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

Parent Directory | Revision Log | Patch

	revision 1220, Wed May 22 15:04:10 2002 UTC	revision 2066, Thu Oct 5 16:26:30 2006 UTC
#	Line 7	Line 7
7
8	The form of an entry should be:	The form of an entry should be:
9
10	Name:
11		Name: Your Name (web page or e-mail)
12		Date: yyyy/mm/dd hh:mm:ss
13		Tag: cvs-tag-for-this-change
14		Description:
15
16		----------------------------------------------------------------------
17		Name: Matthias Blume (http://www.tti-c.org/blume/)
18		Date: 2006/10/05 11:30:00 CDT
19		Tag: blume-20061005-merge-amd64
20		Description:
21
22		Merged code for AMD64 backend (Mike Rainey's work).
23		Everything is hooked up but untested.
24
25		----------------------------------------------------------------------
26		Name: Matthias Blume (http://www.tti-c.org/blume/)
27		Date: 2006/05/17 14:00:00 CST
28		Tag: Release_110_59
29		Description:
30
31		New working version (110.58).  NEW BOOTFILES!
32
33		I am freezing 110.59.  Changes other than the version-number increase:
34
35		eXene:
36
37		* committed changes to eXene from Alley Stoughton:
38		"fixed bugs in X authorization and resource handling, as well
39		as in the pile and viewport widgets"
40
41		Runtime:
42		* fixed linking problem with NetBSD 3.x.
43
44		Lexgen:
45		* lexgen tool handles non-ascii characters in 7-bit mode the same
46		way that ml-lex does
47
48		* lexgen propagates exceptions the same way that ml-lex does
49
50		CML:
51		* Fixed a bug in the SyncVar polling functions (iGetPoll,
52		mTakePoll, and mGetPoll) that could lead to livelock.
53
54		----------------------------------------------------------------------
55		Name: Matthias Blume (http://www.tti-c.org/blume/)
56		Date: 2006/05/12 15:20:00 CST
57		Tag: blume-20060515-makedepend
58		Description:
59
60		Implemented ml-makedepend (i.e., CM.sources) in a better (more robust)
61		way.  This should hopefully fix the ml-makedepend problem permanently.
62
63		----------------------------------------------------------------------
64		Name: Matthias Blume (http://www.tti-c.org/blume/)
65		Date: 2006/05/12 15:15:00 CST
66		Tag: blume-20060512-makedepend
67		Description:
68
69		Fixed long-standing bug with ml-makedepend where it would output a
70		spurious dependency to a non-existing file.  (This is a simple fix.
71		It might need further looking into.)
72
73		----------------------------------------------------------------------
74		Name: Matthias Blume (http://www.tti-c.org/blume/)
75		Date: 2006/04/20 10:30:00 CDT
76		Tag: blume-20060420-heap2exec
77		Description:
78
79		Committed patches received from Johannes 5 Joemann (joemann@befree.free.de)
80		that enable heap2exec under Linux and FreeBSD.
81
82		----------------------------------------------------------------------
83		Name: John Reppy
84		Date: 2006/04/14
85		Tag:
86		Description:
87
88		MLRISC changes:
89		1) renamed GAS_PSEUDO_OPS to AS_PSEUDO_OPS and put it in its own file.
90		2) added support for NOTB and XORB operators in pseudo-op expressions
91		3) added DarwinPseudoOp functor that supports Darwin's assembler syntax.
92		4) added support for 64-bit integer literals
93
94		----------------------------------------------------------------------
95		Name: Matthias Blume (http://www.tti-c.org/blume/)
96	Date: yyyy/mm/dd	Date: yyyy/mm/dd
97	Tag: <post-commit CVS tag>	Tag: Release_110_58
98		Description:
99
100		New working version (110.58).  NEW BOOTFILES!
101
102		----------------------------------------------------------------------
103		Name: Matthias Blume (http://www.tti-c.org/blume/)
104		Date: 2006/03/01 22:15:00 CST
105		Tag: blume-20060301-lexgen
106		Description:
107
108		Incorporated several bugfixes to lexgen.  Compiler now compiles to
109		fixpoint when using lexgen instead of ml-lex.
110
111		----------------------------------------------------------------------
112		Name: Matthias Blume (http://www.tti-c.org/blume/)
113		Date: 2006/02/28 22:45:00 CST
114		Tag: blume-20060228-lexgen
115		Description:
116
117		Removed ml-flex and added lexgen instead, using Aaron Turon's newly
118		provided tarball.  The generated lexers still have problems.
