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 1044, Mon Jan 28 21:36:08 2002 UTC revision 1836, Fri Oct 14 15:16:07 2005 UTC
# Line 13  Line 13 
13  Description:  Description:
14    
15  ----------------------------------------------------------------------  ----------------------------------------------------------------------
16    Name: Allen Leung (leunga (at) reservoir (dot) com)
17    Date: 2005/10/14 11:15:00 EST
18    Tag: leunga-20051014-x86-asm
19    Description:
20    
21    Removed some debugging code in file x86Asm.sml.
22    The function emit_operand was printing out debugging output.
23    
24    ----------------------------------------------------------------------
25    Name: John Reppy
26    Date: 2005/07/27
27    Tag:
28    Description:
29    
30    Fixed ml-lex to recognize "\r" as representing carriage return.
31    
32    ----------------------------------------------------------------------
33    Name: John Reppy
34    Date: 2005/07/27
35    Tag:
36    Description:
37    
38    Fixed ml-yacc to work on files with non-native end-of-line
39    encodings (e.g., Windows text file on a Unix system).
40    
41    ----------------------------------------------------------------------
42    Name: John Reppy
43    Date: 2005/07/20
44    Tag:
45    Description:
46    
47    Added changes from Dominic Evans (oldmanuk (at) gmail (dot) com)
48    to support HPUX 11.
49    
50    ----------------------------------------------------------------------
51    Name: John Reppy
52    Date: 2005/07/06
53    Tag:
54    Description:
55    
56    Changes to the SML/NJ library.  See smlnj-lib/CHANGES for details.
57    
58    ----------------------------------------------------------------------
59    Name: Matthias Blume (blume (at) tti - c (dot) org)
60    Date: 2005/07/06 00:45:00 CDT
61    Tag: blume-20050706-slice-copy
62    Description:
63    
64    Fixed reversed logic for deciding whether to "copy up" or "copy down"
65    in *-array-slice.sml.
66    
67    ----------------------------------------------------------------------
68    Name: Allen Leung (leunga (at) reservoir (dot) com)
69    Date: 2005/05/31 17:00:00 EST
70    Tag: leunga-20050531-cygwin-fault-2
71    Description:
72    
73    A typo in the cygwin code fixed.
74    
75    ----------------------------------------------------------------------
76    Name: Allen Leung (leunga (at) reservoir (dot) com)
77    Date: 2005/05/31 16:47:00 EST
78    Tag: leunga-20050531-cygwin-fault
79    Description:
80    
81    Updated Cygwin's fault/signal handling to match the Windows version.
82    Updated the export list.
83    
84    ----------------------------------------------------------------------
85    Name: Matthias Blume (blume (at) tti - c (dot) org)
86    Date: 2005/05/18 13:35:00 CDT
87    Tag: Release_110_54
88    Description:
89    
90    New working version (110.54).  NEW BOOTFILES!
91    
92    ----------------------------------------------------------------------
93    Name: Matthias Blume (blume (at) tti - c (dot) org)
94    Date: 2005/05/18 11:58:00 CDT
95    Tag: blume-20050518-installer
96    Description:
97    
98    Added support scripts for Mac OS X PackageMaker and modified
99    config/install.sh so that it supports re-dumping a heap image
100    after customization.
101    
102    ----------------------------------------------------------------------
103    Name: Matthias Blume (blume (at) tti - c (dot) org)
104    Date: 2005/05/18 10:55:00 CDT
105    Tag: blume-20050518-realdiv-noovld
106    Description:
107    
108    Un-overloaded / to work around bug in overloading resolution code.
109    
110    ----------------------------------------------------------------------
111    Name: Matthias Blume (blume (at) tti - c (dot) org)
112    Date: 2005/05/16 23:50:00 CDT
113    Tag: blume-20050516-redump-heap
114    Description:
115    
116    Added mechanism for re-creating a heap file for the interactive system
117    after configuration variables have been changed.
118    
119       CM.redump_heap : string -> unit
120    
121    This is much like SMLofNJ.exportML, but starting from the resulting
122    heap does not return to the caller of CM.redump_heap but
123    restarts the interactive system from scratch.  The original call of
124    CM.redump_heap does not return but ends the interactive session.
125    Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
126    
127    Internally, redump_heap winds the dynamic execution context back to
128    the point where the original heap image was created and re-executes
129    the heap image generation code in the boot code.
130    
131    ----------------------------------------------------------------------
132    Name: Matthias Blume (blume (at) tti - c (dot) org)
133    Date: 2005/05/09 21:55:00 CDT
134    Tag: blume-20050509-word64
135    Description:
136    
137    Added a hack to the existing hack known as Word64 to make fromString
138    behave correctly.  I am still not sure whether Word64.scan will work
139    as specified with respect to the interaction of radix and prefix.
140    
141    ----------------------------------------------------------------------
142    Name: Allen Leung (leunga (at) reservoir (dot) com)
143    Date: 2005/05/04 11:50:00 EST
144    Tag: leunga-20050504-checkgc
145    Description:
146    
147    Added a gc protocol checking phase.  This phase is enabled with
148    the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
149    
150    ----------------------------------------------------------------------
151    Name: Matthias Blume (blume (at) tti - c (dot) org)
152    Date: 2005/05/04 10:45:00 CDT
153    Tag: blume-20050504-intinf
154    Description:
155    
156    Fixed a bug in the implementation of div and mod for IntInf.
157    Thanks to Neophytos Michael for reporting the problem.
158    
159    ----------------------------------------------------------------------
160    Name: Matthias Blume (blume (at) tti - c (dot) org)
161    Date: 2005/05/04 10:35:00 CDT
162    Tag: blume-20050504-join
163    Description:
164    
165    Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
166    
167    ----------------------------------------------------------------------
168    Name: Matthias Blume (blume (at) tti - c (dot) org)
169    Date: 2005/02/28 23:40:00 CST
170    Tag: blume-20050228-mVar
171    Description:
172    
173    Fixed serious bug (brown paper bag variety) in new implementation of
174    structure Atom in CML.  (I had accidentally used a mailbox instead of
175    an mvar, leaving the door open for races.)
176    
177    ----------------------------------------------------------------------
178    Name: Matthias Blume (blume (at) tti - c (dot) org)
179    Date: 2005/02/25 15:00:00 CST
180    Tag: Release_110_53
181    Description:
182    
183    New working version (110.53).  NEW BOOTFILES!
184    
185    ----------------------------------------------------------------------
186    Name: Matthias Blume (blume (at) tti - c (dot) org)
187    Date: 2005/02/25 14:50:00 CST
188    Tag: blume-20050225-susp
189    Description:
190    
191    Brought back SMLofNJ.Susp.  The underlying suspension type is the one
192    implemented in Core -- which means that it is the same as the one used
193    by the lazy extension.
194    
195    ----------------------------------------------------------------------
196    Name: Matthias Blume (blume (at) tti - c (dot) org)
197    Date: 2005/02/24 16:50:00 CST
198    Tag: blume-20050224-cml-atom
199    Description:
200    
201    Simpler and at the same time more general implementation of structure
202    Atom in CML.
203    
204    ----------------------------------------------------------------------
205    Name: Matthias Blume (blume (at) tti - c (dot) org)
206    Date: 2005/02/15 17:35:00 CST
207    Tag: blume-20050215-tools
208    Description:
209    
210    Created new "tools" directory under "src" and moved "TraceDebugProf"
211    there.
212    
213    ----------------------------------------------------------------------
214    Name: Matthias Blume (blume (at) tti - c (dot) org)
215    Date: 2005/02/10 17:55:00 CST
216    Tag: blume-20050210-longlong
217    Description:
218    
219    Implemented "long long" arguments and results for NLFFI.  (Only the
220    PPC/MacOS implementation is complete, the other backends still need to
221    be updated.)
222    
223    ----------------------------------------------------------------------
224    Name: Matthias Blume (blume (at) tti - c (dot) org)
225    Date: 2005/01/24 17:40:00 CST
226    Tag: blume-20050124-mlyacc
227    Description:
228    
229    Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
230    problem with certain "as" patterns which previously got rendered
231    using incorrect syntax.
232    
233    ----------------------------------------------------------------------
234    Name: Matthias Blume (blume (at) tti - c (dot) org)
235    Date: 2005/01/18 12:00:00 CST
236    Tag: blume-20050118-profile
237    Description:
238    
239    Made time profiling code (interrupt handler) in runtime system aware
240    of new array representation.
241    
242    ----------------------------------------------------------------------
243    Name: Matthias Blume (blume (at) tti - c (dot) org)
244    Date: 2005/01/14 18:00:00 CST
245    Tag: blume-20050114-heap2exec
246    Description:
247    
248    Implemented new (but still experimental) heap2exec facility.  This is
249    tested under Mac OS X and should work under Linux (will test shortly).
250    It will probably also work on the Sparc (will test some time later).
251    
252      - removed old "HACKED_STANDALONE" hack from runtime
253    
254    To be able to test this, uncomment the request for "heap2asm" in
255    config/targets prior to installation.  (Notice that this is different
256    from "heap2exec" mentioned below.  Not a typo.)
257    
258    To perform an actual test, run the command
259    
260       $ bin/heap2exec heapfile execfile
261    
262    (You can put heap2exec on your shell's path.)
263    
264    For example, run
265    
266      $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
267    
268    This will create a standalone executable called "mly" which you can
269    then invoke directly as a command.
270    
271    ----------------------------------------------------------------------
272    Name: Matthias Blume (blume (at) tti - c (dot) org)
273    Date: 2005/01/07 11:44:00 CST
274    Tag: blume-20050107-mlstring
275    Description:
276    
277    fixed off-by-one error in ML_STRING macro (globals.c)
278    
279    ----------------------------------------------------------------------
280    Name: Matthias Blume (blume (at) tti - c (dot) org)
281    Date: 2004/12/23 18:00:00 CST
282    Tag: blume-20041223-santa
283    Description:
284    
285    Made ml-build script "smarter" (but only very little).
286    
287    ----------------------------------------------------------------------
288    Name: Matthias Blume (blume (at) tti - c (dot) org)
289    Date: 2004/12/21 15:05:00 CST
290    Tag: blume-20041221-longlong
291    Description:
292    
293    * Implemented access to signed and unsigned long long data in NLFFI.
294    (The parameter-passing part of the picture has not complete. But
295    data structure access seems to work.)
296    
297    * Fixed CM's incorrect assumption that the PPC is little-endian.
298      (On the Mac, it is big-endian.  And that's currently our only
299      PPC platform.)
300    
301    ----------------------------------------------------------------------
302    Name: Matthias Blume (blume (at) tti - c (dot) org)
303    Date: 2004/12/21 12:50:00 CST
304    Tag: blume-20041221-memory
305    Description:
306    
307    Some cleanup in the $c/memory.cm library: separated some concerns by
308    moving allocation code and memory access code each into their own
309    files.
310    
311    ----------------------------------------------------------------------
312    Name: Allen Leung (leunga (at) reservoir (dot) com)
313    Date: 2004/12/17 16:12:00 EST
314    Tag: leunga-20041217-cygwin-smlnj-home
315    Description:
316    
317        The Unix I/O library of SML/NJ on cygwin does not understand
318    Windows style pathname, so problems arise when SMLNJ_HOME is set
319    to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
320    to a POSIX pathname on cygwin.
321    
322    ----------------------------------------------------------------------
323    Name: Matthias Blume (blume (at) tti - c (dot) org)
324    Date: 2004/12/16 13:35:00 CST
325    Tag: Release_110_52
326    Description:
327    
328    Last-minute changes incorporated into 110.52.  Release tag moved.
329    
330    The changes:
331       - HashString.hashString' -> HashString.hashSubstring
332       - bug fix in UnivariateStats
333    
334    ----------------------------------------------------------------------
335    Name: Matthias Blume (blume (at) tti - c (dot) org)
336    Date: 2004/12/15 23:40:00 CST
337    Tag: blume-20041215-hashSubstring
338    Description:
339    
340    - HashString.hashString' -> HashString.hashSubstring
341    - corresponding changes in atom.sml
342    - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
343      added some comments
344    
345    ----------------------------------------------------------------------
346    Name: Matthias Blume (blume (at) tti - c (dot) org)
347    Date: 2004/12/15 15:30:00 CST
348    Tag: (Release_110_52)
349    Description:
350    
351    New working version (110.52).  NEW BOOTFILES!
352    
353    ----------------------------------------------------------------------
354    Name: Matthias Blume (blume (at) tti - c (dot) org)
355    Date: 2004/12/15 12:45:00 CST
356    Tag: blume-20041215-spaces
357    Description:
358    
359    More on the space problem (this time for Win32).
360    
361    ----------------------------------------------------------------------
362    Name: Matthias Blume (blume (at) tti - c (dot) org)
363    Date: 2004/12/14 17:30:00 CST
364    Tag: blume-20041214-spaces
365    Description:
366    
367    Hacked some of the scripts (in particular: the installer) to cope with
368    spaces in filenames a bit better.  But beware: the current "solution"
369    is likely still full of bugs and inherently incomplete.  (We need to
370    do away with those shell scripts for a comprehensive solution.)
371    
372    ----------------------------------------------------------------------
373    Name: Matthias Blume (blume (at) tti - c (dot) org)
374    Date: 2004/12/13 14:45:00 CST
375    Tag: blume-20041213-ml-makedepend
376    Description:
377    
378    Fixed bug in code for ml-makedepend.
379    
380    ----------------------------------------------------------------------
381    Name: Matthias Blume (blume (at) tti - c (dot) org)
382    Date: 2004/12/09 16:30:00 CST
383    Tag: blume-20041209-statistics
384    Description:
385    
386    Added two simple but potentially useful statistics modules to SML/NJ Library.
387    (See CHANGES file there.)
388    
389    ----------------------------------------------------------------------
390    Name: Matthias Blume (blume (at) tti - c (dot) org)
391    Date: 2004/12/01 16:50:00 CST
392    Tag: blume-20041201-atom
393    Description:
394    
395    smlnj-lib:
396    
397      Added function HashString.hashString' for substrings.
398      Hand-inlined CharVector.foldl into HashString (for speed).
399      Modified implementation of structure Atom to avoid extracting
400      strings from substrings unless necessary.
401      (Also see CHANGES file for smlnj-lib.)
402    
403    ----------------------------------------------------------------------
404    Name: Matthias Blume (blume (at) tti - c (dot) org)
405    Date: 2004/11/24 22:15:00 CST
406    Tag: blume-20041124-cml
407    Description:
408    
409    Made sure CML compiles when Position = Int64.
410    
411    ----------------------------------------------------------------------
412    Name: Matthias Blume (blume (at) tti - c (dot) org)
413    Date: 2004/11/24 14:30:00 CST
414    Tag: blume-20041124-position
415    Description:
416    
417    The compiler can now be compiled in a mode that makes structure Position
418    equal to Int64.  The default, however, is unchanged (Position = Int31)
419    for the time being.
420    
421    To enable 64-bit positions, use the following procedure:
422    
423    1. Start sml
424    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
425    3. Type
426          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
427    4. Run CMB.make() as usual.
428    
429    This is barely tested.  The only test so far was a little SML program
430    counting the number of characters in an 8-gigabyte file by
431    reading it character-by-character.  That test was successful.
432    
433    In support of 64-bit positions, a number of new functions have been
434    added to the runtime system.
435    
436    ----------------------------------------------------------------------
437    Name: Matthias Blume (blume (at) tti - c (dot) org)
438    Date: 2004/11/23 14:45:00 CST
439    Tag: blume-20041123-useFile
440    Description:
441    
442    Fixed a problem with unhelpful error messages related to problems with
443    .cm- or .sml files that appear as part of the sml command line.
444    
445    ----------------------------------------------------------------------
446    Name: Matthias Blume (blume (at) tti - c (dot) org)
447    Date: 2004/11/18 15:40:00 CST
448    Tag: Release_110_51
449    Description:
450    
451    New working version (110.51).  NEW BOOTFILES!
452    
453    ----------------------------------------------------------------------
454    Name: Matthias Blume (blume (at) tti - c (dot) org)
455    Date: 2004/11/18 15:35:00 CST
456    Tag: <none>
457    Description:
458    
459    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
460    Joemann).
461    
462    ----------------------------------------------------------------------
463    Name: Allen Leung (leunga (at) reservoir (dot) com)
464    Date: 2004/11/17 16:05:21 EST 2004
465    Tag: leunga-20041117-mlrisc-live-kill
466    Description:
467    
468       Added support for MLTree constructs LIVE and KILL
469    to all the architectures.
470    
471    ----------------------------------------------------------------------
472    Name: Matthias Blume (blume (at) tti - c (dot) org)
473    Date: 2004/11/13 00:20:00 CST
474    Tag: blume-20041113-versiontool
475    Description:
476    
477    - Stripped down the versiontool: It now only handles the version number.
478      The date string is generated at bootstrap time (during makeml).
479    
480    - In a previous commit, fixed a minor issue with how polyequal is being
481      translated.  In particular, the code now "looks through" abstractions.
482      This results in slightly fewer polyEqual warnings and hopefully slightly
483      more efficient code.  Important examples for where this matters are
484      the new int64 and word64 types.
485    
486    ----------------------------------------------------------------------
487    Name: Matthias Blume (blume (at) tti - c (dot) org)
488    Date: 2004/11/12 00:30:00 CST
489    Tag: blume-20041112-int64
490    Description:
491    
492    Structure Int64 fully hooked in.  (The implementation is not very
493    efficient, though.)
494    
495    ----------------------------------------------------------------------
496    Name: Matthias Blume (blume (at) tti - c (dot) org)
497    Date: 2004/11/11 17:30:00 CST
498    Tag: blume-20041111-more64
499    Description:
500    
501    All the pieces of Word64 are now there, with the exception of the
502    conversions from and to LargeWord. (Eventually these need to be identities,
503    but for the time being they don't even make sense because LargeWord is
504    32-bit wide.)
