLS stuff and put
2984       it into its own library $/controls-lib.cm.  The source tree for
2985       this is under src/smlnj-lib/Controls.
2986    
2987    2. Changed the names of types and functions in this interface, so they
2988       make a bit more "sense":
2989    
2990          module -> registry
2991          'a registry -> 'a group
2992    
2993    3. The interface now deals in ref cells only.  The getter/setter interface
2994       is (mostly) gone.
2995    
2996    4. Added a function that lets one register an already-existing ref cell.
2997    
2998    5. Made the corresponding modifications to the rest of the code so that
2999       everything compiles again.
3000    
3001    6. Changed the implementation of Controls.MLRISC back to something closer
3002       to the original.  In particular, this module (and therefore MLRISC)
3003       does not depend on Controls.  There now is some link-time code in
3004       int-sys.sml that registers the MLRISC controls with the Controls
3005       module.
3006    
3007    CM:
3008    ---
3009    
3010      * One can now specify the lambda-split aggressiveness in init.cmi.
3011    
3012    ----------------------------------------------------------------------
3013    Name: Allen Leung
3014    Date: 2002/03/13 17:30:00 EST
3015    Tag: leunga-20020313-x86-fp-unary
3016    Description:
3017    
3018    Bug fix for:
3019    
3020    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
3021    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
3022    > - fun f(x,(y,z)) = Real.~ y;
3023    > [autoloading]
3024    > [autoloading done]
3025    >       fchsl   (%eax), 184(%esp)
3026    > Error: MLRisc bug: X86MCEmitter.emitInstr
3027    >
3028    > uncaught exception Error
3029    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
3030    
3031    The problem was that the code generator did not generate any fp registers
3032    in this case, and the ra didn't know that it needed to run the X86FP phase to
3033    translate the pseudo fp instruction.   This only happened with unary fp
3034    operators in certain situations.
3035    
3036    ----------------------------------------------------------------------
3037    Name: Matthias Blume
3038    Date: 2002/03/13 14:00:00 EST
3039    Tag: blume-20020313-overload-etc
3040    Description:
3041    
3042    1. Added _overload as a synonym for overload for backward compatibility.
3043       (Control.overloadKW must be true for either version to be accepted.)
3044    
3045    2. Fixed bug in install script that caused more things to be installed
3046       than what was requested in config/targets.
3047    
3048    3. Made CM aware of the (_)overload construct so that autoloading
3049       works.
3050    
3051    ----------------------------------------------------------------------
3052    Name: Matthias Blume
3053    Date: 2002/03/12 22:03:00 EST
3054    Tag: blume-20020312-url
3055    Description:
3056    
3057    Forgot to update BOOT and srcarchiveurl.
3058    
3059    ----------------------------------------------------------------------
3060    Name: Matthias Blume
3061    Date: 2002/03/12 17:30:00 EST
3062    Tag: blume-20020312-version110392
3063    Description:
3064    
3065    Yet another version number bump (because of small changes to the
3066    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
3067    
3068    Changes:
3069    
3070      The new pid generation scheme described a few weeks ago was overly
3071      complicated.  I implemented a new mechanism that is simpler and
3072      provides a bit more "stability":  Once CM has seen a compilation
3073      unit, it keeps its identity constant (as long as you do not delete
3074      those crucial CM/GUID/* files).  This means that when you change
3075      an interface, compile, then go back to the old interface, and
3076      compile again, you arrive at the original pid.
3077    
3078      There now also is a mechanism that instructs CM to use the plain
3079      environment hash as a module's pid (effectively making its GUID
3080      the empty string).  For this, "noguid" must be specified as an
3081      option to the .sml file in question within its .cm file.
3082      This is most useful for code that is being generated by tools such
3083      as ml-nlffigen (because during development programmers tend to
3084      erase the tool's entire output directory tree including CM's cached
3085      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
3086      revert to the old, broken behavior of SML/NJ, but in specific cases
3087      where there is no danger of interface confusion, its use is ok
3088      (I think).
3089    
3090      ml-nlffigen by default generates "noguid" annotations.  They can be
3091      turned off by specifying -guid in its command line.
