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