505    
506    Also started to add similar support for Int64, but major pieces of that
507    are still missing.
508    
509    ----------------------------------------------------------------------
510    Name: Matthias Blume (blume (at) tti - c (dot) org)
511    Date: 2004/11/11 00:15:00 CST
512    Tag: blume-20041111-word64
513    Description:
514    
515    Structure Word64 is now (almost) complete, word literals and patterns
516    seem to work.  There are a few odd pieces missing.  In particular,
517    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
518    at the moment.
519    
520    Making Word64 official would mean that LargeWord becomes Word64.  But
521    this requires extreme care because most word-word conversions have to
522    go through LargeWord, so making a mistake means loss of efficiency or
523    worse.  Eventually there will be a solution similar to (but actually
524    simpler than) what I did with IntInf.
525    
526    ----------------------------------------------------------------------
527    Name: Matthias Blume (blume (at) tti - c (dot) org)
528    Date: 2004/11/10 18:12:00 CST
529    Tag: blume-20041110-64bit
530    Description:
531    
532    More 64-bit hacking (but still not even half-way there yet).
533    Also, some assorted improvements to the handling of 8-bit words.
534    
535    ----------------------------------------------------------------------
536    Name: Matthias Blume (blume (at) tti - c (dot) org)
537    Date: 2004/11/09 17:50:00 CST
538    Tag: <none>
539    Description:
540    
541    Started adding some infrastructure for supporting 64-bit int- and
542    word-types.  (Still in its very early stages.)
543    
544    ----------------------------------------------------------------------
545    Name: Matthias Blume (blume (at) tti - c (dot) org)
546    Date: 2004/10/28 10:45:00 CDT
547    Tag: Release_110_50
548    Description:
549    
550    New working version (110.50).  NEW BOOTFILES!
551    
552                 =====================
553    
554    Also:
555    
556      - Changend config/srcarchiveurl from a file just
557        containing the URL string into a file containing
558        shell script code.  The code has access to the $VERSION variable.
559      - Made corresponding changes to config/install.sh and config/unpack.
560      - Default contents of config/srcarchiveurl uses $VERSION and
561        normally does not have to be edited to reflect a version change.
562    
563      (As a result, a version change can be done by just editing
564       config/version, the rest is now automatic.)
565    
566    ----------------------------------------------------------------------
567    Name: Matthias Blume (blume (at) tti - c (dot) org)
568    Date: 2004/10/27 17:50:00 CDT
569    Tag: blume-20041027-btrace-msg
570    Description:
571    
572    BackTrace.monitor now also reports the source of the exception that
573    triggered the trace.
574    
575    ----------------------------------------------------------------------
576    Name: Matthias Blume (blume (at) tti - c (dot) org)
577    Date: 2004/10/27 17:20:00 CDT
578    Tag: blume-20041027-x86-c-calls
579    Description:
580    
581    This is the HISTORY entry for two earlier commits, both concerning
582    the x86 c-calls code in MLRISC:
583    
584      - added a missing LOAD in the code that deals with struct arguments
585      - made sure the caller does not add the wrong number of bytes to the
586        stack pointer after a call of a function returning a struct
587        (the callee already pops the implicit argument which points to
588         the space reserved for the result)
589    
590    ----------------------------------------------------------------------
591    Name: Allen Leung (leunga (at) reservoir (dot) com)
592    Date: 2004/10/24 14:00:00 EST
593    Tag: leunga-20041024-x86-gas-fucomip
594    Description:
595    
596       John discovered a bug in the syntax of fucomip.
597       The opcodes FU?COMIP? have been changed to
598    
599           fu?comip? %st(i), %st
600    
601    ----------------------------------------------------------------------
602    Name: Matthias Blume (blume (at) tti - c (dot) org)
603    Date: 2004/10/20 15:06:00 CDT
604    Tag: blume-20041020-standalone-backtrace
605    Description:
606    
607    Added a mechanism for getting back-trace information from standalone
608    programs.  Here is how it works:
609    
610       1. The part of the program from which you want to get backtrace
611          information (usually the whole program) should be wrapped with
612          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
613          main program could be modified from something like
614    
615             fun main (pgm, args) = ...
616    
617         to
618    
619            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
620    
621       2. To be able to access BackTrace.monitor, you have to add
622          library $smlnj-tdp/plugins.cm to the .cm file that contains your
623          main function.
624    
625       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
626          CM might have created in the past for your project).
627    
628       4. Build the system using this command line:
629    
630             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
631                      myprog.cm MyProg.main myprog
632    
633          instead of the usual
634    
635             ml-build myprog.cm MyProg.main myprog
636    
637    I changed a library name:
638    
639       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
640    
641    New libraries:
642    
643       $smlnj-tdp/back-trace.cm
644          -- when loaded causes the back-trace plugin to be installed
645       $smlnj-tdp/coverage.cm
646          -- when loaded causes the coverage plugin to be installed
647    
648    ----------------------------------------------------------------------
649    Name: Matthias Blume (blume (at) tti - c (dot) org)
650    Date: 2004/10/18 16:45:00 CDT
651    Tag: blume-20041018-groupowner
652    Description:
653    
654    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
655    
656    Eliminated group owner specs from .cm files throughout the source tree.
657    
658    ----------------------------------------------------------------------
659    Name: Matthias Blume (blume (at) tti - c (dot) org)
660    Date: 2004/10/15 15:45:00 CDT
661    Tag: blume-20041015-coverage
662    Description:
663    
664    * Test coverage tool added!
665    * Further reorganization of tracing-, debugging-, and profiling support:
666    
667        - moved original BTImp -- now called BackTrace -- into a separate
668          library called $/trace-debug-profile.cm
669        - eliminated all mentions of BTrace from SMLofNJ.Internals
670        - only the instrumentation mechanism is now left in the compiler proper
671        - BackTrace module is a plugin which is NOT plugged in by default
672        - Coverage module is another such plugin
673    
674    To get the benefits of any of these plugin modules, the code in
675    question must be compiled with tdp instrumentation turned on.  This
676    can be done by setting SMLofNJ.Internals.TDP.mode to true.
677    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
678    
679    Plugins are selected at link time.  (Pre-compiled instrumented code
680    can be re-loaded with different plugins in effect.)  When an
681    instrumented module is linked, whatever plugins are at that time
682    enabled will come into effect for that module.
683    
684    To enable the back-trace plugin, load library $/trace-debug-profile.cm
685    and invoke BackTrace.install() (e.g., from the interactive prompt).
686    To enable the coverage plugin, load the same library and invoke
687    Coverage.install().
688    
689    Back-traces are generated automatically on uncaught exceptions and
690    when the code in question explicitly invokes BackTrace.trigger().
691    
692    Coverage (and execution frequency-) information must be queried
693    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
694    
695    ----------------------------------------------------------------------
696    Name: Matthias Blume (blume (at) tti - c (dot) org)
697    Date: 2004/10/14 17:40:00 CDT
698    Tag: blume-20041014-tdp-core
699    Description:
700    
701    Snapshot of a significant overhaul of how the trace/debug/profile support
702    is hooked into the system (specifically: Core and SMLofNJ.Internals).
703    
704    ----------------------------------------------------------------------
705    Name: Matthias Blume (blume (at) tti - c (dot) org)
706    Date: 2004/10/13 16:34:00 CDT
707    Tag: blume-20041013-tdp
708    Description:
709    
710    Some rationalization of names:
711    
712       structure BTrace -> structure TDPInstrument
713       etc.
714    
715    This is is preparation of using the original back-trace
716    instrumentation for other purposes.  "TDP" stands for
717    Trace/Debug/Profile.
718    
719    The control flag controlling whether instrumentation is on or off is now
720    registered under a different name, so instead of running sml as
721    
722     sml -Cinstrument.btrace-mode=true
723    
724    one has to say
725    
726     sml -Ctdp.instrument=true
727    
728    ----------------------------------------------------------------------
729    Name: Matthias Blume (blume (at) tti - c (dot) org)
730    Date: 2004/10/11 16:37:00 CDT
731    Tag: blume-20041011-regions
732    Description:
733    
734    Made some minor modifications to elabcore.sml to have source regions
735    be propagated more tightly -- resulting in better (i.e., smaller)
736    regions being reported in error- and debug messages.
737    
738    ----------------------------------------------------------------------
739    Name: Matthias Blume (blume (at) tti - c (dot) org)
740    Date: 2004/10/08 22:50:00 CDT
741    Tag: blume-20041008-cmkw
742    Description:
743    
744    Fixed handling of keywords in .cm files: After seeing "is" the lexer
745    treats subsequent occurrences of "group", "library", "source", "is",
746    "*", and "-" as ordinary identifiers rather than keywords.
747    
748    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
749    is supposed to accept a tool argument called "source", but this did
750    not work because of the clash with the keyword.
751    
752    ----------------------------------------------------------------------
753    Name: Matthias Blume (blume (at) tti - c (dot) org)
754    Date: 2004/10/07 16:00:00 CDT
755    Tag: blume-20041007-cleanup
756    Description:
757    
758    Assorted cleanup work:
759    
760      - got rid of intstrmap in favor of using the library's
761        hash table implementation
762      - threw out most of the pathnames stuff, as it was not used anyway
763      - simplified tokentable implementation
764      - fixed some minor spelling errors
765    
766    ----------------------------------------------------------------------
767    Name: Matthias Blume (blume (at) tti - c (dot) org)
768    Date: 2004/10/06 15:15:15 CDT
769    Tag: blume-20041006-handler
770    Description:
771    
772    Cleaned up the absyn to reflect the invariant that HANDLE always
773    carries a FNexp as part of the type definition.  This eliminates some
774    superfluous sanity checks at runtime down the road.
775    
776    Some minor cleanup of the btrace code.
777    
778    ----------------------------------------------------------------------
779    Name: Matthias Blume (blume (at) tti - c (dot) org)
780    Date: 2004/10/01 10:20:30 CDT
781    Tag: blume-20041001-slave
782    Description:
783    
784    Added hack to make slave mode work in the presence of the version
785    tool.  (Still, since the master does two passes over the code for
786    CMB.make, the release number gets bumped twice when slaves are
787    attached.  I don't know if this is worth fixing...)
788    
789    ----------------------------------------------------------------------
790    Name: Matthias Blume (blume (at) tti - c (dot) org)
791    Date: 2004/09/30 10:55:00 CDT
792    Tag: blume-20040930-version
793    Description:
794    
795    * Moved the "version" magic into its own little library under
796      src/system/smlnj/internal.  This avoids expensive reconstruction of
797      a stable src/compiler/core.cm.
798    
799    * At the same time, structure CompilerVersion is now known as
800      structure SMLNJVersion.
801    
802    * Arranged for the version tool to NOT kick in when rebuilding the system
803      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
804      Also, loading the versiontool does not work when rebuilding the system
805      because CM is not properly initialized at that time.
806    
807    ----------------------------------------------------------------------
808    Name: Matthias Blume (blume (at) tti - c (dot) org)
809    Date: 2004/09/29 14:00:00 CDT
810    Tag: blume-20040929-autoversion
811    Description:
812    
813    Implemented some CM magic to have
814        file src/compiler/TopLevel/main/version.sml
815    generated automagically.
816    The version is taken from two files: config/version and config/release.
817    The first is expected to contain a two-part version number such as 110.49.
818    The second should contain a single number, but it may be missing.
819    
820    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
821    time the version tool is loaded (which is the first time you say CMB.make),
822    then the tool will increment the value stored in config/release every
823    time CMB.make is invoked.
824    
825    The binfile format is now insensitive to anything beyond the first
826    two components of a version number, so bumping the release does not render
827    binfiles incompatible.  Auto-bumping can be used to keep track of versions
828    during development without invalidating existing binfiles.
829    
830    In any case, every CMB.make updates the date information in version.sml.
831    (This is the date that is printed in the banner.)
832    
833    ----------------------------------------------------------------------
834    Name: Matthias Blume (blume (at) tti - c (dot) org)
835    Date: 2004/09/28 10:53:00 CDT
836    Tag: blume-20040928-controls
837    Description:
838    
839    Some cleanup of the controls code.
840    
841    ----------------------------------------------------------------------
842    Name: Matthias Blume (blume (at) tti - c (dot) org)
843    Date: 2004/09/27 22:08:00 CDT
844    Tag: blume-20040927-controls
845    Description:
846    
847    Added two pieces of functionality to the Controls interface:
848    
849    1.   val save'restore: 'a control -> unit -> unit
850    
851       grabs the current value of the control in stage 1 and restores it
852       in stage 2.
853    
854    2.   val set' : 'a control * 'a -> unit -> unit
855    
856       stores the given value into the control in stage 2 (i.e., delayed)
857       but does all error checking in stage 1.
858       (This is for string controls that need to do parse their argument --
859        something that might fail.  In some cases, notably in CM, one
860        already knows the intended argument but wants to delay the actual
861        assignment until a time when error recovery would be more difficult.)
862    
863    Changed the handling of controls in tool arguments to classes "sml" and
864    "lazysml":
865       - use Controls.save'restore as a more robust way of restoring the
866         old value (in particular: without having to re-parse the string)
867       - use controls to handle the "overload" keyword in the init group
868         (I believe this change actually fixes a long-standing obscure bug.)
869    
870    ----------------------------------------------------------------------
871    Name: Matthias Blume (blume (at) tti - c (dot) org)
872    Date: 2004/09/27 17:00:00 CDT
873    Tag: blume-20040927-lazysml
874    Description:
875    
876    Added a new tool class called "lazysml" to CM's tool chest.  The only
877    difference to "sml" is that compilation is done with Control.lazysml
878    set to true.  A source of class "lazysml" is automatically recognized
879    by a file name suffix of ".lml".
880    
881    In addition to the above feature, the original class "sml" now also
882    supports a tool argument "lazy" which has the same effect.  As a
883    result, the following three lines are equivalent:
884    
885        foo.sml : lazysml
886        foo.sml : sml (lazy)
887        foo.sml (lazy)
888    
889    The setting goes into effect both during parsing and during
890    compilation.  The original setting is restored right after parsing and
891    after compilation, respectively.
892    
893    In addition to all the above, there is also a general mechanism to set
894    ANY of the "controls" that are available at the command line via
895    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
896    apply as well. (In fact, "lazy" is implemented as a special case of
897    the general mechanism.)
898    
899    The .cm file syntax uses a new keyword tool argument called "with".
900    There are several ways of indicating the desired settings:
901    
902        foo.sml (with:parser.quotations=true)
903        foo.sml (with:(name:parser.quotations value:true))
904        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
905        foo.sml (with:(name1=value1 name2=value2 ...))
906        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
907    
908        etc.
909    
910    Another possible abbreviation is to leave out the =v or value:v part
911    if the name refers to a boolean control (in which case the value is
912    taken to be true).  Thus, one could get lazy sml also by saying:
913    
914        foo.sml (with:parser.lazy-keyword=true)
915        foo.sml (with:parser.lazy-keyword)
916        foo.sml (with:(name:parser.lazy-keyword value:true))
917        foo.sml (with:(name:parser.lazy-keyword))
918    
919    ----------------------------------------------------------------------
920    Name: Matthias Blume (blume (at) tti - c (dot) org)
921    Date: 2004/09/24 16:22:00 CDT
922    Tag: blume-20040924-ppc-long-branch
923    Description:
924    
925    Turned message about "emiting long form of branch" off by default.
926    Added a control flag to turn it back on when desired.
927    
928    ----------------------------------------------------------------------
929    Name: Matthias Blume (blume (at) tti - c (dot) org)
930    Date: 2004/09/24 16:05:00 CDT
931    Tag: blume-20040924-rounding
932    Description:
933    
934    Applied patch for setting rounding modes under Mac OS X.  Thanks to
935    Melissa O'Neill for providing the code!
936    
937    ----------------------------------------------------------------------
938    Name: Matthias Blume (blume (at) tti - c (dot) org)
939    Date: 2004/09/23 17:30:00 CDT
940    Tag: blume-20040923-envvars
941    Description:
942    
943    1. Changed definition of type ControlRegistry.registry_tree to
944       include control_info (i.e., the name of the controlling environment
945       variable).
946    
947    2. Added command-line flags -e and -E to print the names of environment
948       variables that can be used to control internal settings.
949       (This uses the new API mentioned in 1.)
950    
951    ----------------------------------------------------------------------
952    Name: Matthias Blume (blume (at) tti - c (dot) org)
953    Date: 2004/09/13 16:50:00 CDT
954    Tag: Release_110_49
955    Description:
956    
957    New working version (110.49).  NEW BOOTFILES!
958    
959    ----------------------------------------------------------------------
960    Name: Matthias Blume
961    Date: 2004/09/13 16:20:00 CDT
962    Tag: blume-20040913-config-mlrisc
963    Description:
964    
965    Put target "mlrisc" back into the default list.
966    (There is no harm in having it, and some users have expressed their
967    wish to have "mlrisc" included by default.)
968    
969    ----------------------------------------------------------------------
970    Name: John Reppy
971    Date: 2004/09/13
972    Tag: jhr-20040913-signals
973    Description:
974    
975    Fixed the signal masking code to properly nest mask/unmask operations
976    on a per-signal basis.
977    
978    ----------------------------------------------------------------------
979    Name: Matthias Blume (blume (at) tti - c (dot) org)
980    Date: 2004/09/08 13:20:00 CDT
981    Tag: blume-20040908-heap-magic
982    Description:
983    
984    Bumped the heap macig number to 0x09082004 to account for the changed
985    layout of the ML frame under MacOS X.
986    
987    ----------------------------------------------------------------------
988    Name: Allen Leung (leunga (at) reservoir (dot) com)
989    Date: 2004/09/03 11:26:00 EST
990    Tag: leunga-20040903-cygwin-install
991    Description:
992    
993       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
994    The Cygwin runtime is turned on by setting the environment
995    variable SMLNJ_CYGWIN_RUNTIME to 1.