3092    
3093    ----------------------------------------------------------------------
3094    Name: Lal George
3095    Date: 2002/03/12 12 14:42:36 EST
3096    Tag: george-20020312-frequency-computation
3097    Description:
3098    
3099    Integrated jump chaining and static block frequency into the
3100    compiler. More details and numbers later.
3101    
3102    ----------------------------------------------------------------------
3103    Name: Lal George
3104    Date: 2002/03/11 11 22:38:53 EST
3105    Tag: george-20020311-jump-chain-elim
3106    Description:
3107    
3108    Tested the jump chain elimination on all architectures (except the
3109    hppa).  This is on by default right now and is profitable for the
3110    alpha and x86, however, it may not be profitable for the sparc and ppc
3111    when compiling the compiler.
3112    
3113    The gc test will typically jump to a label at the end of the cluster,
3114    where there is another jump to an external cluster containing the actual
3115    code to invoke gc. This is to allow factoring of common gc invocation
3116    sequences. That is to say, we generate:
3117    
3118            f:
3119               testgc
3120               ja   L1      % jump if above to L1
3121    
3122            L1:
3123               jmp L2
3124    
3125    
3126    After jump chain elimination the 'ja L1' instructions is converted to
3127    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
3128    up being implemented in their long form (if L2 is far away) using:
3129    
3130            jbe     L3      % jump if below or equal to L3
3131            jmp     L2
3132         L3:
3133            ...
3134    
3135    
3136    For large compilation units L2  may be far away.
3137    
3138    
3139    ----------------------------------------------------------------------
3140    Name: Matthias Blume
3141    Date: 2002/03/11 13:30:00 EST
3142    Tag: blume-20020311-mltreeeval
3143    Description:
3144    
3145    A functor parameter was missing.
3146    
3147    ----------------------------------------------------------------------
3148    Name: Allen Leung
3149    Date: 2002/03/11 10:30:00 EST
3150    Tag: leunga-20020311-runtime-string0
3151    Description:
3152    
3153       The representation of the empty string now points to a
3154    legal null terminated C string instead of unit.  It is now possible
3155    to convert an ML string into C string with InlineT.CharVector.getData.
3156    This compiles into one single machine instruction.
3157    
3158    ----------------------------------------------------------------------
3159    Name: Allen Leung
3160    Date: 2002/03/10 23:55:00 EST
3161    Tag: leunga-20020310-x86-call
3162    Description:
3163    
3164       Added machine generation for CALL instruction (relative displacement mode)
3165    
3166    ----------------------------------------------------------------------
3167    Name: Matthias Blume
3168    Date: 2002/03/08 16:05:00
3169    Tag: blume-20020308-entrypoints
3170    Description:
3171    
3172    Version number bumped to 110.39.1.  NEW BOOTFILES!
3173    
3174    Entrypoints: non-zero offset into a code object where execution should begin.
3175    
3176    - Added the notion of an entrypoint to CodeObj.
3177    - Added reading/writing of entrypoint info to Binfile.
3178    - Made runtime system bootloader aware of entrypoints.
3179    - Use the address of the label of the first function given to mlriscGen
3180      as the entrypoint.  This address is currently always 0, but it will
3181      not be 0 once we turn on block placement.
3182    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3183      with entry points) from mlriscGen.
3184    
3185    ----------------------------------------------------------------------
3186    Name: Allen Leung
3187    Date: 2002/03/07 20:45:00 EST
3188    Tag: leunga-20020307-x86-cmov
3189    Description:
3190    
3191       Bug fixes for CMOVcc on x86.
3192    
3193       1. Added machine code generation for CMOVcc
3194       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3195       3. CMOVcc cannot have an immediate operand as argument.
3196    
3197    ----------------------------------------------------------------------
3198    Name: Matthias Blume
3199    Date: 2002/03/07 16:15:00 EST
3200    Tag: blume-20020307-controls
3201    Description:
3202    
3203    This is a very large but mostly boring patch which makes (almost)
3204    every tuneable compiler knob (i.e., pretty much everything under
3205    Control.* plus a few other things) configurable via both the command
3206    line and environment variables in the style CM did its configuration
3207    until now.
3208    
3209    Try starting sml with '-h' (or, if you are brave, '-H')
3210    
3211    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3212    implements the underlying generic mechanism.
3213    
3214    The interface to some of the existing such facilities has changed somewhat.
3215    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3216    (The getFoo interface is still there for backward-compatibility, but its
3217    use is deprecated.)
3218    
3219    The ml-build script passes -Cxxx=yyy command-line arguments through so
3220    that one can now twiddle the compiler settings when using this "batch"
3221    compiler.
3222    
3223    TODO items:
3224    
3225    We should go through and throw out all controls that are no longer
3226    connected to anything.  Moreover, we should go through and provide
3227    meaningful (and correct!) documentation strings for those controls
3228    that still are connected.
3229    
3230    Currently, multiple calls to Controls.new are accepted (only the first
3231    has any effect).  Eventually we should make sure that every control
3232    is being made (via Controls.new) exactly once.  Future access can then
3233    be done using Controls.acc.
3234    
3235    Finally, it would probably be a good idea to use the getter-setter
3236    interface to controls rather than ref cells.  For the time being, both
3237    styles are provided by the Controls module, but getter-setter pairs are
3238    better if thread-safety is of any concern because they can be wrapped.
3239    
3240    *****************************************
3241    
3242    One bug fix: The function blockPlacement in three of the MLRISC
3243    backpatch files used to be hard-wired to one of two possibilities at
3244    link time (according to the value of the placementFlag).  But (I
3245    think) it should rather sense the flag every time.
3246    
3247    *****************************************
3248    
3249    Other assorted changes (by other people who did not supply a HISTORY entry):
3250    
3251    1. the cross-module inliner now works much better (Monnier)
3252    2. representation of weights, frequencies, and probabilities in MLRISC
3253       changed in preparation of using those for weighted block placement
3254       (Reppy, George)
3255    
3256    ----------------------------------------------------------------------
3257    Name: Lal George
3258    Date: 2002/03/07 14:44:24 EST 2002
3259    Tag: george-20020307-weighted-block-placement
3260    
3261    Tested the weighted block placement optimization on all architectures
3262    (except the hppa) using AMPL to generate the block and edge frequencies.
3263    Changes were required in the machine properties to correctly
3264    categorize trap instructions. There is an MLRISC flag
3265    "weighted-block-placement" that can be used to enable weighted block
3266    placement, but this will be ineffective without block/edge
3267    frequencies (coming soon).
3268    
3269    
3270    ----------------------------------------------------------------------
3271    Name: Lal George
3272    Date: 2002/03/05 17:24:48 EST
3273    Tag: george-20020305-linkage-cluster
3274    
3275    In order to support the block placement optimization, a new cluster
3276    is generated as the very first cluster (called the linkage cluster).
3277    It contains a single jump to the 'real' entry point for the compilation
3278    unit. Block placement has no effect on the linkage cluster itself, but
3279    all the other clusters  have full freedom in the manner in which they
3280    reorder blocks or functions.
3281    
3282    On the x86 the typical linkage code that is generated is:
3283       ----------------------
3284            .align 2
3285       L0:
3286            addl    $L1-L0, 72(%esp)
3287            jmp     L1
3288    
3289    
3290            .align  2
3291       L1:
3292       ----------------------
3293    
3294    72(%esp) is the memory location for the stdlink register. This
3295    must contain the address of the CPS function being called. In the
3296    above example, it contains the address of  L0; before
3297    calling L1 (the real entry point for the compilation unit), it
3298    must contain the address for L1, and hence
3299    
3300            addl $L1-L0, 72(%esp)
3301    
3302    I have tested this on all architectures except the hppa.The increase
3303    in code size is of course negligible
3304    
3305    ----------------------------------------------------------------------
3306    Name: Allen Leung
3307    Date: 2002/03/03 13:20:00 EST
3308    Tag: leunga-20020303-mlrisc-tools
3309    
3310      Added #[ ... ] expressions to mlrisc tools
3311    
3312    ----------------------------------------------------------------------
3313    Name: Matthias Blume
3314    Date: 2002/02/27 12:29:00 EST
3315    Tag: blume-20020227-cdebug
3316    Description:
3317    
3318    - made types in structure C and C_Debug to be equal
3319    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3320    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3321    
3322    ----------------------------------------------------------------------
3323    Name: Matthias Blume
3324    Date: 2002/02/26 12:00:00 EST
3325    Tag: blume-20020226-ffi
3326    Description:
3327    
3328    1. Fixed a minor bug in CM's "noweb" tool:
3329       If numbering is turned off, then truly don't number (i.e., do not
3330       supply the -L option to noweb).  The previous behavior was to supply
3331       -L'' -- which caused noweb to use the "default" line numbering scheme.