996    
997    ----------------------------------------------------------------------
998    Name: Matthias Blume (blume (at) tti - c (dot) org)
999    Date: 2004/08/31 17:15:00 CDT
1000    Tag: blume-20040831-core
1001    Description:
1002    
1003    Added some exports to src/compiler/core.cm upon request by J. Joemann.
1004    
1005    ----------------------------------------------------------------------
1006    Name: Matthias Blume (blume (at) tti - c (dot) org)
1007    Date: 2004/08/30 17:55:00 CDT
1008    Tag: blume-20040830-installer
1009    Description:
1010    
1011    Upon request by Johannes Joemann:
1012    
1013    - improved ML code of installer to fall back to coping when renaming
1014      fails (i.e., when source and target are on different file systems);
1015      the code compiles but has yet to be tested in anger
1016    - removed mlrisc from list of default targets (config/targets)
1017    
1018    ----------------------------------------------------------------------
1019    Name: Matthias Blume (blume (at) tti - c (dot) org)
1020    Date: 2004/08/27 17:20:00 CDT
1021    Tag: blume-20040827-ptreql
1022    Description:
1023    
1024    Added ptreql primop to structure InlineT (upon request from Larry
1025    Paulson).
1026    
1027    ----------------------------------------------------------------------
1028    Name: Allen Leung (leunga (at) reservoir (dot) com)
1029    Date: 2004/08/15 21:21:00 EST
1030    Tag: leunga-110_48-udgraph
1031    Description:
1032    
1033    Another bug fix from Carl Hauser:
1034    
1035    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
1036    > 48c48
1037    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
1038    > e::L)
1039    > ---
1040    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
1041    > Without this, any deletion of an edge in an undirected graph does severe
1042    > violence to the graph.
1043    
1044    ----------------------------------------------------------------------
1045    Name: Allen Leung (leunga (at) reservoir (dot) com)
1046    Date: 2004/08/10 23:35:00 EST
1047    Tag: leunga-110_48-ppc
1048    Description:
1049    
1050       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
1051    
1052    ----------------------------------------------------------------------
1053    Name: Matthias Blume (blume (at) tti - c (dot) org)
1054    Date: 2004/08/10 12:00:00 CDT
1055    Tag: Release_110_48
1056    Description:
1057    
1058    New working version (110.48).  NEW BOOTFILES!
1059    
1060    ----------------------------------------------------------------------
1061    Name: Allen Leung (leunga (at) reservoir (dot) com)
1062    Date: 2004/08/09 12:21:00 EST
1063    Tag: leunga-110_47-dijsktra
1064    Description:
1065    
1066    Bug fix from Carl Hauser:
1067    
1068    single_source_shortest_paths in dijkstra.sml was observed to get wrong
1069    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1070    
1071    The problem is that following the expression A.update(dist,s,Num.zero)
1072    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1073    
1074    ----------------------------------------------------------------------
1075    Name: Matthias Blume (blume (at) tti - c (dot) org)
1076    Date: 2004/08/06 18:10:00 CDT
1077    Tag: blume-20040806-cmdline
1078    Description:
1079    
1080    Fiddled with handling of command-line options:
1081    
1082      * sml now quits after processing the command line
1083        if -H, -S, -h<n>, or -s<n> appears as the last
1084        command-line argument
1085      * a new option -q terminates the session when encountered on
1086        the command line; subsequent arguments will be ignored
1087      * bug fixes: short (erroneous) arguments are no longer ignored
1088        completely
1089    
1090    ----------------------------------------------------------------------
1091    Name: Allen Leung (leunga (at) reservoir (dot) com)
1092    Date: 2004/08/04 18:17:00 EST
1093    Tag: leunga-110_47-ppc-ibm-asm
1094    Description:
1095    
1096      - Added minimal IBM assembly syntax support for PowerPC.
1097    
1098      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1099        been altered in the runtime.  We need an automatic way to keep the file
1100        in sync.
1101    
1102    ----------------------------------------------------------------------
1103    Name: Matthias Blume (blume (at) tti - c (dot) org)
1104    Date: 2004/08/04 14:00:00 CDT
1105    Tag: Release_110_47
1106    Description:
1107    
1108    New working version (110.47).  NEW BOOTFILES!
1109    
1110    ----------------------------------------------------------------------
1111    Name: Matthias Blume (blume (at) tti - c (dot) org)
1112    Date: 2004/08/03 14:25:00 CDT
1113    Tag: blume-20040803-callingconv
1114    Description:
1115    
1116    Added low-level support for choosing C calling conventions by
1117    twiddling the type of rawccall.  (See
1118    src/compiler/Semant/types/cproto.sml for details.)
1119    
1120    ----------------------------------------------------------------------
1121    Name: Matthias Blume (blume (at) tti - c (dot) org)
1122    Date: 2004/08/02 15:55:00 CDT
1123    Tag: blume-20040802-backout
1124    Description:
1125    
1126    Backed out of change to win32-filesys.c.  The earlier patch to
1127    get_file_time caused CM to produce files with the wrong time stamp.
1128    
1129    ----------------------------------------------------------------------
1130    Name: Matthias Blume (blume (at) tti - c (dot) org)
1131    Date: 2004/08/02 14:45:00 CDT
1132    Tag: blume-20040802-nlffi-win32
1133    Description:
1134    
1135    Added NLFFI support for Win32, adapted from a patch provided by David
1136    Hansel.  This is currently completely untested.  Also, the issue
1137    concerning stdcall vs. ccall is still unresolved.
1138    
1139    ----------------------------------------------------------------------
1140    Name: Matthias Blume (blume (at) tti - c (dot) org)
1141    Date: 2004/07/30 17:55:00 CDT
1142    Tag: blume-20040730-various
1143    Description:
1144    
1145    Gearing up towards 110.47...
1146    
1147    - various minor bugfixes to ml-nlffigen
1148    - a beginning of a manual for nlffi
1149    
1150    - eliminated 'export name=value' in config/install.sh as this does
1151      not work with certain versions of /bin/sh
1152      (Thanks to David King at Motorola for catching this.)
1153    
1154    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1155      - added a test for tm==NULL to gmtime.c and localtime.c
1156      - applied patch for incorrect GetFileTime under win32
1157      - toSeconds -> toMilliseconds in Win32/win32-process.sml
1158    
1159    ----------------------------------------------------------------------
1160    Name: Matthias Blume (blume (at) tti - c (dot) org)
1161    Date: 2004/07/21 18:20:00 CDT
1162    Tag: blume-20040721-nlffigen
1163    Description:
1164    
1165    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1166      for a typedef to an incomplete type, but leave out the "typ" member.
1167      (This is just for consistency.)
1168    - Started to produce what is supposed to become better (i.e., comprehensive)
1169      documentation of what ml-nlffigen does and produces.
1170    
1171    ----------------------------------------------------------------------
1172    Name: Matthias Blume (blume (at) tti - c (dot) org)
1173    Date: 2004/07/14 16:25:00 CDT
1174    Tag: blume-20040714-union
1175    Description:
1176    
1177    Added C_UNION to c-calls/c-types.sml and updated the machinery
1178    (ml-nlffigen, cproto.sml) that conveys C function interface
1179    information to the code generator.
1180    
1181    However, the actual architecture-specific implementation of function
1182    arguments and results that are C unions is still not implemented.
1183    
1184    ----------------------------------------------------------------------
1185    Name: Allen Leung (leunga (at) reservoir (dot) com)
1186    Date: 2004/07/14 14:38:00 EST
1187    Tag: leunga-110_46_1-ppc-lwzu
1188    Description:
1189    
1190        Added these instructions to the PowerPC architecture:
1191        LBZU(X), LHZU(X), LWZU(X),
1192        STWU(X), STFDU, STFSU
1193    
1194        etc...
1195    
1196        Note: I haven't added their instruction encoding into the description.
1197    
1198    ----------------------------------------------------------------------
1199    Name: Allen Leung (leunga (at) reservoir (dot) com)
1200    Date: 2004/07/13 15:04:00 EST
1201    Tag: leunga-110_46_1-ppc-lwarx
1202    Description:
1203    
1204       Added the two instructions LWARX and STWCX to the PowerPC
1205    instruction set.
1206    
1207       A (untested) rewrite of loop-structure.sml.  The old version
1208    is completely broken.
1209    
1210    ----------------------------------------------------------------------
1211    Name: Matthias Blume (blume (at) tti - c (dot) org)
1212    Date: 2004/07/13 13:50:00 CDT
1213    Tag: blume-20040713-nlffi
1214    Description:
1215    
1216    - use paramAlloc to report c-calls with too many arguments
1217      (for PPC version where parameter area is pre-allocated)
1218    - added ccall_maxargspace to machspec (to implement the above)
1219    - made "make" commend in CM's "make" tool configurable
1220    - added option (default: on) for passing the name of the SML/NJ's "bin"
1221      directory to "make"; the call looks like this:
1222    
1223         make <options> SMLNJ_BINDIR=<dir> <target>
1224    
1225      This can be used by the Makefile to, e.g., pick the "right" version
1226      of ml-nlffigen.
1227    - minor code tweaks
1228    
1229    ----------------------------------------------------------------------
1230    Name: Matthias Blume (blume (at) tti - c (dot) org)
1231    Date: 2004/07/12 22:50:00 CDT
1232    Tag: blume-110_46_1-macosx-nlffi
1233    Description:
1234    
1235    NLFFI under Mac OS X now working (sort of).  This is largely untested,
1236    though.
1237    
1238    Note:  1. You have to make a new, clean build of the runtime system.
1239           2. There are new BOOTFILES, you have to use them!
1240              (Doing the bootstrap process yourself would be *very* painful!
1241               If you absolutely have to do it, build the system under
1242               a different architecture and then cross-compile.)
1243    
1244    Version bumped to 110.46.1 to account for runtime data format changes.
1245    
1246    ----------------------------------------------------------------------
1247    Name: Matthias Blume (blume (at) tti - c (dot) org)
1248    Date: 2004/06/18 14:30:00 CDT
1249    Tag: blume-20040618-unix
1250    Description:
1251    
1252    Changed the implementation of structure Unix so that the same stream
1253    is returned every time one of the {text,bin}{In,Out}streamOf functions
1254    is invoked on the same proc.  This is not what the spec currently
1255    says -- although IMO it arguably should.  (See discussion below.)
1256    
1257    ----------------------------------------------------------------------
1258    Name: Matthias Blume (blume (at) tti - c (dot) org)
1259    Date: 2004/06/17 18:15:00 CDT
1260    Tag: Release_110_46
1261    Description:
1262    
1263    New working version (110.46).  NEW BOOTFILES!
1264    
1265    ----------------------------------------------------------------------
1266    Name: Matthias Blume (blume (at) tti - c (dot) org)
1267    Date: 2004/06/17 17:20:00 CDT
1268    Tag: blume-20040617-timer-unix
1269    Description:
1270    
1271    Changed the interface of structures Timer and Unix to match the most
1272    recent Basis spec.
1273    
1274    In the case of Unix there still seems to be an open/weird issue:
1275    
1276       The {text,bin}{In,Out}streamOf functions are supposed to create
1277       fresh streams whenever they are called -- as opposed to have them
1278       return the same stream every time.  This design is supposed to
1279       prevent space leaks caused by proc values hanging on to streams.
1280    
1281       The reap function, on the other hand, is supposed to close the
1282       streams.  This cannot be done without having a handle on the
1283       stream in proc after all...
1284    
1285    I took the liberty to implement the following stopgap solution:
1286    
1287      The proc value hangs on to the most recently created stream(s).
1288      Reap closes those.  If either or both of the two streams hadn't
1289      been created at all yet, then reap will close the corresponding
1290      file descriptors directly.
1291    
1292    PS: I don't understand the original space leak argument anymore.  If
1293    a proc hangs on to the imperative stream, then I/O operations on those
1294    will advance the state of the cached stream and avoid the space leak.
1295    
1296    ----------------------------------------------------------------------
1297    Name: Matthias Blume (blume (at) tti - c (dot) org)
1298    Date: 2004/05/28 16:45:00 CDT
1299    Tag: blume-20040528-basis
1300    Description:
1301    
1302    Added signature PACK_REAL and exported functor PrimIO.
1303    
1304    ----------------------------------------------------------------------
1305    Name: Matthias Blume (blume (at) tti - c (dot) org)
1306    Date: 2004/05/25 16:00:00 CDT
1307    Tag: blume-20040525-group-owner
1308    Description:
1309    
1310    CM now ignores (but still accepts) the "owner" information in group
1311    descriptions.  The owner of a group is its next enclosing
1312    library. Each group must have a unique owner.  (There is a virtual
1313    "toplevel" library that owns groups which are not nested within a real
1314    library.)  Previously, each group had to explicitly declare its owner,
1315    and CM would check that such a declaration is correct.  The new scheme
1316    is to have CM check that for each group there is precisely one owning
1317    library.
1318    
1319    The advantage of the new scheme is that the programmer no longer needs
1320    to maintain the somewhat annoying owner information.  The downside is
1321    that CM cannot enforce the ownership rule across multiple runs of
1322    CM.make.  Fortunately, enclosing the same group in two different
1323    libraries A and B which are not part of the same program does not
1324    cause real problems.
1325    
1326    ----------------------------------------------------------------------
1327    Name: Matthias Blume (blume (at) tti - c (dot) org)
1328    Date: 2004/05/20 16:00:00 CDT
1329    Tag: blume-20040520-win32
1330    Description:
1331    
1332    Made the win32 version work again.  (Strangely, a misplaced comma had
1333    slipped into win32-process.c which prevented the runtime from being
1334    compiled correctly.)
1335    
1336    Also, included a minor addition to ml-build.bat analogous to what was
1337    done in blume-20040519-ml-build.
1338    
1339    ----------------------------------------------------------------------
1340    Name: Matthias Blume (blume (at) tti - c (dot) org)
1341    Date: 2004/05/19 22:10:00 CDT
1342    Tag: blume-20040519-ml-build
1343    Description:
1344    
1345    Arranged for ml-build to clean up after itself a little bit better.
1346    The script generates a temporary SML source file and compiles it using
1347    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1348    gets rid of those at the end, so they don't accumulate under .cm.
1349    
1350    This required a minor change to install.sh because the name of the
1351    metadata directory (default: .cm) is actually configurable at
1352    installation time.
1353    
1354    ----------------------------------------------------------------------
1355    Name: Matthias Blume (blume (at) tti - c (dot) org)
1356    Date: 2004/05/18 15:50:00 CDT
1357    Tag: blume-20040518-mkreader
1358    Description:
1359    
1360    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1361    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1362    
1363    ----------------------------------------------------------------------
1364    Name: Matthias Blume (blume (at) tti - c (dot) org)
1365    Date: 2004/05/11 14:35:00 CDT
1366    Tag: blume-20040511-win32sock
1367    Description:
1368    
1369    Added previously missing support for many socket-related functions
1370    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1371    for the voluminous patch!
1372    
1373    (I have not tested this patch under win32 yet.)
1374    
1375    Here is David's e-mail:
1376    
1377    Hi,
1378    
1379    Attached to this email you find a diff against sml/nj 110.45
1380    that will enable socket support under Windows.
1381    
1382    To apply the patch (using unix or cygwin)
1383    1) gunzip runtime.diff.gz
1384    2) "cd" into "src/runtime" in the source tree of a fresh
1385       110.45 installation.
1386    3) patch -p 1 < [your/path/to]runtime.diff
1387    
1388    The code compiles fine but has NOT yet been extensively tested.
1389    I only ran a few tests for basic socket client functionality
1390    (which worked fine).  Especially the functions that use ioctl
1391    are not tested at all and might not work (see below).
1392    
1393    I implemented this since we want to move to a newer version of sml/nj
1394    but need socket support in order to use it.  This is the first time I
1395    even had a look at the sml/nj source,  so please review my changes
1396    before making this part of the distribution!  Here are a few issues
1397    that I think might be better for someone to solve who is more
1398    familiar with the sml/nj source (and socket programming):
1399    
1400    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1401      exception since I could not figure out what the windows equivalent
1402      of these functions is
1403    
1404    - In sockets-osdep.h there are a some #include statements that are
1405      only used in a few files that include sockets-osdep.h
1406    
1407    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1408      does not process its return value since I don't know how to
1409      report an error upwards.
1410    
1411    - It would probably be good to have a call to WSACleanup() when
1412      the library is unloaded (if there is such a possibility).
1413      Otherwise I think Windows will take care of this automatically
1414      when the process finishes.
1415    
1416    - I used ioctlsocket() as a replacement for ioctl() but I have
1417      no idea if that is actually the proper replacement on Windows.
1418    
1419    - All these issues are marked in the code by "FIXME" comments.
1420    
1421    We use sml/nj extensively in our products and are quite happy
1422    with it.  I hope this contribution will help you.
1423    
1424    Keep up the good work!
1425    
1426    David
1427    
1428    ----------------------------------------------------------------------
1429    Name: Matthias Blume (blume (at) tti - c (dot) org)
1430    Date: 2004/05/11 14:20:00 CDT
1431    Tag: blume-20040511-installml
1432    Description:
1433    
1434    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1435    the patch.)
1436    
1437    ----------------------------------------------------------------------
1438    Name: Matthias Blume (blume (at) tti - c (dot) org)
1439    Date: 2004/05/11 14:05:00 CDT
1440    Tag: blume-20040511-nlffi-netbsd
1441    Description:
1442    
1443    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1444    the patch.)
1445    
1446    ----------------------------------------------------------------------
1447    Name: Matthias Blume (blume (at) tti - c (dot) org)
1448    Date: 2004/05/11 12:05:00 CDT
1449    Tag: blume-20040511-exports
1450    Description:
1451    
1452    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1453    export lists in compiler-related .cm-files.