3332       Thanks to Chris Richards for pointing this out (and supplying the fix).
3333    
3334    2. Once again, I reworked some aspects of the FFI:
3335    
3336       A. The incomplete/complete type business:
3337    
3338       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3339         gone!
3340       - ML types representing an incomplete type are now *equal* to
3341         ML types representing their corresponding complete types (just like
3342         in C).  This is still safe because ml-nlffigen will not generate
3343         RTTI for incomplete types, nor will it generate functions that
3344         require access to such RTTI.   But when ML code generated from both
3345         incomplete and complete versions of the C type meet, the ML types
3346         are trivially interoperable.
3347    
3348         NOTE:  These changes restore the full generality of the translation
3349         (which was previously lost when I eliminated functorization)!
3350    
3351       B. Enum types:
3352    
3353       - Structure C now has a type constructor "enum" that is similar to
3354         how the "su" constructor works.  However, "enum" is not a phantom
3355         type because each "T enum" has values (and is isomorphic to
3356         MLRep.Signed.int).
3357       - There are generic access operations for enum objects (using
3358         MLRep.Signed.int).
3359       - ml-nlffigen will generate a structure E_foo for each "enum foo".
3360         * The structure contains the definition of type "mlrep" (the ML-side
3361         representation type of the enum).  Normally, mlrep is the same
3362         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3363         then mlrep will be defined as a datatype -- thus facilitating
3364         pattern matching on mlrep values.
3365         ("-ec" will be suppressed if there are duplicate values in an
3366          enumeration.)
3367         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3368         will be generated for each C enum constant xxx.
3369         * Conversion functions m2i and i2m convert between mlrep and
3370         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
3371         * Coversion functions c and ml convert between mlrep and "tag enum".
3372         * Access functions (get/set) fetch and store mlrep values.
3373       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3374         enumerations are merged into one single enumeration represented by
3375         structure E_'.
3376    
3377    ----------------------------------------------------------------------
3378    Name: Allen Leung
3379    Date: 2002/02/25 04:45:00 EST
3380    Tag: leunga-20020225-cps-spill
3381    
3382    This is a new implementation of the CPS spill phase.
3383    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3384    In case of problems, replace it with the old file spill.sml
3385    
3386    The current compiler runs into some serious performance problems when
3387    constructing a large record.  This can happen when we try to compile a
3388    structure with many items.  Even a very simple structure like the following
3389    makes the compiler slow down.
3390    
3391        structure Foo = struct
3392           val x_1 = 0w1 : Word32.int
3393           val x_2 = 0w2 : Word32.int
3394           val x_3 = 0w3 : Word32.int
3395           ...
3396           val x_N = 0wN : Word32.int
3397        end
3398    
3399    The following table shows the compile time, from N=1000 to N=4000,
3400    with the old compiler:
3401    
3402    N
3403    1000   CPS 100 spill                           0.04u  0.00s  0.00g
3404           MLRISC ra                               0.06u  0.00s  0.05g
3405              (spills = 0 reloads = 0)
3406           TOTAL                                   0.63u  0.07s  0.21g
3407    
3408    1100   CPS 100 spill                           8.25u  0.32s  0.64g
3409           MLRISC ra                               5.68u  0.59s  3.93g
3410              (spills = 0 reloads = 0)
3411           TOTAL                                   14.71u  0.99s  4.81g
3412    
3413    1500   CPS 100 spill                           58.55u  2.34s  1.74g
3414           MLRISC ra                               5.54u  0.65s  3.91g
3415              (spills = 543 reloads = 1082)
3416           TOTAL                                   65.40u  3.13s  6.00g
3417    
3418    2000   CPS 100 spill                           126.69u  4.84s  3.08g
3419           MLRISC ra                               0.80u  0.10s  0.55g
3420              (spills = 42 reloads = 84)
3421           TOTAL                                   129.42u  5.10s  4.13g
3422    
3423    3000   CPS 100 spill                           675.59u  19.03s  11.64g
3424           MLRISC ra                               2.69u  0.27s  1.38g
3425              (spills = 62 reloads = 124)
3426           TOTAL                                   682.48u  19.61s  13.99g
3427    
3428    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
3429           MLRISC ra                               4.96u  0.27s  2.72g
3430              (spills = 85 reloads = 170)
3431           TOTAL                                   2375.26u  57.21s  48.00g
3432    
3433    As you can see the old cps spill module suffers from some serious
3434    performance problem.  But since I cannot decipher the old code fully,
3435    instead of patching the problems up, I'm reimplementing it
3436    with a different algorithm.  The new code is more modular,
3437    smaller when compiled, and substantially faster
3438    (O(n log n) time and O(n) space).  Timing of the new spill module:
3439    
3440    4000  CPS 100 spill                           0.02u  0.00s  0.00g
3441          MLRISC ra                               0.25u  0.02s  0.15g
3442             (spills=1 reloads=3)
3443          TOTAL                                   7.74u  0.34s  1.62g
3444    
3445    Implementation details:
3446    
3447    As far as I can tell, the purpose of the CPS spill module is to make sure the
3448    number of live variables at any program point (the bandwidth)
3449    does not exceed a certain limit, which is determined by the
3450    size of the spill area.
3451    
3452    When the bandwidth is too large, we decrease the register pressure by
3453    packing live variables into spill records.  How we achieve this is
3454    completely different than what we did in the old code.
3455    
3456    First, there is something about the MLRiscGen code generator
3457    that we should be aware of:
3458    
3459    o MLRiscGen performs code motion!
3460    
3461       In particular, it will move floating point computations and
3462       address computations involving only the heap pointer to
3463       their use sites (if there is only a single use).
3464       What this means is that if we have a CPS record construction
3465       statement
3466    
3467           RECORD(k,vl,w,e)
3468    
3469       we should never count the new record address w as live if w
3470       has only one use (which is often the case).
3471    
3472       We should do something similar to floating point, but the transformation
3473       there is much more complex, so I won't deal with that.
3474    
3475    Secondly, there are now two new cps primops at our disposal:
3476    
3477     1. rawrecord of record_kind option
3478        This pure operator allocates some uninitialized storage from the heap.
3479        There are two forms:
3480    
3481         rawrecord NONE [INT n]  allocates a tagless record of length n
3482         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
3483                                     and initializes the tag.
3484    
3485     2. rawupdate of cty
3486          rawupdate cty (v,i,x)
3487          Assigns to x to the ith component of record v.
3488          The storelist is not updated.
3489    
3490    We use these new primops for both spilling and increment record construction.
3491    
3492     1. Spilling.
3493    
3494        This is implemented with a linear scan algorithm (but generalized
3495        to trees).  The algorithm will create a single spill record at the
3496        beginning of the cps function and use rawupdate to spill to it,
3497        and SELECT or SELp to reload from it.  So both spills and reloads
3498        are fine-grain operations.  In contrast, in the old algorithm
3499        "spills" have to be bundled together in records.
3500    
3501        Ideally, we should sink the spill record construction to where
3502        it is needed.  We can even split the spill record into multiple ones
3503        at the places where they are needed.  But CPS is not a good
3504        representation for global code motion, so I'll keep it simple and
3505        am not attempting this.
3506    
3507     2. Incremental record construction (aka record splitting).
3508    
3509        Long records with many component values which are simulatenously live
3510        (recall that single use record addresses are not considered to
3511         be live) are constructed with rawrecord and rawupdate.
3512        We allocate space on the heap with rawrecord first, then gradually
3513        fill it in with rawupdate.  This is the technique suggested to me
3514        by Matthias.
3515    
3516        Some restrictions on when this is applicable:
3517        1. It is not a VECTOR record.  The code generator currently does not handle
3518           this case. VECTOR record uses double indirection like arrays.
3519        2. All the record component values are defined in the same "basic block"
3520           as the record constructor.  This is to prevent speculative
3521           record construction.