1454    
1455    ----------------------------------------------------------------------
1456    Name: Matthias Blume (blume (at) tti - c (dot) org)
1457    Date: 2004/05/11 11:35:00 CDT
1458    Tag: blume-20040511-allsource
1459    Description:
1460    
1461    The installer now honors the "src-smlnj" target again, although its meaning
1462    has changed from "all sources required for the compiler" to "all sources
1463    the installer knows about".  In other words, if you enable "src-smlnj"
1464    in the "targets" file, then the installer will pull in sources for
1465    everything.  (Notice that this refers to source code only.  Compiled
1466    code is still only installed for modules that were requested explicitly
1467    or which are required for other modules that were requested explicitly.)
1468    
1469    ----------------------------------------------------------------------
1470    Name: Matthias Blume (blume (at) tti - c (dot) org)
1471    Date: 2004/04/23 17:40:00 CDT
1472    Tag: blume-20040423-ieee-scan
1473    Description:
1474    
1475    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1476    in the exponent calculation we get INF or ZERO (depending on the mantissa
1477    and the sign of the exponent).
1478    
1479    ----------------------------------------------------------------------
1480    Name: Matthias Blume (blume (at) tti - c (dot) org)
1481    Date: 2004/04/23 10:40:00 CDT
1482    Tag: blume-20040423-ml-build
1483    Description:
1484    
1485    The ml-build script now terminates with a non-0 status when something
1486    goes wrong.
1487    
1488    ----------------------------------------------------------------------
1489    Name: Matthias Blume (blume (at) tti - c (dot) org)
1490    Date: 2004/04/22 16:35:00 CDT
1491    Tag: blume-20040422-Option
1492    Description:
1493    
1494    Made exception Option to be the same as exception Option.Option
1495    (as it should be).
1496    
1497    ----------------------------------------------------------------------
1498    Name: Allen Leung (leunga (at) reservoir (dot) com)
1499    Date: 2004/03/19 14:40:00 EST
1500    Tag: leunga-20040319-cygwin-nlffi
1501    Description:
1502    
1503    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1504    of SML/NJ.  The problem is that
1505    
1506       lib = dlopen(NULL, ...)
1507       f   = dlsym(lib, "malloc");
1508    
1509    does not work on Windows unless we explicitly export symbols
1510    such as 'malloc' during linking.  We fixed this by explicitly
1511    exporting the required symbols with the magic gcc incantation:
1512    
1513        -Wl,--export-all cygwin.def
1514    
1515    where cygwin.def is a file containing all the symbols that
1516    we wish to export.
1517    
1518    I suspect this is a Windows problem and we'll have to
1519    do the same (somehow with windows compilers) when
1520    we build the native win32 version with the system
1521    calls LoadLibrary/GetProcAddress.
1522    
1523    ----------------------------------------------------------------------
1524    Name: Matthias Blume (blume (at) tti - c (dot) org)
1525    Date: 2004/03/04 16:35:00 CST
1526    Tag: blume-20040304-intinf-fmt
1527    Description:
1528    
1529    Fixed problem with IntInf.fmt (sign would show up on the right instead
1530    of on the left for BIN, OCT, and HEX).
1531    
1532    ----------------------------------------------------------------------
1533    Name: Matthias Blume (blume (at) tti - c (dot) org)
1534    Date: 2004/03/04 11:25:00 CST
1535    Tag: blume-20040304-symlinks
1536    Description:
1537    
1538    Fixed problem with installer script (unix only) where bin/ml-yacc and
1539    friends pointed (via symlinks) to absolute locations instead of just
1540    .run-sml.  This was reported by Vesa A Norrman.
1541    
1542    ----------------------------------------------------------------------
1543    Name: Matthias Blume (blume (at) tti - c (dot) org)
1544    Date: 2004/02/13 14:50:00 CST
1545    Tag: Release_110_45
1546    Description:
1547    
1548    New working version (110.45). New bootfiles.
1549    
1550    ----------------------------------------------------------------------
1551    Name: Matthias Blume (blume (at) tti - c (dot) org)
1552    Date: 2004/01/26 15:15:15 CST
1553    Tag: blume-20040126-toplevel
1554    Description:
1555    
1556    Improved handling of exceptions at the interactive toplevel.
1557    
1558    ----------------------------------------------------------------------
1559    Name: Matthias Blume (blume (at) tti - c (dot) org)
1560    Date: 2004/01/26 11:25:00
1561    Tag: blume-20040126-app
1562    Description:
1563    
1564    Type of top-level "app" corrected.
1565    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1566    
1567    ----------------------------------------------------------------------
1568    Name: Matthias Blume (blume (at) tti - c (dot) org)
1569    Date: 2003/11/18 17:10 CST
1570    Tag: blume-20031118-basis-fiddle
1571    Description:
1572    
1573    - changed Timer interface to what might become the spec
1574    - POSIX_FLAGS -> BIT_FLAGS according to spec
1575    - some other minor discrepancies wrt. spec eliminated
1576    
1577    ----------------------------------------------------------------------
1578    Name: Matthias Blume (blume (at) tti - c (dot) org)
1579    Date: 2003/11/06 12:00:00 CST
1580    Tag: Release_110_44
1581    Description:
1582    
1583    New working version (110.44). New bootfiles.
1584    
1585    ----------------------------------------------------------------------
1586    Name: Matthias Blume (blume (at) tti - c (dot) org)
1587    Date: 2003/11/04 11:50:00 CST
1588    Tag: blume-20031104-move-libraries
1589    Description:
1590    
1591    Eliminated the "dont_move_libraries" directive in config/targets.
1592    (The mechanism was broken and could not be fixed easily.  Moreover,
1593    there does not seem to be any reason not to move all libraries into
1594    lib during installation.  I originally implemented this directive as a
1595    backward-compatibility feature when I first introduced the new CM.
1596    Now that things have been stable for a long time and going back to the
1597    old CM is not an option, there is no reason to keep it around.)
1598    
1599    ----------------------------------------------------------------------
1600    Name: Matthias Blume (blume (at) tti - c (dot) org)
1601    Date: 2003/11/03 16:00:00 CST
1602    Tag: blume-20031103-installdir
1603    Description:
1604    
1605    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1606    Richards for pointing out the problem and providing the solution.)
1607    
1608    ----------------------------------------------------------------------
1609    Name: Matthias Blume (blume (at) tti - c (dot) org)
1610    Date: 2003/10/01 17:05:00 CDT
1611    Tag: blume-20031001-lal-mlrisc
1612    Description:
1613    
1614    MLRISC bug fix from Lal.
1615    
1616    ----------------------------------------------------------------------
1617    Name: Matthias Blume (blume (at) tti - c (dot) org)
1618    Date: 2003/09/30 16:10:00 CDT
1619    Tag: blume-20030930-primio-bat
1620    Description:
1621    
1622    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1623    2. Improved .bat files (for Win32 port) to make things work under Win95.
1624       (thanks to Aaron S. Hawley for this one)
1625    
1626    ----------------------------------------------------------------------
1627    Name: Matthias Blume (blume (at) tti - c (dot) org)
1628    Date: 2003/09/26 16:05:00 CDT
1629    Tag: blume-20030926-wrappriv
1630    Description:
1631    
1632    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1633    
1634    ----------------------------------------------------------------------
1635    Name: Matthias Blume (blume (at) tti - c (dot) org)
1636    Date: 2003/09/26 15:00:00 CDT
1637    Tag: blume-20030926-110_43_3
1638    Description:
1639    
1640    - additional cleanup
1641    - version number bump, NEW BOOTFILES
1642    
1643    ----------------------------------------------------------------------
1644    Name: Matthias Blume (blume (at) tti - c (dot) org)
1645    Date: 2003/09/26 12:00:00 CDT
1646    Tag: blume-20030926-ppautoload
1647    Description:
1648    
1649    I modified the read-eval-print loop so that the autoloader gets
1650    invoked whenever the prettyprinter tries to look up a symbol that
1651    is not currently defined in the toplevel environment but which
1652    appears in CM's autoload registry.  As a result, we see far fewer of
1653    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1654    
1655    In addition to this I tried to clean up some pieces of the Basis
1656    implementation (e.g., Socket, Word8Array) in order to prevent other
1657    instances of these ?.Foo.Bar.xxx names from being printed.
1658    
1659    The mechanism that picks names for types still needs some work, though.
1660    (Right now it seems that if there is a type A.t which is defined to
1661    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1662    "?.B.u" although the perhaps more sensible solution would be to use
1663    "A.t" in this case.  In other words, the prettyprinter should follow
1664    a chain of DEFtycs not farther than there are corresponding toplevel
1665    names in the current environment.)
1666    
1667    ----------------------------------------------------------------------
1668    Name: Matthias Blume (blume (at) tti - c (dot) org)
1669    Date: 2003/09/24 16:31:00 CDT
1670    Tag: blume-20030924-installer
1671    Description:
1672    
1673    Another installer tweak:  All the ML code for the installer is now
1674    compiled during CMB.make and put into a little library called
1675    $smlnj/installer.cm.  The installation then simply invokes
1676    
1677       sml -m $smlnj/installer.cm
1678    
1679    and everything happens automagically.
1680    
1681    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1682    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1683            and BINDIR.
1684    
1685    The new scheme guarantees that the ML code responsible for the installation
1686    is in sync with the APIs of the main system.  Also, the installer is
1687    somewhat faster because the installer script is precompiled.
1688    
1689    ----------------------------------------------------------------------
1690    Name: Matthias Blume (blume (at) tti - c (dot) org)
1691    Date: 2003/09/24 15:35:00 CDT
1692    Tag: blume-20030924-synsock
1693    Description:
1694    
1695    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1696    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1697    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1698    implementations of non-blocking operations from CML's Socket
1699    structure.
1700    
1701    ----------------------------------------------------------------------
1702    Name: Matthias Blume (blume (at) tti - c (dot) org)
1703    Date: 2003/09/24 15:10:05 CDT
1704    Tag: blume-20030924-sockets
1705    Description:
1706    
1707    1. Fixed SOCKET API and implementation to match Basis spec.
1708       This required changing the internal representation of sockets to one
1709       that remembers (for each socket file descriptor) whether it is currently
1710       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1711       call is made only if the state actually needs to change).
1712    
1713    2. OS-specific details of sockets were moved into separate files, thus
1714       making it possible to unify the bulk of the socket implementations
1715       between Unix and Win32.
1716    
1717    3. CML's socket API changed accordingly.
1718       (Note that we need to remove non-blocking functions from this API
1719       since they are redundant in the case of CML!)
1720    
1721    4. CML's socket implementation now makes use of non-blocking functions
1722       provided by Basis, thus removing all OS-dependent code from this part
1723       of CML.
1724    
1725    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1726    
1727    ----------------------------------------------------------------------
1728    Name: Matthias Blume (blume (at) tti - c (dot) org)
1729    Date: 2003/09/22 12:10:00 CDT
1730    Tag: blume-20030922-110_43_2
1731    Description:
1732    
1733    Made a new interim version and bootfiles for developer's bootstrapping
1734    convenience.
1735    
1736       110.43.2 -- NEW BOOTFILES
1737    
1738    ----------------------------------------------------------------------
1739    Name: Matthias Blume (blume (at) tti - c (dot) org)
1740    Date: 2003/09/19 15:55:00 CDT
1741    Tag: blume-20030919-cmdir
1742    Description:
1743    
1744    1. new-install.sh -> install.sh
1745    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1746    3. tweaked installer so that another name instead of .cm can be chosen
1747       at install time (by setting the CM_DIR_ARC environment variable
1748       during installation); once installation is complete, the name is
1749       fixed
1750    
1751    ----------------------------------------------------------------------
1752    Name: Matthias Blume (blume (at) tti - c (dot) org)
1753    Date: 2003/09/18 16:00:00 CDT
1754    Tag: blume-20030918-110_43_1
1755    Description:
1756    
1757    Made a new interim version and bootfiles for developer's bootstrapping
1758    convenience.
1759    
1760       110.43.1 -- NEW BOOTFILES
1761    
1762    ----------------------------------------------------------------------
1763    Name: Matthias Blume (blume (at) tti - c (dot) org)
1764    Date: 2003/09/18 15:20:00 CDT
1765    Tag: blume-20030918-misc
1766    Description:
1767    
1768    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
1769       this seems to be controversial at the moment) and used those in
1770       Posix.ProcEnv.times.
1771    
1772    2. Added Time.{from,to}Nanoseconds to Time.
1773    
1774    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
1775       For example, fromLargeInt never needs to look at more than 3 "big
1776       digits" to get its 53 bits of precision.
1777    
1778    ----------------------------------------------------------------------
1779    Name: Matthias Blume (blume (at) tti - c (dot) org)
1780    Date: 2003/09/17 16:30:00 CDT
1781    Tag: blume-20030917-real32-slices
1782    Description:
1783    
1784    Added an entry to the primitive environment
1785    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
1786    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
1787    
1788    Removed some of the "magic" constants in real64.sml and replaced them
1789    with code that generates these values from their corresponding
1790    integer counterparts.
1791    
1792    Made all(?) the slice-related changes to the Basis and made everything
1793    compile again...
1794    
1795    ----------------------------------------------------------------------
1796    Name: Matthias Blume (blume (at) tti - c (dot) org)
1797    Date: 2003/09/15 17:45:00 CDT
1798    Tag: blume-20030915-rbase
1799    Description:
1800    
1801    Fixed bug in Real.fromLargeInt.
1802    
1803    ----------------------------------------------------------------------
1804    Name: Matthias Blume (blume (at) tti - c (dot) org)
1805    Date: 2003/09/13 18:11:00 CDT
1806    Tag: blume-20030913-libinstall
1807    Description:
1808    
1809    Minor bugfix in config/libinstall (set anchor with path to
1810    standalone tool after installing it, otherwise libraries that
1811    need ml-lex or ml-yacc won't compile the first time the installer
1812    runs).
1813    
1814    ----------------------------------------------------------------------
1815    Name: Matthias Blume (blume (at) tti - c (dot) org)
1816    Date: 2003/09/12 11:45:00 CDT
1817    Tag: blume-20030912-various
1818    Description:
1819    
1820     - fixed bug in Real.toLargeInt
1821     - fixed bug in Posix.ProcEnv.times
1822     - changed inputLine functions to return an option
1823     - minor installer improvements / bugfixes
1824     - changed default @SMLalloc parameter for x86/celeron to 64k
1825    
1826    ----------------------------------------------------------------------
1827    Name: Matthias Blume (blume (at) tti - c (dot) org)
1828    Date: 2003/09/09 22:00:00 CDT
1829    Tag: Release_110_43
1830    Description:
1831    
1832        New working release 110.43.  New bootfiles.
1833    
1834    ----------------------------------------------------------------------
1835    Name: Matthias Blume (blume (at) tti - c (dot) org)
1836    Date: 2003/09/09 19:20:00 CDT
1837    Tag: blume-20030909-installer
1838    Description:
1839    
1840    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
1841    Modified config/install.bat to take advantage of it.  Also modified
1842    config/install.sh (and called it config/new-install.sh) to take advantage
1843    of it on Unix systems.  (The SML code is (supposed to be) platform-
1844    independent.)
1845    
1846    The installer can now install everything under Win32
1847    as well as under *nix as long as it compiles.
1848    
1849    Other changes:
1850    
1851      - made CML compile again under Win32
1852      - made eXene compile under Win32 (by providing a fake structure UnixSock
1853        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
1854      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
1855        same as type int;  under Win32 it isn't
1856      - fixed some slice-related problems in the win32-specific parts of CML
1857      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
1858        (under Win32, the volume name is case-insensitive, and the
1859        OS.Path code compares volume names for equality)
1860    
1861    ----------------------------------------------------------------------
1862    Name: Matthias Blume (blume (at) tti - c (dot) org)
1863    Date: 2003/09/08 11:55:00 CDT
1864    Tag: blume-20030908-fullpath
1865    Description:
1866    
1867    Made Win32 version of OS.FileSys.fullPath return current directory
1868    when given an empty string.  This is what the spec says, and incidentally,
1869    CM depends on it.  (CM otherwise goes into an infinite loop in certain
1870    cases when presented with the name of a non-existing .cm file.)
1871    
1872    ----------------------------------------------------------------------
1873    Name: Matthias Blume (blume (at) tti - c (dot) org)
1874    Date: 2003/09/04 16:30:00 CDT
1875    Tag: blume-20030905-slices-etc
1876    Description:
1877    
1878      1. Changed interface to vectors and arrays in Basis to match
1879         (draft) Basis spec.
1880      2. Added signatures and implementations of slices according to
1881         Basis spec.
1882      3. Edited source code throughout the system to make it compile again
1883         under 1. and 2.  (In some cases code had to be added to have it
1884         match the new signatures.)
1885      4. MLRISC should be backward-compatible: the copies of the originals
1886         of files that needed to change under 3. were retained, the .cm files
1887         check the compiler version number and use old versions when
1888         appropriate.
1889      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
1890         dirstream -> string option (in accordance with Basis spec).
1891      6. When generating code that counts lines, ml-lex used function
1892         CharVector.foldli, taking advantage of its old interface.
1893         This has been replaced with the corresponding code from
1894         CharVectorSlice.  (html-lex must be re-lexed!)
1895      7. BitArray in smlnj-lib/Util has been extended/modified to match the
1896         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
1897      8. Removed temporary additions (fromInternal, toInternal) from the
1898         (now obsolete) IntInf in smlnj-lib/Util.
1899      9. Cleaned up structure Byte.
1900     10. Added localOffset, scan, and fromString to Date (according to spec).
1901         Cleaned/corrected implementation of Date.
1902         (Still need to check for correctness; implement better canonicalizeDate.)
1903     11. Added "scan" to signature IEEE_REAL.
1904     12. Some improvements to IntInf [in particular: efficiency-hack for
1905         mod and rem when second operand is 2 (for parity checks).]
1906     13. Changed representation of type Time.time, using a single IntInf.int
1907         value counting microseconds.  This considerably simplified the
1908         implementation of structure Time.  We now support negative time
1909         values; scan and fromString handle signs.