3522    
3523    ----------------------------------------------------------------------
3524    Name: Allen Leung
3525    Date: 2002/02/22 01:02:00 EST
3526    Tag: leunga-20020222-mlrisc-tools
3527    
3528    Minor bug fixes in the parser and rewriter
3529    
3530    ----------------------------------------------------------------------
3531    Name: Allen Leung
3532    Date: 2002/02/21 20:20:00 EST
3533    Tag: leunga-20020221-peephole
3534    
3535    Regenerated the peephole files.  Some contained typos in the specification
3536    and some didn't compile because of pretty printing bugs in the old version
3537    of 'nowhere'.
3538    
3539    ----------------------------------------------------------------------
3540    Name: Allen Leung
3541    Date: 2002/02/19 20:20:00 EST
3542    Tag: leunga-20020219-mlrisc-tools
3543    Description:
3544    
3545       Minor bug fixes to the mlrisc-tools library:
3546    
3547       1.  Fixed up parsing colon suffixed keywords
3548       2.  Added the ability to shut the error messages up
3549       3.  Reimplemented the pretty printer and fixed up/improved
3550           the pretty printing of handle and -> types.
3551       4.  Fixed up generation of literal symbols in the nowhere tool.
3552       5.  Added some SML keywords to to sml.sty
3553    
3554    ----------------------------------------------------------------------
3555    Name: Matthias Blume
3556    Date: 2002/02/19 16:20:00 EST
3557    Tag: blume-20020219-cmffi
3558    Description:
3559    
3560    A wild mix of changes, some minor, some major:
3561    
3562    * All C FFI-related libraries are now anchored under $c:
3563        $/c.cm      --> $c/c.cm
3564        $/c-int.cm  --> $c/internals/c-int.cm
3565        $/memory.cm --> $c/memory/memory.cm
3566    
3567    * "make" tool (in CM) now treats its argument pathname slightly
3568      differently:
3569        1. If the native expansion is an absolute name, then before invoking
3570           the "make" command on it, CM will apply OS.Path.mkRelative
3571           (with relativeTo = OS.FileSys.getDir()) to it.
3572        2. The argument will be passed through to subsequent phases of CM
3573           processing without "going native".  In particular, if the argument
3574           was an anchored path, then "make" will not lose track of that anchor.
3575    
3576    * Compiler backends now "know" their respective C calling conventions
3577      instead of having to be told about it by ml-nlffigen.  This relieves
3578      ml-nlffigen from one of its burdens.
3579    
3580    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
3581    
3582    * Export C_DEBUG and C_Debug from $c/c.cm.
3583    
3584    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
3585      subtyping relationship between incomplete pointers and their complete
3586      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
3587      with the convention of instantiating 'o with ('t, 'c) obj whenever
3588      the pointer target type is complete.  In the incomplete case, 'o
3589      will be instantiated with some "'c iobj" -- a type obtained by
3590      using one of the functors PointerToIncompleteType or PointerToCompleteType.
3591    
3592      Operations that work on both incomplete and complete pointer types are
3593      typed as taking an 'o ptr while operations that require the target to
3594      be known are typed as taking some ('t, 'c) obj ptr.
3595    
3596      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
3597      where void is an eqtype without any values.  This makes it possible
3598      to work on voidptr values using functions meant to operate on light
3599      incomplete pointers.
3600    
3601    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
3602      been vastly simplified.
3603    
3604    ----------------------------------------------------------------------
3605    Name: Matthias Blume
3606    Date: 2002/02/19 10:48:00 EST
3607    Tag: blume-20020219-pqfix
3608    Description:
3609    
3610    Applied Chris Okasaki's bug fix for priority queues.
3611    
3612    ----------------------------------------------------------------------
3613    Name: Matthias Blume
3614    Date: 2002/02/15 17:05:00
3615    Tag: Release_110_39
3616    Description:
3617    
3618    Last-minute retagging is becoming a tradition... :-(
3619    
3620    This is the working release 110.39.
3621    
3622    ----------------------------------------------------------------------
3623    Name: Matthias Blume
3624    Date: 2002/02/15 16:00:00 EST
3625    Tag: Release_110_39-orig
3626    Description:
3627    
3628    Working release 110.39.  New bootfiles.
3629    
3630    (Update: There was a small bug in the installer so it wouldn't work
3631    with all shells.  So I retagged. -Matthias)
3632    
3633    ----------------------------------------------------------------------
3634    Name: Matthias Blume
3635    Date: 2002/02/15 14:17:00 EST
3636    Tag: blume-20020215-showbindings
3637    Description:
3638    
3639    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
3640    the latter can be useful for exploring what bindings are available at
3641    the interactive prompt.  (The first function returns only the list
3642    of symbols that are really bound, the second prints those but also the
3643    ones that CM's autoloading mechanism knows about.)