1910     14. Functor PrimIO now takes two additional arguments (VectorSlice and
1911         ArraySlice).
1912    
1913    ----------------------------------------------------------------------
1914    Name: Matthias Blume (blume (at) tti - c (dot) org)
1915    Date: 2003/08/28 17:00:00 CDT
1916    Tag: blume-20030828-intinf
1917    Description:
1918    
1919    This is a major update which comes with a version number bump
1920    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
1921    and an implementation of IntInf in the Basis.
1922    
1923    There are a fairly large number of related changes and updates throughout
1924    the system:
1925    
1926    Basis:
1927            - Implemented IntInf.
1928            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
1929            - Added some missing Real64 operations, most notably Real.toLargeInt.
1930            - Added FixedInt as a synonym for Int32.
1931    
1932    compiler:
1933            * Added support for a built-in intinf type.
1934              - literals
1935              - pattern matching
1936              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
1937              - overloading on literals and operations
1938    
1939            This required adding a primitive type intinf, some additional
1940            primops, and implementations for several non-trivial intinf
1941            operations in Core.  (The intinf type is completely abstract
1942            to the compiler; all operations get delegated back to the Core.)
1943    
1944            * Intinf equality is handled by polyequal.  However, the compiler
1945            does not print its usual warning in this case (since polyequal
1946            is the right thing to do there).
1947    
1948            * Improved the organization of structure InlineT.
1949    
1950            * A word about conversion primops:
1951            If conversions involving intinf do not cancel out during
1952            CPS contract, then the compiler must insert calls to Core functions.
1953            Since all core access must be resolved already during the FLINT
1954            translate phase, it would be too late a the time of CPS contract
1955            to add new Core calls.  For this reason, conversion primops
1956            for intinf carry two arguments:  1. the numeric argument that
1957            they are supposed to convert, and 2. the Core function that
1958            can help with this conversion if necessary.  If CPS contract
1959            eliminates a primop, then the associated Core function becomes
1960            dead and goes away.  Intinf conversion primops that do not get
1961            eliminated by CPS contract get rewritten into calls of their
1962            core functions by a separate, new phase.
1963    
1964    interactive system:
1965            - Control.Print.intinfDepth controls max length of intinf constants
1966            being printed. (Analogous to Control.Print.stringDepth.)
1967            - Cleanup in printutil and pputil: got rid of unused stuff and
1968            duplicates; replaced some of the code with code that makes better
1969            use of library functionality.
1970    
1971    CM:
1972            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
1973            are no longer silent.
1974    
1975    CKIT:
1976            Fixed mismatched uses of Int32 and LargeInt.  I always decided
1977            in favor of LargeInt -- which is now the same as IntInf.
1978            CKIT-knowledgable people should check whether this is what's
1979            intended and otherwise change things back to using Int32 or
1980            FixedInt.
1981    
1982    Throughout the code:
1983            Started using IntInf.int literals and built-in operations
1984            (e.g., comparison with 0) where this seems appropriate.
1985    
1986    
1987    ----------------------------------------------------------------------
1988    Name: Dave MacQueen (dbm@cs.uchicago.edu)
1989    Date: 2003/08/13 11:36:00 CDT
1990    Tag: dbm-20030813-mcz-merge1
1991    Description:
1992    
1993    Merging changes from the mcz-branch development branch into trunk.
1994    These changes involve replacement of the emulated old prettyprinter
1995    interface with direct use of the SML/NJ Lib PP library, and fixing
1996    of a couple of bugs (895, 1186) relating to error messages.  A new
1997    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
1998    has been added.
1999    
2000    ----------------------------------------------------------------------
2001    Name: Matthias Blume (blume (at) tti - c (dot) org)
2002    Date: 2003/08/11 15:45:00 CDT
2003    Tag: blume-20030811-windows
2004    Description:
2005    
2006    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
2007    
2008        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
2009    
2010    This patch restores SML/NJ's ability to run under win32.  There are a
2011    number of changes, including fixes for several bugs that had gone
2012    unnoticed until now:
2013    
2014      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
2015        trying to run the win32 version from within cygwin.)
2016    
2017      - There are a number of simple .bat scripts that substitute for their
2018        corresponding Unix shell-scripts.  (See below.)
2019    
2020      - The internals of ml-build have been modified slightly.  The main
2021        difference is that instead of calling ".link-sml" (or link-sml.bat)
2022        using OS.Process.system, the ML process delegates this task back
2023        to the script.  Otherwise problems arise in mixed environments such
2024        as Cygwin where scripts look and work like Unix scripts, but
2025        where OS.Process.system cannot run them.
2026    
2027      - In CM, the srcpath pickler used native pathname syntax -- which
2028        is incorrect in the case of cross-compilation.  The new pickle format
2029        is independent of platform-specific naming conventions.
2030    
2031      - Path configuration files (such as lib/pathconfig) can now choose
2032        between native and standard syntax.  Placing a line of the form
2033    
2034            standard!
2035    
2036        into the file causes all subsequent paths to be interpreted using
2037        CM standard pathname syntax (= Unix conventions); a line
2038    
2039            native!
2040    
2041        switches back to native style.  This was needed so that
2042        path config files can be written portably, see src/system/pathconfig.
2043    
2044      - Runtime system:
2045    
2046          - win32-filesys.c:  get_file_time and set_file_time now
2047            access modification time, not creation time.
2048    
2049          - I/O code made aware of new array representation.
2050    
2051          - Bug fixes in X86.prim.masm.
2052    
2053       - src/system/makeml made aware of win32. (For use under cygwin
2054            and other Unix-environments for windows.)
2055    
2056       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
2057         which caused BinIO.inputAll to fail consistently.
2058    
2059    .bat scripts:
2060    
2061       Windows .bat scripts assume that SMLNJ_HOME is defined.
2062    
2063       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
2064         applications (sml, ml-yacc, ml-lex).
2065       - ml-build.bat: analogous to ml-build.
2066       - config\install.bat:  Analogous to config/install.sh.  This requires
2067         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2068         (nmake etc. must be on the path, and vcvars32 must have been run.)
2069         Moreover, sources for ml-lex and ml-yacc need to exist under src,
2070         and the bootfile hierarchy must have been unpacked under
2071         sml.boot.x86-win32.
2072         The script is very primitive and does a poor job at error checking.
2073         It only installs the base system, ml-lex, and ml-yacc.  No other
2074         libraries are being installed (i.e., you get only those that
2075         are part of the compiler.)
2076       - link-sml.bat:  analogous to .link-sml, but not currently used
2077    
2078    Unrelated bug fixes:
2079    
2080      - ml-nlffigen now exports structures ST_* corresponding to incomplete
2081        types.
2082      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
2083        otherwise.)
2084    
2085    ----------------------------------------------------------------------
2086    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2087    Date: 2003/06/17
2088    Tag: macqueen-20030617-bug895
2089    Description:
2090    
2091    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2092    Tag will be used for new development branch (mcz-branch) for
2093    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2094    
2095    ----------------------------------------------------------------------
2096    Name: Matthias Blume (blume (at) tti - c (dot) org)
2097    Date: 2003/05/27 16:55:00 CDT
2098    Tag: blume-20030527-polyeq
2099    Description:
2100    
2101    Tried to eliminated most cases of polymorphic equality.
2102    
2103    ----------------------------------------------------------------------
2104    Name: Matthias Blume (blume (at) tti - c (dot) org)
2105    Date: 2003/05/21 17:45:00 CDT
2106    Tag: blume-20030517-complete
2107    Description:
2108    
2109    Two changes:
2110    
2111      1. Added a flag for controlling whether non-exhaustive bindings will
2112         be treated as errors (default is false).
2113      2. Cleaned up the *entire* source tree so that CMB.make goes through
2114         without a single non-exhaustive match- or bind warning.
2115    
2116    ----------------------------------------------------------------------
2117    Name: Matthias Blume (blume (at) tti - c (dot) org)
2118    Date: 2003/05/17 10:20:00 CDT
2119    Tag: blume-20030517-absyn
2120    Description:
2121    
2122    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2123    
2124       This mainly affects the quality of error messages.  However, some
2125       of the code is now more straightforward than before. (Treatment of
2126       the above four constructs in translate.sml is much simpler than
2127       the "macro-expansion" that was going on before.  Plus, the mach-
2128       compiler no longer gets invoked just to be able to compile an
2129       if-expression.)
2130    
2131    2. The ErrorMsg.Error exception is now caught and absorbed by the
2132       interactive loop.
2133    
2134    ----------------------------------------------------------------------
2135    Name: Allen Leung
2136    Date: 2003/05/16 13:05:00 CDT
2137    Tag: leunga-20030516-cygwin-runtime
2138    Description:
2139    
2140        Ported the runtime system to cygwin, which uses the unix
2141    x86-unix bin files.  Missing/buggy features:
2142    
2143       o getnetbyname, getnetbyaddr: these functions seem to be missing in
2144         the Cygwin library.
2145       o Ctrl-C handling may be flaky.
2146       o Windows system calls and Windows I/O are not supported.
2147    
2148       A new set of binfiles is located at:
2149    
2150         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2151    
2152       This is only needed for bootstrapping the cygwin version of smlnj.
2153       Other x86 versions can use the existing binfiles.
2154    
2155    ----------------------------------------------------------------------
2156    Name: Matthias Blume
2157    Date: 2003/04/08 15:42:00 CDT
2158    Tag: blume-20030408-listpair
2159    Description:
2160    
2161    1. Added a target 'mlrisc' to installer.
2162    
2163    2. Added missing elements to structure ListPair.
2164    
2165    ----------------------------------------------------------------------
2166    Name: Allen Leung
2167    Date: 2003/01/07 10:40:00 EST
2168    Tag: leunga-20030107-int-rem
2169    Description:
2170    
2171    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2172    The arguments to the SUBL instruction were swapped.
2173    
2174    ----------------------------------------------------------------------
2175    Name: Matthias Blume
2176    Date: 2002/12/12 16:25:00 EST
2177    Tag: blume-20021212-risc-ra
2178    Description:
2179    
2180    Fixed a serious bug in the rewrite code for FP spilling/reloading that
2181    sent the RA into an infinite loop when floating point registers get
2182    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2183    110.37 and 110.38.)
2184    There was another set of potential problems related to the handling of
2185    MLRISC annotations (but those did not yet cause real problems, apparently).
2186    
2187    ----------------------------------------------------------------------
2188    Name: Matthias Blume
2189    Date: 2002/12/06 22:40:00 EST
2190    Tag: blume-20021206-cm-fileid
2191    Description:
2192    
2193    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2194    This fixes the problem of CM getting confused by files that suddenly
2195    change their identity (e.g., by getting unlinked and recreated by some
2196    text editor such as vi).  There might be a better/cheaper/cleaner way
2197    of doing this, but for now this will have to do.
2198    
2199    ----------------------------------------------------------------------
2200    Name: Matthias Blume
2201    Date: 2002/10/28 09:50:00 EST
2202    Tag: blume-20021028-typecheck
2203    Description:
2204    
2205    Exported structure Typecheck from $smlnj/viscomp/core.cm.
2206    
2207    ----------------------------------------------------------------------
2208    Name: Matthias Blume
2209    Date: 2002/10/17 09:10:00 EDT
2210    Tag: Release_110_42
2211    Description:
2212    
2213    In good old tradition, there has been a slight hiccup so that we have
2214    to patch 110.42 after the fact.   The old release tag has been replaced
2215    (see below).
2216    
2217    The change solves a problem with two competing approaches the
2218    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2219    each other's way.
2220    
2221    This change only affects the runtime system code and the installer script.
2222    (No new bootfiles.)
2223    
2224    ----------------------------------------------------------------------
2225    Name: Matthias Blume
2226    Date: 2002/10/16 12:00:00 EDT
2227    Tag: Release_110_42_removed
2228    Description:
2229    
2230    New working release.  New bootfiles.
2231    
2232    ----------------------------------------------------------------------
2233    Name: Matthias Blume
2234    Date: 2002/10/10 13:10:00 EDT
2235    Tag: blume-20021010-ppc-divs
2236    Description:
2237    
2238    The mltree operator DIVS must be implemented with an overflow check on
2239    the PPC because the hardware indicates divide-by-zero using "overflow" as
2240    well.
2241    
2242    ----------------------------------------------------------------------
2243    Name: Matthias Blume
2244    Date: 2002/07/23 11:45:00
2245    Tag: blume-20020723-smlnj-home
2246    Description:
2247    
2248    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2249    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2250    CM_PATHCONFIG is also set), the path configuration file is assumed
2251    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2252    the entire tree to some other place and everything will "just work".
2253    
2254    (Companion commands such as ml-build and ml-makedepend also sense this
2255    variable.)
2256    
2257    ----------------------------------------------------------------------
2258    Name: Matthias Blume
2259    Date: 2002/07/12 21:19:00 EDT
2260    Tag: blume-20020712-liveness
2261    Description:
2262    
2263    Exported two useful "step" functions from liveness module (MLRISC).
2264    
2265    ----------------------------------------------------------------------
2266    Name: Matthias Blume
2267    Date: 2002/07/05 16:00 EDT
2268    Tag: Release_110_41
2269    Description:
2270    
2271    New working release.  New bootfiles.
2272    
2273    ----------------------------------------------------------------------
2274    Name: Matthias Blume
2275    Date: 2002/07/05 10:25:00 EDT
2276    Tag: blume-20020705-btimp
2277    Description:
2278    
2279    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2280    other clients can set up backtracing support.
2281    
2282    ----------------------------------------------------------------------
2283    Name: Matthias Blume
2284    Date: 2002/06/25 17:23:00 EDT
2285    Tag: blume-20020625-fpmax
2286    Description:
2287    
2288    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2289    case.  (The sense of the isNaN test was reversed -- which made min and
2290    max always return their first argument.)
2291    
2292    ----------------------------------------------------------------------
2293    Name: Matthias Blume
2294    Date: 2002/06/11
2295    Tag: blume-20020611-unixpath
2296    Description:
2297    
2298    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2299    
2300    ----------------------------------------------------------------------
2301    Name: Matthias Blume
2302    Date: 2002/06/10 16:35:00 EDT
2303    Tag: blume-20020610-ieeereal
2304    Description:
2305    
2306    I back-ported my implementation of IEEEReal.fromString from the
2307    idlbasis-devel branch so that we can test it.
2308    
2309    Another small change is that ppDec tries to give more information
2310    than just "<sig>" in the case of functors.  However, this code is
2311    broken in some mysterious way if the functor's body's signature
2312    has not been declared by ascription but gets inferred from the
2313    implementation.  This needs fixing...
2314    
2315    ----------------------------------------------------------------------
2316    Name: Matthias Blume
2317    Date: 2002/05/31
2318    Tag: blume-20020531-btrace-mode
2319    Description:
2320    
2321    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2322    the wayside when I switched over to using Controls everywhere.)
2323    
2324    ----------------------------------------------------------------------
2325    Name: Lal George
2326    Date: 2002/05/23 12:21:40 EDT
2327    Tag: george-20020523-visual-labels
2328    Description:
2329    
2330    Labels are now displayed in the graphical output to make
2331    the fall-through and target blocks obvious.
2332    
2333    ----------------------------------------------------------------------
2334    Name: Matthias Blume
2335    Date: 2002/05/22 11:03:00 EDT
2336    Tag: blume-20020522-shrink
2337    Description:
2338    
2339    John tweaked yesterday's fix for 1131 to handle an out-of-memory
2340    situation that comes up when allocating huge arrays.
2341    
2342    ----------------------------------------------------------------------
2343    Name: Matthias Blume
2344    Date: 2002/05/21 16:00:00 EDT
2345    Tag: Release_110_40
2346    Description:
2347    
2348    New working release (110.40).  New bootfiles.
2349    
2350    [Also: John Reppy fixed GC bug 1131.]
2351    
2352    ----------------------------------------------------------------------
2353    Name: Matthias Blume
2354    Date: 2002/05/21 12:35:00 EDT
2355    Tag: blume-20020521-cmdoc
2356    Description:
2357    
2358    CM documentation update.
2359    
2360    ----------------------------------------------------------------------
2361    Name: Matthias Blume
2362    Date: 2002/05/21 10:55:00 EDT
2363    Tag: blume-20020521-misc
2364    Description:
2365    
2366    - John tweaked runtime to be silent on heap export (except when
2367      GC messages are on).
2368    - I added a few more things (cross-compiling versions of CMB) to
2369      config/preloads (as suggestions).
2370    
2371    ----------------------------------------------------------------------
2372    Name: Matthias Blume
2373    Date: 2002/05/20 22:25:00 EDT
2374    Tag: blume-20020520-controls
2375    Description:
2376    
2377    - Added ControlUtil structure to control-lib.cm.
2378    - Use it throughout.
2379    - Used Controls facility to define MLRISC controls (as opposed to
2380      registering MLRISC control ref cells with Controls after the fact)
2381    - Fixed messed-up controls priorities.
2382    
2383    * Removed again all the stuff from config/preloads that one wouldn't
2384      be able to preload at the time the initial heap image is built.
2385      (Many libraries, e.g., CML, do not exist yet at this time.  The
2386       only libraries that can be preloaded via config/preloads are those
2387       that come bundled with the bootfiles.)
2388    
2389    ----------------------------------------------------------------------
2390    Name: Matthias Blume
2391    Date: 2002/05/20 10:59:00 EDT
2392    Tag: blume-20020520-preloads
2393    Description:
2394    
2395    Added a lot of commented-out suggestions for things to be included
2396    in config/preloads.
2397    
2398    ----------------------------------------------------------------------
2399    Name: Allen Leung
2400    Date: 2002/05/18 14:20:00 EDT
2401    Tag: leunga-20020518-mdl
2402    Description:
2403    
2404    o Made the mdl tool stuff compile and run again.