3644    
3645    ----------------------------------------------------------------------
3646    Name: Matthias Blume
3647    Date: 2002/02/15 12:08:00 EST
3648    Tag: blume-20020215-iptrs
3649    Description:
3650    
3651    Two improvements to ml-nlffigen:
3652    
3653      1. Write files only if they do not exist or if their current contents
3654         do not coincide with what's being written.  (That is, avoid messing
3655         with the time stamps unless absolutely necessary.)
3656    
3657      2. Implement a "repository" mechanism for generated files related
3658         to "incomplete pointer types".   See the README file for details.
3659    
3660    ----------------------------------------------------------------------
3661    Name: Matthias Blume
3662    Date: 2002/02/14 11:50:00 EST
3663    Tag: blume-20020214-quote
3664    Description:
3665    
3666    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
3667    because of the new and improved tag generation scheme.  (Thanks to Allen
3668    Leung for pointing it out.)
3669    
3670    ----------------------------------------------------------------------
3671    Name: Lal George
3672    Date: 2002/02/14 09:55:27 EST 2002
3673    Tag: george-20020214-isabelle-bug
3674    Description:
3675    
3676    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
3677    of Isabelle on the x86.
3678    
3679    From Allen:
3680    -----------
3681     I've found the problem:
3682    
3683         in ra-core.sml, I use the counter "blocked" to keep track of the
3684         true number of elements in the freeze queue.  When the counter goes
3685         to zero, I skip examining the queue.  But I've messed up the
3686         bookkeeping in combine():
3687    
3688             else ();
3689             case !ucol of
3690               PSEUDO => (if !cntv > 0 then
3691                     (if !cntu > 0 then blocked := !blocked - 1 else ();
3692                                        ^^^^^^^^^^^^^^^^^^^^^^^
3693                      moveu := mergeMoveList(!movev, !moveu)
3694                     )
3695                  else ();
3696    
3697         combine() is called to coalesce two nodes u and v.
3698         I think I was thinking that if the move counts of u and v are both
3699         greater than zero then after they are coalesced then one node is
3700         removed from the freeze queue.  Apparently I was thinking that
3701         both u and v are of low degree, but that's clearly not necessarily true.
3702    
3703    
3704    02/12/2002:
3705        Here's the patch.  HOL now compiles.
3706    
3707        I don't know how this impact on performance (compile
3708        time or runtime).  This bug caused the RA (especially on the x86)
3709        to go thru the potential spill phase when there are still nodes on the
3710        freeze queue.
3711    
3712    
3713    
3714    
3715    ----------------------------------------------------------------------
3716    Name: Matthias Blume
3717    Date: 2002/02/13 22:40:00 EST
3718    Tag: blume-20020213-fptr-rtti
3719    Description:
3720    
3721    Fixed a bug in ml-nlffigen that was introduced with one of the previous
3722    updates.
3723    
3724    ----------------------------------------------------------------------
3725    Name: Matthias Blume
3726    Date: 2002/02/13 16:41:00 EST
3727    Tag: blume-20020213-cmlpq
3728    Description:
3729    
3730    Added new priority queue export symbols (which have just been added to
3731    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
3732    not compile and the installer would choke.)
3733    
3734    ----------------------------------------------------------------------
3735    Name: Matthias Blume
3736    Date: 2002/02/13 16:15:00 EST
3737    Tag: blume-20020213-various
3738    Description:
3739    
3740    1. More tweaks to ml-nlffigen:
3741    
3742       - better internal datastructures (resulting in slight speedup)
3743       - "-match" option requires exact match
3744       - "localized" gensym counters (untagged structs/unions nested within
3745         other structs/unions or within typedefs get a fresh counter; their
3746         tag will be prefixed by a concatenation of their parents' tags)
3747       - bug fixes (related to calculation of transitive closure of types
3748         to be included in the output)
3749    
3750    2. Minor Basis updates:
3751    
3752       - added implementations for List.collate and Option.app
3753    
3754    ----------------------------------------------------------------------
3755    Name: Matthias Blume
3756    Date: 2002/02/11 15:55:00 EST
3757    Tag: blume-20020211-gensym
3758    Description:
3759    
3760    Added a "-gensym" option to command line of ml-nlffigen.  This can be
3761    used to specify a "stem" -- a string that is inserted in all "gensym'd"
3762    names (ML structure names that correspond to unnamed C structs, unions,
3763    and enums), so that separate runs of ml-nlffigen do not clash.
3764    
3765    ----------------------------------------------------------------------
3766    Name: Matthias Blume
3767    Date: 2002/02/11 12:05:00 EST
3768    Tag: blume-20020211-gensml
3769    Description:
3770    
3771    A quick fix for a problem with GenSML (in the pgraph-util library):
3772    Make generation of toplevel "local" optional.  (Strictly speaking,
3773    signature definitions within "local" are not legal SML.)
3774    
3775    Other than that: updates to INSTALL and cm/TODO.
3776    
3777    ----------------------------------------------------------------------
3778    Name: Matthias Blume
3779    Date: 2002/02/08 15:00:00 EST
3780    Tag: blume-20020208-uniquepid
3781    Description:
3782    
3783    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
3784    
3785    1. The installer (config/install.sh) has gotten smarter:
3786    
3787         - Configuration options are a bit easier to specify now
3788           (in config/targets).
3789         - Bug in recognizing .tar.bz2 files fixed.
3790         - Installer automatically resolves dependencies between
3791           configuration options (e.g., if you ask for eXene, you will
3792           also get cml -- regardless whether you asked for it or not).
3793         - Installer can run in "quieter mode" by setting the environment
3794           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
3795           "completely silent", though.
3796         - Build HashCons library as part of smlnj-lib.
3797    
3798    2. A new scheme for assigning persistent identifiers to compilation
3799       units (and, by extension, to types etc.) has been put into place.
3800       This fixes a long-standing bug where types and even dynamic values
3801       can get internally confused, thereby compromising type safety
3802       (abstraction) and dynamic correctness.  See
3803    
3804         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
3805    
3806       for an example of how things could go wrong until now.
3807    
3808       The downside of the new scheme is that pids are not quite as
3809       persistent as they used to be: CM will generate a fresh pid
3810       for every compilation unit that it thinks it sees for the first
3811       time.  That means that if you compile starting from a clean, fresh
3812       source tree at two different times, you end up with different
3813       binaries.
3814    
3815       Cutoff recompilation, however, has not been compromised because
3816       CM keeps pid information in special caches between runs.
3817    
3818    ----------------------------------------------------------------------
3819    Name: Lal George
3820    Date: 2002/02/07 15:34:13 EST 2002
3821    Tag: <none>
3822    Description:
3823    
3824    Compilers that generate assembly code may produce  global labels
3825    whose value is resolved at link time. The various peephole optimization
3826    modules did not take this in account.
3827    
3828    TODO. The Labels.addrOf function should really return an option
3829    type so that clients are forced to deal with this issue, rather
3830    than an exception being raised.
3831    
3832    ----------------------------------------------------------------------
3833    Name: Lal George
3834    Date: 2002/02/06 13:55:02 EST
3835    Tag: george-20020206-ra-breakup
3836    Description:
3837    
3838    1. A bug fix from Allen.
3839    
3840        A typo causes extra fstp %st(0)'s to be generated at compensation
3841        edges, which might cause stack underflow traps at runtime.  This
3842        occurs in fft where there are extraneous fstps right before the 'into'
3843        trap instruction (in this case they are harmless since none of the
3844        integers overflow.)
3845    
3846    2. Pulled out various utility modules that were embedded in the modules
3847       of the register allocator. I need these modules for other purposes, but
3848       they are not complete enough to put into a library (just yet).
3849    ----------------------------------------------------------------------
3850    Name: Matthias Blume
3851    Date: 2002/01/31 16:05:00 EST
3852    Tag: blume-20020131-sparc-ccalls
3853    Description:
3854    
3855    1. C-calls on Sparc needlessly allocated a huge chunk (96 bytes)
3856       of extra stack space by mistake.  Fixed.
3857