2405    o I've disabled all the stuff that depends on RTL specifications; they
2406      are all badly broken anyway.
2407    
2408    ----------------------------------------------------------------------
2409    Name: Matthias Blume
2410    Date: 2002/05/17 16:49:00 EDT
2411    Tag: blume-20020517-controls
2412    Description:
2413    
2414    0. John Reppy made several modifications to the SML/NJ library.
2415       In particular, there is a shiny new controls-lib.cm.
2416    
2417    1. Pushed new controls interface through compiler so that everything
2418       compiles again.
2419    
2420    2. Added FormatComb and FORMAT_COMB to the CML version of the
2421       SML/NJ library (so that CML compiles again).
2422    
2423    3. Modified init scripts because XXX_DEFAULT environment variables
2424       are no longer with us.  (Boot-time initialization is now done
2425       using the same environment variables that are also used for
2426       startup-time initialization of controls.)
2427    
2428    ----------------------------------------------------------------------
2429    Name: Lal George
2430    Date: 2002/05/15 09:20:10 EDT
2431    Tag: george-20020515-pseudo-op-decls
2432    Description:
2433    
2434            All pseudo-ops emitted before the first segment declaration
2435    such as TEXT, DATA, and BSS directives are assumed to be global
2436    declarations and are emitted first in the assembly file. This is
2437    useful in a number of situations where one has pseudo-ops that are not
2438    specific to any segment, and also works around the constraint that one
2439    cannot have client pseudo-ops in the TEXT segment.
2440    
2441            Because no segment is associated with these declarations it is
2442    an error to allocate any space or objects before the first segment
2443    directive and an exception will be raised. However, we cannot make
2444    this check for client pseudo-ops.
2445    
2446            These top level declarations are a field in the CFG graph_info.
2447    In theory you can continue to add to this field after the CFG has been
2448    built -- provided you know what you are doing;-)
2449    
2450    ----------------------------------------------------------------------
2451    Name: Matthias Blume
2452    Date: 2002/05/13 16:40:00 EDT
2453    Tag: blume-20020513-pp-etc
2454    Description:
2455    
2456    A few minor bugfixes:
2457    
2458      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2459        (Bogus printouts for redefined bindings still occur.  Compiler
2460        bug should no longer occur now.  We need to redo the prettyprinter
2461        from scratch.)
2462    
2463      - CM pathname printer now also adds escape sequences for ( and )
2464    
2465      - commend and docu fixes for ml-nlffi
2466    
2467    ----------------------------------------------------------------------
2468    Name: Matthias Blume
2469    Date: 2002/05/10 16:40:00 EDT
2470    Tag: blume-20020510-erg-textio
2471    Description:
2472    
2473    Applied the following bugfix provided by Emden Gansner:
2474    
2475        Output is corrupted when outputSubstr is used rather than output.
2476    
2477        The problem occurs when a substring
2478    
2479            ss = (s, dataStart, dataLen)
2480    
2481        where dataStart > 0, fills a stream buffer with avail bytes left.
2482        avail bytes of s, starting at index dataStart, are copied into the
2483        buffer, the buffer is flushed, and then the remaining dataLen-avail
2484        bytes of ss are copied into the beginning of the buffer. Instead of
2485        starting this copy at index dataStart+avail in s, the current code
2486        starts the copy at index avail.
2487    
2488        Fix:
2489        In text-io-fn.sml, change line 695 from
2490             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2491        to
2492             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2493    
2494    ----------------------------------------------------------------------
2495    Name: Matthias Blume
2496    Date: 2002/04/12 13:55:00 EDT
2497    Tag: blume-20020412-assyntax
2498    Description:
2499    
2500    1. Grabbed newer assyntax.h from the XFree86 project.
2501    2. Fiddled with how to compile X86.prim.asm without warnings.
2502    3. (Very) Minor cleanup in CM.
2503    
2504    ----------------------------------------------------------------------
2505    Name: Matthias Blume
2506    Date: 2002/04/01 (no joke!) 17:07:00 EST
2507    Tag: blume-20020401-x86div
2508    Description:
2509    
2510    Added full support for div/mod/rem/quot on the x86, using the machine
2511    instruction's two results (without clumsily recomputing the remainder)
2512    directly where appropriate.
2513    
2514    Some more extensive power-of-two support was added to the x86 instruction
2515    selector (avoiding expensive divs, mods, and muls where they can be
2516    replaced with cheaper shifts and masks).  However, this sort of thing
2517    ought to be done earlier, e.g., within the CPS optimizer so that
2518    all architectures benefit from it.
2519    
2520    The compiler compiles to a fixed point, but changes might be somewhat
2521    fragile nevertheless.  Please, report any strange things that you might
2522    see wrt. div/mod/quot/rem...
2523    
2524    ----------------------------------------------------------------------
2525    Name: Matthias Blume
2526    Date: 2002/03/29 17:22:00
2527    Tag: blume-20020329-div
2528    Description:
2529    
2530    Fixed my broken div/mod logic.  Unfortunately, this means that the
2531    inline code for div/mod now has one more comparison than before.
2532    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2533    The problem was with quotient = 0, because that alone does not tell
2534    us which way the rounding went.  One then has to look at whether
2535    remainder and divisor have the same sign...  :(
2536    
2537    Anyway, I replaced the bootfiles with fresh ones...
2538    
2539    ----------------------------------------------------------------------
2540    Name: Matthias Blume
2541    Date: 2002/03/29 14:10:00 EST
2542    Tag: blume-20020329-inlprims
2543    Description:
2544    
2545    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2546    
2547    Primops have changed. This means that the bin/boot-file formats have
2548    changed as well.
2549    
2550    To make sure that there is no confusion, I made a new version.
2551    
2552    
2553    CHANGES:
2554    
2555    * removed REMT from mltree (remainder should never overflow).
2556    
2557    * added primops to deal with divisions of all flavors to the frontend
2558    
2559    * handled these primops all the way through so they map to their respective
2560      MLRISC support
2561    
2562    * used these primops in the implementation of Int, Int32, Word, Word32
2563    
2564    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2565    
2566    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2567    
2568    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2569      floating point
2570    
2571    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2572      and Real (but Real.abs maps to a separate floating-point-only primop)
2573    
2574    
2575    TODO items:
2576    
2577    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2578      instructions because the machine instruction encoder cannot handle
2579      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2580      REM in the first place, and REM is supported by the x86 machine coder.)
2581    
2582    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2583      selection phase.  (The two can be streamlined because the hardware
2584      delivers both quotient and remainder at the same time anyway.)
2585    
2586    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2587      (Currently the behavior is inconsistent both across architectures and
2588      wrt. the draft Basis spec.)
2589    
2590    * Word8 should eventually be handled natively, too.
2591    
2592    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2593      as if there currently is no execution path that could trigger it in
2594      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2595      hold for things like multiplication and division.)
2596    
2597    ----------------------------------------------------------------------
2598    Name: Matthias Blume
2599    Date: 2002/03/27 16:27:00 EST
2600    Tag: blume-20020327-mlrisc-divisions
2601    Description:
2602    
2603    Added support for all four division operations (ML's div, mod, quot,
2604    and rem) to MLRISC.  In the course of doing so, I also rationalized
2605    the naming (no more annoying switch-around of DIV and QUOT), by
2606    parameterizing the operation by div_rounding_mode (which can be either
2607    DIV_TO_ZERO or DIV_TO_NEGINF).
2608    
2609    The generic MLTreeGen functor takes care of compiling all four
2610    operations down to only round-to-zero div.
2611    
2612    Missing pieces:
2613    
2614      * Doing something smarter than relying on MLTreeGen on architectures
2615        like, e.g., the x86 where hardware division delivers both quotient and
2616        remainder at the same time.  With this, the implementation of the
2617        round-to-neginf operations could be further streamlined.
2618    
2619      * Remove inlining support for div/mod/rem from the frontend and replace it
2620        with primops that get carried through to the backend.  Do this for all
2621        int and word types.
2622    
2623    ----------------------------------------------------------------------
2624    Name: Matthias Blume
2625    Date: 2002/03/25 17:25:00 EST
2626    Tag: blume-20020325-divmod
2627    Description:
2628    
2629    I improved (hopefully without breaking them) the implementation of Int.div,
2630    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2631    advantage of the following observations:
2632    
2633      Let  q = x quot y      r = x rem y
2634           d = x div  y      m = x mod y
2635    
2636    where "quot" is the round-to-zero version of integer division that
2637    hardware usually provides.  Then we have:
2638    
2639         r = x - q * y        where neither the * nor the - will overflow
2640         d = if q >= 0 orelse x = q * y then q else q - 1
2641                              where neither the * nor the - will overflow
2642         m = if q >= 0 orelse r = 0 then r else r + y
2643                              where the + will not overflow
2644    
2645    This results in substantial simplification of the generated code.
2646    The following table shows the number of CFG nodes and edges generated
2647    for
2648            fun f (x, y) = x OPER y
2649            (* with OPER \in div, mod, quot, rem *)
2650    
2651    
2652        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2653        --------------------------------------------------------
2654         div |         24 |         39 |         12 |         16
2655         mod |         41 |         71 |         12 |         16
2656        quot |          8 |         10 |          8 |         10
2657         rem |         10 |         14 |          8 |         10
2658    
2659    
2660    ----------------------------------------------------------------------
2661    Name: Matthias Blume
2662    Date: 2002/03/25 22:06:00 EST
2663    Tag: blume-20020325-cprotobug
2664    Description:
2665    
2666    Fixed a bug in cproto (c prototype decoder).
2667    
2668    ----------------------------------------------------------------------
2669    Name: Matthias Blume
2670    Date: 2002/03/25 16:00:00 EST
2671    Tag: blume-20020325-raw-primops
2672    Description:
2673    
2674    I did some cleanup to Allen's new primop code and
2675    replaced yesterday's bootfiles with new ones.
2676    (But they are stored in the same place.)
2677    
2678    ----------------------------------------------------------------------
2679    Name: Matthias Blume
2680    Date: 2002/03/24 22:40:00 EST
2681    Tag: blume-20020324-bootfiles
2682    Description:
2683    
2684    Made the bootfiles that Allen asked for.
2685    
2686    ----------------------------------------------------------------------
2687    Name: Allen Leung
2688    Date: 2002/03/23 15:50:00 EST
2689    Tag: leunga-20020323-flint-cps-rcc-primops
2690    Description:
2691    
2692      1. Changes to FLINT primops:
2693    
2694        (* make a call to a C-function;
2695         * The primop carries C function prototype information and specifies
2696         * which of its (ML-) arguments are floating point. C prototype
2697         * information is for use by the backend, ML information is for
2698         * use by the CPS converter. *)
2699      | RAW_CCALL of { c_proto: CTypes.c_proto,
2700                       ml_args: ccall_type list,
2701                       ml_res_opt: ccall_type option,
2702                       reentrant : bool
2703                     } option
2704       (* Allocate uninitialized storage on the heap.
2705        * The record is meant to hold short-lived C objects, i.e., they
2706        * are not ML pointers.  With the tag, the representation is
2707        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2708        *)
2709      | RAW_RECORD of {tag:bool,sz:int}
2710      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2711    
2712      2.  These CPS primops are now overloaded:
2713    
2714           rawload of {kind:numkind}
2715           rawstore of {kind:numkind}
2716    
2717          The one argument form is:
2718    
2719             rawload {kind} address
2720    
2721          The two argument form is:
2722    
2723             rawload {kind} [ml object, byte-offset]
2724    
2725      3. RAW_CCALL/RCC now takes two extra arguments:
2726    
2727         a. The first is whether the C call is reentrant, i.e., whether
2728            ML state should be saved and restored.
2729         b. The second argument is a string argument specifying the name of
2730            library and the C function.
2731    
2732         These things are currently not handled in the code generator, yet.
2733    
2734      4. In CProto,
2735    
2736         An encoding type of "bool" means "ml object" and is mapped into
2737         C prototype of PTR.  Note that "bool" is different than "string",
2738         even though "string" is also mapped into PTR, because "bool"
2739         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2740    
2741      5. Pickler/unpicker
2742    
2743         Changed to handle RAW_RECORD and newest RAW_CCALL
2744    
2745      6. MLRiscGen,
2746    
2747         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2748         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2749            in the file CodeGen/cpscompile/cps-c-calls.sml
2750    
2751      7. Added the conditional move operator
2752    
2753             condmove of branch
2754    
2755         to cps.  Generation of this is still buggy so it is currently
2756         disabled.
2757    
2758    ----------------------------------------------------------------------
2759    Name: Lal George
2760    Date: 2002/03/22 14:18:25 EST
2761    Tag: george-20020322-cps-branch-prob
2762    Description:
2763    
2764    Implemented the Ball-Larus branch prediction-heuristics, and
2765    incorporated graphical viewers for control flow graphs.
2766    
2767    Ball-Larus Heuristics:
2768    ---------------------
2769    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
2770    
2771    By design it uses the Dempster-Shafer theory for combining
2772    probabilities.  For example, in the function:
2773    
2774        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
2775    
2776    the ball-larus heuristics predicts that the n=0 is unlikely
2777    (OH-heuristic), and the 'then' branch is unlikely because of the
2778    RH-heuristic -- giving the 'then' branch an even lower combined
2779    probability using the Dempster-Shafer theory.
2780    
2781    Finally, John Reppy's loop analysis in MLRISC, further lowers the
2782    probability of the 'then' branch because of the loop in the else
2783    branch.
2784    
2785    
2786    Graphical Viewing:
2787    ------------------
2788    I merely plugged in Allen's graphical viewers into the compiler. The
2789    additional code is not much. At the top level, saying:
2790    
2791            Control.MLRISC.getFlag "cfg-graphical-view" := true;
2792    
2793    will display the graphical view of the control flow graph just before
2794    back-patching.  daVinci must be in your path for this to work. If
2795    daVinci is not available, then the default viewer can be changed
2796    using:
2797    
2798            Control.MLRISC.getString "viewer"
2799    
2800    which can be set to "dot" or "vcg" for the corresponding viewers. Of
2801    course, these viewers must be in your path.
2802    
2803    The above will display the compilation unit at the level of clusters,
2804    many of which are small, boring, and un-interesting. Also setting:
2805    
2806            Control.MLRISC.getInt "cfg-graphical-view_size"
2807    
2808    will display clusters that are larger than the value set by the above.
2809    
2810    
2811    ----------------------------------------------------------------------
2812    Name: Matthias Blume
2813    Date: 2002/03/21 22:20:00 EST
2814    Tag: blume-20020321-kmp-bugfix
2815    Description:
2816    
2817    Changed the interface to the KMP routine in PreString and fixed
2818    a minor bug in one place where it was used.
2819    
2820    ----------------------------------------------------------------------
2821    Name: Allen Leung
2822    Date: 2002/03/21 20:30:00 EST
2823    Tag: leunga-20020321-cfg
2824    Description:
2825    
2826      Fixed a potential problem in cfg edge splitting.
2827    
2828    ----------------------------------------------------------------------
2829    Name: Allen Leung
2830    Date: 2002/03/21 17:15:00 EST
2831    Tag: leunga-20020321-x86-fp-cfg
2832    Description:
2833    
2834      1. Recoded the buggy parts of x86-fp.
2835    
2836         a. All the block reordering code has been removed.
2837            We now depend on the block placement phases to do this work.
2838    
2839         b. Critical edge splitting code has been simplified and moved into the
2840            CFG modules, as where they belong.
2841    
2842         Both of these were quite buggy and complex.  The code is now much, much
2843         simpler.
2844    
2845      2. X86 backend.
2846    
2847         a. Added instructions for 64-bit support.  Instruction selection for
2848            64-bit has not been committed, however, since that
2849            requires changes to MLTREE which haven't been approved by
2850            Lal and John.
2851    
2852         b. Added support for FUCOMI and FUCOMIP when generating code for
2853            PentiumPro and above.  We only generate these instructions in
2854            the fast-fp mode.
2855    
2856         c. Added cases for JP and JNP in X86FreqProps.
2857    
2858      3. CFG
2859    
2860         CFG now has a bunch of methods for edge splitting and merging.
2861    
2862      4. Machine description.
2863    
2864         John's simplification of MLTREE_BASIS.fcond broke a few machine
2865         description things:
2866    
2867         rtl-build.{sig,sml} and hppa.mdl fixed.
2868    
2869         NOTE: the machine description stuff in the repository is still broken.
2870               Again, I can't put my fixes in because that involves
2871               changes to MLTREE.
2872    
2873    ----------------------------------------------------------------------
2874    Name: Matthias Blume
2875    Date: 2002/03/20 15:55:00 EST
2876    Tag: blume-20020320-kmp
2877    Description:
2878    
2879    Implemented Knuth-Morris-Pratt string matching in PreString and used
2880    it for String.isSubstring, Substring.isSubstring, and
2881    Substring.position.
2882    
2883    (Might need some stress-testing.  Simple examples worked fine.)
2884    
2885    ----------------------------------------------------------------------
2886    Name: Matthias Blume
2887    Date: 2002/03/19 16:37:00 EST
2888    Tag: blume-20020319-witnesses
2889    Description:
2890    
2891    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
2892    
2893    This implements a generic mechanism for changing constness qualifiers
2894    anywhere within big C types without resorting to outright "casts".
2895    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
2896    modify the constness at the outermost level.)
2897    The implementation of "convert" is based on the idea of "witness"
2898    values -- values that are not used by the operation but whose types
2899    "testify" to their applicability.  On the implementation side, "convert"
2900    is simply a projection (returning its second curried argument).  With
2901    cross-module inlining, it should not result in any machine code being
2902    generated.
2903    
2904    ----------------------------------------------------------------------
2905    Name: Matthias Blume
2906    Date: 2002/03/15 16:40:00 EST
2907    Tag: blume-20020315-basis
2908    Description:
2909    
2910    Provided (preliminary?) implementations for
2911    
2912      {String,Substring}.{concatWith,isSuffix,isSubstring}
2913    
2914    and
2915    
2916      Substring.full
2917    
2918    Those are in the Basis spec but they were missing in SML/NJ.
2919    
2920    ----------------------------------------------------------------------
2921    Name: Matthias Blume
2922    Date: 2002/03/14 21:30:00 EST
2923    Tag: blume-20020314-controls
2924    Description:
2925    
2926    Controls:
2927    ---------
2928    
2929    1. Factored out the recently-added Controls : CONTROLS stuff and put
2930       it into its own library $/controls-lib.cm.  The source tree for
2931       this is under src/smlnj-lib/Controls.
2932    
2933    2. Changed the names of types and functions in this interface, so they
2934       make a bit more "sense":
2935    
2936          module -> registry
2937          'a registry -> 'a group
2938    
2939    3. The interface now deals in ref cells only.  The getter/setter interface
2940       is (mostly) gone.
2941    
2942    4. Added a function that lets one register an already-existing ref cell.
2943    
2944    5. Made the corresponding modifications to the rest of the code so that
2945       everything compiles again.
2946    
2947    6. Changed the implementation of Controls.MLRISC back to something closer
2948       to the original.  In particular, this module (and therefore MLRISC)
2949       does not depend on Controls.  There now is some link-time code in
2950       int-sys.sml that registers the MLRISC controls with the Controls
2951       module.
2952    
2953    CM:
2954    ---
2955    
2956      * One can now specify the lambda-split aggressiveness in init.cmi.
2957    
2958    ----------------------------------------------------------------------
2959    Name: Allen Leung
2960    Date: 2002/03/13 17:30:00 EST
2961    Tag: leunga-20020313-x86-fp-unary
2962    Description:
2963    
2964    Bug fix for:
2965    
2966    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
2967    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
2968    > - fun f(x,(y,z)) = Real.~ y;
2969    > [autoloading]
2970    > [autoloading done]
2971    >       fchsl   (%eax), 184(%esp)
2972    > Error: MLRisc bug: X86MCEmitter.emitInstr
2973    >
2974    > uncaught exception Error
2975    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
2976    
2977    The problem was that the code generator did not generate any fp registers
2978    in this case, and the ra didn't know that it needed to run the X86FP phase to
2979    translate the pseudo fp instruction.   This only happened with unary fp
2980    operators in certain situations.
2981    
2982    ----------------------------------------------------------------------
2983    Name: Matthias Blume
2984    Date: 2002/03/13 14:00:00 EST
2985    Tag: blume-20020313-overload-etc
2986    Description:
2987    
2988    1. Added _overload as a synonym for overload for backward compatibility.
2989       (Control.overloadKW must be true for either version to be accepted.)
2990    
2991    2. Fixed bug in install script that caused more things to be installed
2992       than what was requested in config/targets.
2993    
2994    3. Made CM aware of the (_)overload construct so that autoloading
2995       works.
2996    
2997    ----------------------------------------------------------------------
2998    Name: Matthias Blume
2999    Date: 2002/03/12 22:03:00 EST
3000    Tag: blume-20020312-url
3001    Description:
3002    
3003    Forgot to update BOOT and srcarchiveurl.
3004    
3005    ----------------------------------------------------------------------
3006    Name: Matthias Blume
3007    Date: 2002/03/12 17:30:00 EST
3008    Tag: blume-20020312-version110392
3009    Description:
3010    
3011    Yet another version number bump (because of small changes to the
3012    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
3013    
3014    Changes:
3015    
3016      The new pid generation scheme described a few weeks ago was overly
3017      complicated.  I implemented a new mechanism that is simpler and
3018      provides a bit more "stability":  Once CM has seen a compilation
3019      unit, it keeps its identity constant (as long as you do not delete
3020      those crucial CM/GUID/* files).  This means that when you change
3021      an interface, compile, then go back to the old interface, and
3022      compile again, you arrive at the original pid.
3023    
3024      There now also is a mechanism that instructs CM to use the plain
3025      environment hash as a module's pid (effectively making its GUID
3026      the empty string).  For this, "noguid" must be specified as an
3027      option to the .sml file in question within its .cm file.
3028      This is most useful for code that is being generated by tools such
3029      as ml-nlffigen (because during development programmers tend to
3030      erase the tool's entire output directory tree including CM's cached
3031      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
3032      revert to the old, broken behavior of SML/NJ, but in specific cases
3033      where there is no danger of interface confusion, its use is ok
3034      (I think).
3035    
3036      ml-nlffigen by default generates "noguid" annotations.  They can be
3037      turned off by specifying -guid in its command line.
3038    
3039    ----------------------------------------------------------------------
3040    Name: Lal George
3041    Date: 2002/03/12 12 14:42:36 EST
3042    Tag: george-20020312-frequency-computation
3043    Description:
3044    
3045    Integrated jump chaining and static block frequency into the
3046    compiler. More details and numbers later.
3047    
3048    ----------------------------------------------------------------------
3049    Name: Lal George
3050    Date: 2002/03/11 11 22:38:53 EST
3051    Tag: george-20020311-jump-chain-elim
3052    Description:
3053    
3054    Tested the jump chain elimination on all architectures (except the
3055    hppa).  This is on by default right now and is profitable for the
3056    alpha and x86, however, it may not be profitable for the sparc and ppc
3057    when compiling the compiler.
3058    
3059    The gc test will typically jump to a label at the end of the cluster,
3060    where there is another jump to an external cluster containing the actual
3061    code to invoke gc. This is to allow factoring of common gc invocation
3062    sequences. That is to say, we generate:
3063    
3064            f:
3065               testgc
3066               ja   L1      % jump if above to L1
3067    
3068            L1:
3069               jmp L2
3070    
3071    
3072    After jump chain elimination the 'ja L1' instructions is converted to
3073    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
3074    up being implemented in their long form (if L2 is far away) using:
3075    
3076            jbe     L3      % jump if below or equal to L3
3077            jmp     L2
3078         L3:
3079            ...
3080    
3081    
3082    For large compilation units L2  may be far away.
3083    
3084    
3085    ----------------------------------------------------------------------
3086    Name: Matthias Blume
3087    Date: 2002/03/11 13:30:00 EST
3088    Tag: blume-20020311-mltreeeval
3089    Description:
3090    
3091    A functor parameter was missing.
3092    
3093    ----------------------------------------------------------------------
3094    Name: Allen Leung
3095    Date: 2002/03/11 10:30:00 EST
3096    Tag: leunga-20020311-runtime-string0
3097    Description:
3098    
3099       The representation of the empty string now points to a
3100    legal null terminated C string instead of unit.  It is now possible
3101    to convert an ML string into C string with InlineT.CharVector.getData.
3102    This compiles into one single machine instruction.
3103    
3104    ----------------------------------------------------------------------
3105    Name: Allen Leung
3106    Date: 2002/03/10 23:55:00 EST
3107    Tag: leunga-20020310-x86-call
3108    Description:
3109    
3110       Added machine generation for CALL instruction (relative displacement mode)
3111    
3112    ----------------------------------------------------------------------
3113    Name: Matthias Blume
3114    Date: 2002/03/08 16:05:00
3115    Tag: blume-20020308-entrypoints
3116    Description:
3117    
3118    Version number bumped to 110.39.1.  NEW BOOTFILES!
3119    
3120    Entrypoints: non-zero offset into a code object where execution should begin.
3121    
3122    - Added the notion of an entrypoint to CodeObj.
3123    - Added reading/writing of entrypoint info to Binfile.
3124    - Made runtime system bootloader aware of entrypoints.
3125    - Use the address of the label of the first function given to mlriscGen
3126      as the entrypoint.  This address is currently always 0, but it will
3127      not be 0 once we turn on block placement.
3128    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3129      with entry points) from mlriscGen.
3130    
3131    ----------------------------------------------------------------------
3132    Name: Allen Leung
3133    Date: 2002/03/07 20:45:00 EST
3134    Tag: leunga-20020307-x86-cmov
3135    Description:
3136    
3137       Bug fixes for CMOVcc on x86.
3138    
3139       1. Added machine code generation for CMOVcc
3140       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3141       3. CMOVcc cannot have an immediate operand as argument.
3142    
3143    ----------------------------------------------------------------------
3144    Name: Matthias Blume
3145    Date: 2002/03/07 16:15:00 EST
3146    Tag: blume-20020307-controls
3147    Description:
3148    
3149    This is a very large but mostly boring patch which makes (almost)
3150    every tuneable compiler knob (i.e., pretty much everything under
3151    Control.* plus a few other things) configurable via both the command
3152    line and environment variables in the style CM did its configuration
3153    until now.
3154    
3155    Try starting sml with '-h' (or, if you are brave, '-H')
3156    
3157    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3158    implements the underlying generic mechanism.
3159    
3160    The interface to some of the existing such facilities has changed somewhat.
3161    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3162    (The getFoo interface is still there for backward-compatibility, but its
3163    use is deprecated.)
3164    
3165    The ml-build script passes -Cxxx=yyy command-line arguments through so
3166    that one can now twiddle the compiler settings when using this "batch"
3167    compiler.
3168    
3169    TODO items:
3170    
3171    We should go through and throw out all controls that are no longer
3172    connected to anything.  Moreover, we should go through and provide
3173    meaningful (and correct!) documentation strings for those controls
3174    that still are connected.
3175    
3176    Currently, multiple calls to Controls.new are accepted (only the first
3177    has any effect).  Eventually we should make sure that every control
3178    is being made (via Controls.new) exactly once.  Future access can then
3179    be done using Controls.acc.
3180    
3181    Finally, it would probably be a good idea to use the getter-setter
3182    interface to controls rather than ref cells.  For the time being, both
3183    styles are provided by the Controls module, but getter-setter pairs are
3184    better if thread-safety is of any concern because they can be wrapped.
3185    
3186    *****************************************
3187    
3188    One bug fix: The function blockPlacement in three of the MLRISC
3189    backpatch files used to be hard-wired to one of two possibilities at
3190    link time (according to the value of the placementFlag).  But (I
3191    think) it should rather sense the flag every time.
3192    
3193    *****************************************
3194    
3195    Other assorted changes (by other people who did not supply a HISTORY entry):
3196    
3197    1. the cross-module inliner now works much better (Monnier)
3198    2. representation of weights, frequencies, and probabilities in MLRISC
3199       changed in preparation of using those for weighted block placement
3200       (Reppy, George)
3201    
3202    ----------------------------------------------------------------------
3203    Name: Lal George
3204    Date: 2002/03/07 14:44:24 EST 2002
3205    Tag: george-20020307-weighted-block-placement
3206    
3207    Tested the weighted block placement optimization on all architectures
3208    (except the hppa) using AMPL to generate the block and edge frequencies.
3209    Changes were required in the machine properties to correctly
3210    categorize trap instructions. There is an MLRISC flag
3211    "weighted-block-placement" that can be used to enable weighted block
3212    placement, but this will be ineffective without block/edge
3213    frequencies (coming soon).
3214    
3215    
3216    ----------------------------------------------------------------------
3217    Name: Lal George
3218    Date: 2002/03/05 17:24:48 EST
3219    Tag: george-20020305-linkage-cluster
3220    
3221    In order to support the block placement optimization, a new cluster
3222    is generated as the very first cluster (called the linkage cluster).
3223    It contains a single jump to the 'real' entry point for the compilation
3224    unit. Block placement has no effect on the linkage cluster itself, but
3225    all the other clusters  have full freedom in the manner in which they
3226    reorder blocks or functions.
3227    
3228    On the x86 the typical linkage code that is generated is:
3229       ----------------------
3230            .align 2
3231       L0:
3232            addl    $L1-L0, 72(%esp)
3233            jmp     L1
3234    
3235    
3236            .align  2
3237       L1:
3238       ----------------------
3239    
3240    72(%esp) is the memory location for the stdlink register. This
3241    must contain the address of the CPS function being called. In the
3242    above example, it contains the address of  L0; before
3243    calling L1 (the real entry point for the compilation unit), it
3244    must contain the address for L1, and hence
3245    
3246            addl $L1-L0, 72(%esp)
3247    
3248    I have tested this on all architectures except the hppa.The increase
3249    in code size is of course negligible
3250    
3251    ----------------------------------------------------------------------
3252    Name: Allen Leung
3253    Date: 2002/03/03 13:20:00 EST
3254    Tag: leunga-20020303-mlrisc-tools
3255    
3256      Added #[ ... ] expressions to mlrisc tools
3257    
3258    ----------------------------------------------------------------------
3259    Name: Matthias Blume
3260    Date: 2002/02/27 12:29:00 EST
3261    Tag: blume-20020227-cdebug
3262    Description:
3263    
3264    - made types in structure C and C_Debug to be equal
3265    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3266    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3267    
3268    ----------------------------------------------------------------------
3269    Name: Matthias Blume
3270    Date: 2002/02/26 12:00:00 EST
3271    Tag: blume-20020226-ffi
3272    Description:
3273    
3274    1. Fixed a minor bug in CM's "noweb" tool:
3275       If numbering is turned off, then truly don't number (i.e., do not
3276       supply the -L option to noweb).  The previous behavior was to supply
3277       -L'' -- which caused noweb to use the "default" line numbering scheme.
3278       Thanks to Chris Richards for pointing this out (and supplying the fix).
3279    
3280    2. Once again, I reworked some aspects of the FFI:
3281    
3282       A. The incomplete/complete type business:
3283    
3284       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3285         gone!
3286       - ML types representing an incomplete type are now *equal* to
3287         ML types representing their corresponding complete types (just like
3288         in C).  This is still safe because ml-nlffigen will not generate
3289         RTTI for incomplete types, nor will it generate functions that
3290         require access to such RTTI.   But when ML code generated from both
3291         incomplete and complete versions of the C type meet, the ML types
3292         are trivially interoperable.
3293    
3294         NOTE:  These changes restore the full generality of the translation
3295         (which was previously lost when I eliminated functorization)!
3296    
3297       B. Enum types:
3298    
3299       - Structure C now has a type constructor "enum" that is similar to
3300         how the "su" constructor works.  However, "enum" is not a phantom
3301         type because each "T enum" has values (and is isomorphic to
3302         MLRep.Signed.int).
3303       - There are generic access operations for enum objects (using
3304         MLRep.Signed.int).
3305       - ml-nlffigen will generate a structure E_foo for each "enum foo".
3306         * The structure contains the definition of type "mlrep" (the ML-side
3307         representation type of the enum).  Normally, mlrep is the same
3308         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3309         then mlrep will be defined as a datatype -- thus facilitating
3310         pattern matching on mlrep values.
3311         ("-ec" will be suppressed if there are duplicate values in an
3312          enumeration.)
3313         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3314         will be generated for each C enum constant xxx.
3315         * Conversion functions m2i and i2m convert between mlrep and
3316         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
3317         * Coversion functions c and ml convert between mlrep and "tag enum".
3318         * Access functions (get/set) fetch and store mlrep values.
3319       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3320         enumerations are merged into one single enumeration represented by
3321         structure E_'.
3322    
3323    ----------------------------------------------------------------------
3324    Name: Allen Leung
3325    Date: 2002/02/25 04:45:00 EST
3326    Tag: leunga-20020225-cps-spill
3327    
3328    This is a new implementation of the CPS spill phase.
3329    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3330    In case of problems, replace it with the old file spill.sml
3331    
3332    The current compiler runs into some serious performance problems when
3333    constructing a large record.  This can happen when we try to compile a
3334    structure with many items.  Even a very simple structure like the following
3335    makes the compiler slow down.
3336    
3337        structure Foo = struct
3338           val x_1 = 0w1 : Word32.int
3339           val x_2 = 0w2 : Word32.int
3340           val x_3 = 0w3 : Word32.int
3341           ...
3342           val x_N = 0wN : Word32.int
3343        end
3344    
3345    The following table shows the compile time, from N=1000 to N=4000,
3346    with the old compiler:
3347    
3348    N
3349    1000   CPS 100 spill                           0.04u  0.00s  0.00g
3350           MLRISC ra                               0.06u  0.00s  0.05g
3351              (spills = 0 reloads = 0)
3352           TOTAL                                   0.63u  0.07s  0.21g
3353    
3354    1100   CPS 100 spill                           8.25u  0.32s  0.64g
3355           MLRISC ra                               5.68u  0.59s  3.93g
3356              (spills = 0 reloads = 0)
3357           TOTAL                                   14.71u  0.99s  4.81g
3358    
3359    1500   CPS 100 spill                           58.55u  2.34s  1.74g
3360           MLRISC ra                               5.54u  0.65s  3.91g
3361              (spills = 543 reloads = 1082)
3362           TOTAL                                   65.40u  3.13s  6.00g
3363    
3364    2000   CPS 100 spill                           126.69u  4.84s  3.08g
3365           MLRISC ra                               0.80u  0.10s  0.55g
3366              (spills = 42 reloads = 84)
3367           TOTAL                                   129.42u  5.10s  4.13g
3368    
3369    3000   CPS 100 spill                           675.59u  19.03s  11.64g
3370           MLRISC ra                               2.69u  0.27s  1.38g
3371              (spills = 62 reloads = 124)
3372           TOTAL                                   682.48u  19.61s  13.99g
3373    
3374    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
3375           MLRISC ra                               4.96u  0.27s  2.72g
3376              (spills = 85 reloads = 170)
3377           TOTAL                                   2375.26u  57.21s  48.00g
3378    
3379    As you can see the old cps spill module suffers from some serious
3380    performance problem.  But since I cannot decipher the old code fully,
3381    instead of patching the problems up, I'm reimplementing it
3382    with a different algorithm.  The new code is more modular,
3383    smaller when compiled, and substantially faster
3384    (O(n log n) time and O(n) space).  Timing of the new spill module:
3385    
3386    4000  CPS 100 spill                           0.02u  0.00s  0.00g
3387          MLRISC ra                               0.25u  0.02s  0.15g
3388             (spills=1 reloads=3)
3389          TOTAL                                   7.74u  0.34s  1.62g
3390    
3391    Implementation details:
3392    
3393    As far as I can tell, the purpose of the CPS spill module is to make sure the
3394    number of live variables at any program point (the bandwidth)
3395    does not exceed a certain limit, which is determined by the
3396    size of the spill area.
3397    
3398    When the bandwidth is too large, we decrease the register pressure by
3399    packing live variables into spill records.  How we achieve this is
3400    completely different than what we did in the old code.
3401    
3402    First, there is something about the MLRiscGen code generator
3403    that we should be aware of:
3404    
3405    o MLRiscGen performs code motion!
3406    
3407       In particular, it will move floating point computations and
3408       address computations involving only the heap pointer to
3409       their use sites (if there is only a single use).
3410       What this means is that if we have a CPS record construction
3411       statement
3412    
3413           RECORD(k,vl,w,e)
3414    
3415       we should never count the new record address w as live if w
3416       has only one use (which is often the case).
3417    
3418       We should do something similar to floating point, but the transformation
3419       there is much more complex, so I won't deal with that.
3420    
3421    Secondly, there are now two new cps primops at our disposal:
3422    
3423     1. rawrecord of record_kind option
3424        This pure operator allocates some uninitialized storage from the heap.
3425        There are two forms:
3426    
3427         rawrecord NONE [INT n]  allocates a tagless record of length n
3428         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
3429                                     and initializes the tag.
3430    
3431     2. rawupdate of cty
3432          rawupdate cty (v,i,x)
3433          Assigns to x to the ith component of record v.
3434          The storelist is not updated.
3435    
3436    We use these new primops for both spilling and increment record construction.
3437    
3438     1. Spilling.
3439    
3440        This is implemented with a linear scan algorithm (but generalized
3441        to trees).  The algorithm will create a single spill record at the
3442        beginning of the cps function and use rawupdate to spill to it,
3443        and SELECT or SELp to reload from it.  So both spills and reloads
3444        are fine-grain operations.  In contrast, in the old algorithm
3445        "spills" have to be bundled together in records.
3446    
3447        Ideally, we should sink the spill record construction to where
3448        it is needed.  We can even split the spill record into multiple ones
3449        at the places where they are needed.  But CPS is not a good
3450        representation for global code motion, so I'll keep it simple and
3451        am not attempting this.
3452    
3453     2. Incremental record construction (aka record splitting).
3454    
3455        Long records with many component values which are simulatenously live
3456        (recall that single use record addresses are not considered to
3457         be live) are constructed with rawrecord and rawupdate.
3458        We allocate space on the heap with rawrecord first, then gradually
3459        fill it in with rawupdate.  This is the technique suggested to me
3460        by Matthias.
3461    
3462        Some restrictions on when this is applicable:
3463        1. It is not a VECTOR record.  The code generator currently does not handle
3464           this case. VECTOR record uses double indirection like arrays.
3465        2. All the record component values are defined in the same "basic block"
3466           as the record constructor.  This is to prevent speculative
3467           record construction.
3468    
3469    ----------------------------------------------------------------------
3470    Name: Allen Leung
3471    Date: 2002/02/22 01:02:00 EST
3472    Tag: leunga-20020222-mlrisc-tools
3473    
3474    Minor bug fixes in the parser and rewriter
3475    
3476    ----------------------------------------------------------------------
3477    Name: Allen Leung
3478    Date: 2002/02/21 20:20:00 EST
3479    Tag: leunga-20020221-peephole
3480    
3481    Regenerated the peephole files.  Some contained typos in the specification
3482    and some didn't compile because of pretty printing bugs in the old version
3483    of 'nowhere'.
3484    
3485    ----------------------------------------------------------------------
3486    Name: Allen Leung
3487    Date: 2002/02/19 20:20:00 EST
3488    Tag: leunga-20020219-mlrisc-tools
3489    Description:
3490    
3491       Minor bug fixes to the mlrisc-tools library:
3492    
3493       1.  Fixed up parsing colon suffixed keywords
3494       2.  Added the ability to shut the error messages up
3495       3.  Reimplemented the pretty printer and fixed up/improved
3496           the pretty printing of handle and -> types.
3497       4.  Fixed up generation of literal symbols in the nowhere tool.
3498       5.  Added some SML keywords to to sml.sty
3499    
3500    ----------------------------------------------------------------------
3501    Name: Matthias Blume
3502    Date: 2002/02/19 16:20:00 EST
3503    Tag: blume-20020219-cmffi
3504    Description:
3505    
3506    A wild mix of changes, some minor, some major:
3507    
3508    * All C FFI-related libraries are now anchored under $c:
3509        $/c.cm      --> $c/c.cm
3510        $/c-int.cm  --> $c/internals/c-int.cm
3511        $/memory.cm --> $c/memory/memory.cm
3512    
3513    * "make" tool (in CM) now treats its argument pathname slightly
3514      differently:
3515        1. If the native expansion is an absolute name, then before invoking
3516           the "make" command on it, CM will apply OS.Path.mkRelative
3517           (with relativeTo = OS.FileSys.getDir()) to it.
3518        2. The argument will be passed through to subsequent phases of CM
3519           processing without "going native".  In particular, if the argument
3520           was an anchored path, then "make" will not lose track of that anchor.
3521    
3522    * Compiler backends now "know" their respective C calling conventions
3523      instead of having to be told about it by ml-nlffigen.  This relieves
3524      ml-nlffigen from one of its burdens.
3525    
3526    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
3527    
3528    * Export C_DEBUG and C_Debug from $c/c.cm.
3529    
3530    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
3531      subtyping relationship between incomplete pointers and their complete
3532      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
3533      with the convention of instantiating 'o with ('t, 'c) obj whenever
3534      the pointer target type is complete.  In the incomplete case, 'o
3535      will be instantiated with some "'c iobj" -- a type obtained by
3536      using one of the functors PointerToIncompleteType or PointerToCompleteType.
3537    
3538      Operations that work on both incomplete and complete pointer types are
3539      typed as taking an 'o ptr while operations that require the target to
3540      be known are typed as taking some ('t, 'c) obj ptr.
3541    
3542      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
3543      where void is an eqtype without any values.  This makes it possible
3544      to work on voidptr values using functions meant to operate on light
3545      incomplete pointers.
3546    
3547    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
3548      been vastly simplified.
3549    
3550    ----------------------------------------------------------------------
3551    Name: Matthias Blume
3552    Date: 2002/02/19 10:48:00 EST
3553    Tag: blume-20020219-pqfix
3554    Description:
3555    
3556    Applied Chris Okasaki's bug fix for priority queues.
3557    
3558    ----------------------------------------------------------------------
3559    Name: Matthias Blume
3560    Date: 2002/02/15 17:05:00
3561    Tag: Release_110_39
3562    Description:
3563    
3564    Last-minute retagging is becoming a tradition... :-(
3565    
3566    This is the working release 110.39.
3567    
3568    ----------------------------------------------------------------------
3569    Name: Matthias Blume
3570    Date: 2002/02/15 16:00:00 EST
3571    Tag: Release_110_39-orig
3572    Description:
3573    
3574    Working release 110.39.  New bootfiles.
3575    
3576    (Update: There was a small bug in the installer so it wouldn't work
3577    with all shells.  So I retagged. -Matthias)
3578    
3579    ----------------------------------------------------------------------
3580    Name: Matthias Blume
3581    Date: 2002/02/15 14:17:00 EST
3582    Tag: blume-20020215-showbindings
3583    Description:
3584    
3585    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
3586    the latter can be useful for exploring what bindings are available at
3587    the interactive prompt.  (The first function returns only the list
3588    of symbols that are really bound, the second prints those but also the
3589    ones that CM's autoloading mechanism knows about.)
3590    
3591    ----------------------------------------------------------------------
3592    Name: Matthias Blume
3593    Date: 2002/02/15 12:08:00 EST
3594    Tag: blume-20020215-iptrs
3595    Description:
3596    
3597    Two improvements to ml-nlffigen:
3598    
3599      1. Write files only if they do not exist or if their current contents
3600         do not coincide with what's being written.  (That is, avoid messing
3601         with the time stamps unless absolutely necessary.)
3602    
3603      2. Implement a "repository" mechanism for generated files related
3604         to "incomplete pointer types".   See the README file for details.
3605    
3606    ----------------------------------------------------------------------
3607    Name: Matthias Blume
3608    Date: 2002/02/14 11:50:00 EST
3609    Tag: blume-20020214-quote
3610    Description:
3611    
3612    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
3613    because of the new and improved tag generation scheme.  (Thanks to Allen
3614    Leung for pointing it out.)
3615    
3616    ----------------------------------------------------------------------
3617    Name: Lal George
3618    Date: 2002/02/14 09:55:27 EST 2002
3619    Tag: george-20020214-isabelle-bug
3620    Description:
3621    
3622    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
3623    of Isabelle on the x86.
3624    
3625    From Allen:
3626    -----------
3627     I've found the problem:
3628    
3629         in ra-core.sml, I use the counter "blocked" to keep track of the
3630         true number of elements in the freeze queue.  When the counter goes
3631         to zero, I skip examining the queue.  But I've messed up the
3632         bookkeeping in combine():
3633    
3634             else ();
3635             case !ucol of
3636               PSEUDO => (if !cntv > 0 then
3637                     (if !cntu > 0 then blocked := !blocked - 1 else ();
3638                                        ^^^^^^^^^^^^^^^^^^^^^^^
3639                      moveu := mergeMoveList(!movev, !moveu)
3640                     )
3641                  else ();
3642    
3643         combine() is called to coalesce two nodes u and v.
3644         I think I was thinking that if the move counts of u and v are both
3645         greater than zero then after they are coalesced then one node is
3646         removed from the freeze queue.  Apparently I was thinking that
3647         both u and v are of low degree, but that's clearly not necessarily true.
3648    
3649    
3650    02/12/2002:
3651        Here's the patch.  HOL now compiles.
3652    
3653        I don't know how this impact on performance (compile
3654        time or runtime).  This bug caused the RA (especially on the x86)
3655        to go thru the potential spill phase when there are still nodes on the
3656        freeze queue.
3657    
3658    
3659    
3660    
3661    ----------------------------------------------------------------------
3662    Name: Matthias Blume
3663    Date: 2002/02/13 22:40:00 EST
3664    Tag: blume-20020213-fptr-rtti
3665    Description:
3666    
3667    Fixed a bug in ml-nlffigen that was introduced with one of the previous
3668    updates.
3669    
3670    ----------------------------------------------------------------------
3671    Name: Matthias Blume
3672    Date: 2002/02/13 16:41:00 EST
3673    Tag: blume-20020213-cmlpq
3674    Description:
3675    
3676    Added new priority queue export symbols (which have just been added to
3677    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
3678    not compile and the installer would choke.)
3679    
3680    ----------------------------------------------------------------------
3681    Name: Matthias Blume
3682    Date: 2002/02/13 16:15:00 EST
3683    Tag: blume-20020213-various
3684    Description:
3685    
3686    1. More tweaks to ml-nlffigen:
3687    
3688       - better internal datastructures (resulting in slight speedup)
3689       - "-match" option requires exact match
3690       - "localized" gensym counters (untagged structs/unions nested within
3691         other structs/unions or within typedefs get a fresh counter; their
3692         tag will be prefixed by a concatenation of their parents' tags)
3693       - bug fixes (related to calculation of transitive closure of types
3694         to be included in the output)
3695    
3696    2. Minor Basis updates:
3697    
3698       - added implementations for List.collate and Option.app
3699    
3700    ----------------------------------------------------------------------
3701    Name: Matthias Blume
3702    Date: 2002/02/11 15:55:00 EST
3703    Tag: blume-20020211-gensym
3704    Description:
3705    
3706    Added a "-gensym" option to command line of ml-nlffigen.  This can be
3707    used to specify a "stem" -- a string that is inserted in all "gensym'd"
3708    names (ML structure names that correspond to unnamed C structs, unions,
3709    and enums), so that separate runs of ml-nlffigen do not clash.
3710    
3711    ----------------------------------------------------------------------
3712    Name: Matthias Blume
3713    Date: 2002/02/11 12:05:00 EST
3714    Tag: blume-20020211-gensml
3715    Description:
3716    
3717    A quick fix for a problem with GenSML (in the pgraph-util library):
3718    Make generation of toplevel "local" optional.  (Strictly speaking,
3719    signature definitions within "local" are not legal SML.)
3720    
3721    Other than that: updates to INSTALL and cm/TODO.
3722    
3723    ----------------------------------------------------------------------
3724    Name: Matthias Blume
3725    Date: 2002/02/08 15:00:00 EST
3726    Tag: blume-20020208-uniquepid
3727    Description:
3728    
3729    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
3730    
3731    1. The installer (config/install.sh) has gotten smarter:
3732    
3733         - Configuration options are a bit easier to specify now
3734           (in config/targets).
3735         - Bug in recognizing .tar.bz2 files fixed.
3736         - Installer automatically resolves dependencies between
3737           configuration options (e.g., if you ask for eXene, you will
3738           also get cml -- regardless whether you asked for it or not).
3739         - Installer can run in "quieter mode" by setting the environment
3740           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
3741           "completely silent", though.
3742         - Build HashCons library as part of smlnj-lib.
3743    
3744    2. A new scheme for assigning persistent identifiers to compilation
3745       units (and, by extension, to types etc.) has been put into place.
3746       This fixes a long-standing bug where types and even dynamic values
3747       can get internally confused, thereby compromising type safety
3748       (abstraction) and dynamic correctness.  See
3749    
3750         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
3751    
3752       for an example of how things could go wrong until now.
3753    
3754       The downside of the new scheme is that pids are not quite as
3755       persistent as they used to be: CM will generate a fresh pid
3756       for every compilation unit that it thinks it sees for the first
3757       time.  That means that if you compile starting from a clean, fresh
3758       source tree at two different times, you end up with different
3759       binaries.
3760    
3761       Cutoff recompilation, however, has not been compromised because
3762       CM keeps pid information in special caches between runs.
3763    
3764    ----------------------------------------------------------------------
3765    Name: Lal George
3766    Date: 2002/02/07 15:34:13 EST 2002
3767    Tag: <none>
3768    Description:
3769    
3770    Compilers that generate assembly code may produce  global labels
3771    whose value is resolved at link time. The various peephole optimization
3772    modules did not take this in account.
3773    
3774    TODO. The Labels.addrOf function should really return an option
3775    type so that clients are forced to deal with this issue, rather
3776    than an exception being raised.
3777    
3778    ----------------------------------------------------------------------
3779    Name: Lal George
3780    Date: 2002/02/06 13:55:02 EST
3781    Tag: george-20020206-ra-breakup
3782    Description:
3783    
3784    1. A bug fix from Allen.
3785    
3786        A typo causes extra fstp %st(0)'s to be generated at compensation
3787        edges, which might cause stack underflow traps at runtime.  This
3788        occurs in fft where there are extraneous fstps right before the 'into'
3789        trap instruction (in this case they are harmless since none of the
3790        integers overflow.)
3791    
3792    2. Pulled out various utility modules that were embedded in the modules
3793       of the register allocator. I need these modules for other purposes, but
3794       they are not complete enough to put into a library (just yet).
3795    ----------------------------------------------------------------------
3796    Name: Matthias Blume
3797    Date: 2002/01/31 16:05:00 EST
3798    Tag: blume-20020131-sparc-ccalls
3799    Description:
3800    
3801    1. C-calls on Sparc needlessly allocated a huge chunk (96 bytes)
3802       of extra stack space by mistake.  Fixed.
3803    
3804    2. Bug in logic of handling of command-line options in ml-nlffigen fixed.
3805    
3806    ----------------------------------------------------------------------
3807    Name: Allen Leung
3808    Date: 2002/01/30
3809    Tag: leunga-20020130-nowhere-bug-fix
3810    Description:
3811    
3812       MLRISC bug fixes:
3813       1. Fixed a bindings computation bug in the 'nowhere' program generator tool.
3814       2. MachineInt.fromString was negating its value.
3815    
3816    ----------------------------------------------------------------------
3817    Name: Matthias Blume
3818    Date: 2002/01/29
3819    Tag: blume-20020129-INSTALL
3820    Description:
3821    
3822    - Added somewhat detailed installation instructions (file INSTALL).
3823    - Fixed curl-detection bug in config/install.sh.
3824    - It is now possible to select the URL getter using the URLGETTER
3825      environment variable:
3826    
3827          not set / "unknown"      --> automatic detection (script tries wget,
3828                                       curl, and lynx)
3829          "wget" / "curl" / "lynx" --> use the specified program (script "knows"
3830                                       how to properly invoke them)
3831          other                    --> use $URLGETTER directly, it must take
3832                                       precisely two command-line arguments
3833                                       (source URL and destination file name)
3834    
3835    ----------------------------------------------------------------------
3836    Name: Matthias Blume
3837    Date: 2002/01/28
3838    Tag: blume-20020128-sparc-ccalls
3839    Description:
3840    
3841    - Fixed problem with calculation of "used" registers in sparc-c-calls.
3842    - Make use of the allocParam argument in sparc-c-calls.
3843    
3844    ----------------------------------------------------------------------
3845  Name: Matthias Blume  Name: Matthias Blume
3846  Date: 2002/01/28  Date: 2002/01/28
3847  Tag: blume-20020128-allocParam  Tag: blume-20020128-allocParam
# Line 584  Line 4413 
4413    
4414  ----------------------------------------------------------------------  ----------------------------------------------------------------------
4415  Name: Matthias Blume  Name: Matthias Blume
 >>>>>>> 1.169  
4416  Date: 2001/09/18 15:35:00 EDT  Date: 2001/09/18 15:35:00 EDT
4417  Tag: blume-20010918-readme11036  Tag: blume-20010918-readme11036
4418  Description:  Description:

Legend:
Removed from v.1044  
changed lines
  Added in v.