119
120		----------------------------------------------------------------------
121		Name: Matthias Blume (http://www.tti-c.org/blume/)
122		Date: 2006/02/26 23:30:00 CST
123		Tag: blume-20060226-ml-flex
124		Description:
125
126		Removed ml-flex' dependency on regexp-lib.cm.  Turned installation
127		of ml-flex on by default.
128
129		----------------------------------------------------------------------
130		Name: Matthias Blume (http://www.tti-c.org/blume/)
131		Date: 2006/02/24 16:30:00 CST
132		Tag: blume-20060224-ml-flex
133		Description:
134
135		Added ml-flex sources.  Partially integrated, but should not be
136		turned on yet!  (Read: leave it commented-out in config/targets!)
137
138		----------------------------------------------------------------------
139		Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
140		Date: 2006/02/23
141		Tag: jhr-20060223-x86-64
142		Description:
143		Changes to support building on x86-64 systems (using the 32-bit mode).
144		Also cleaned up signal handling on Linux.  Support for pre-2.2 Linux
145		kernels dropped.
146
147		----------------------------------------------------------------------
148		Name: Allen Leung (leunga (at) reservoir (dot) com)
149		Date: 2006/02/22 15:07:00 EST
150		Tag: leunga-20060222-mdl-gen
151		Description:
152
153		CM has changed.  Updated the script for rebuilding the MLRISC generated
154		files.
155
156		----------------------------------------------------------------------
157		Name: Matthias Blume (http://www.tti-c.org/blume/)
158		Date: 2006/02/14 17:30:00 CST
159		Tag: blume-20060214-intel-mac
160		Description:
161
162		Hooked code for Darwin-specific Intel ABI into main compiler.
163		(This is a temporary solution which relies on the fact that the
164		compiler itself does not use NLFFI.  Eventually we need to divorce
165		intel mac from generic x86 unix code and make separate sets of binaries.)
166
167		----------------------------------------------------------------------
168		Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
169		Date: 2006/02/13
170		Tag: jhr-20060213-intel-mac
171		Description:
172		Changed MLRISC x86 CCalls for partial support of MacOS X ABI.
173
174		----------------------------------------------------------------------
175		Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
176		Date: 2006/02/06
177		Tag: jhr-20060206-intel-mac
178		Description:
179		Changes to support MacOS X on Intel hardware.  The C-calls support in
180		MLRISC must also be updated to support the MacOS X ABI.
181
182		----------------------------------------------------------------------
183		Name: Matthias Blume (http://www.tti-c.org/blume/)
184		Date: 2005/12/16 00:15:00 CST
185		Tag: blume-20051216-cm
186		Description:
187
188		Improved error reporting and handling in CM.
189
190		----------------------------------------------------------------------
191		Name: Matthias Blume (http://www.tti-c.org/blume/)
192		Date: 2005/11/21 14:00:00 CST
193		Tag: blume-20051121-html-lib
194		Description:
195
196		Enabling $/html-lib.cm in config/preloads did not work.  This is fixed
197		now.  (Since the anchor mapping for html-lib.cm is not yet in effect
198		at the time when config/preloads is processed, the library has to
199		be referred to by another name.  In 110.57 this would be
200		$SMLNJ-LIB/HTML/html-lib.cm.  I arranged for $smlnj/smlnj-lib/html-lib.cm
201		to be valid as well.)
202
203		Thanks to Todd Wilson (Fresno) for alerting me to this issue.
204
205		----------------------------------------------------------------------
206		Name: Matthias Blume (http://www.tti-c.org/blume/)
207		Date: 2005/11/19 13:45:00 CST
208		Tag: Release_110_57
209		Description:
210
211		Moved release tag for 110.57.
212
213		----------------------------------------------------------------------
214		Name: Matthias Blume (http://www.tti-c.org/blume/)
215		Date: 2005/11/19 23:35:00 CST
216		Tag: blume-20051119-fish
217		Description:
218
219		Fixed a problem in config/install.sh where it tries to "fish" the
220		name of the CM metadata directory from the wrong place (because
221		the physical location of basis.cm has changed).  Also, corrected the
222		path anchor for $/html-lib.cm.  (Thanks to M. Fluet for pointing out
223		these problems.)
224
225		----------------------------------------------------------------------
226		Name: Matthias Blume (http://www.tti-c.org/blume/)
227		Date: 2005/11/16 14:30:00 CST
228		Tag: blume-orig-110_57
229		Description:
230
231		New working version (110.57).  NEW BOOTFILES!
232
233		UPDATE: The release tag has been moved to include a fix for a brown-paper-
234		bag glitch (see above).
235
236		----------------------------------------------------------------------
237		Name: Matthias Blume (http://www.tti-c.org/blume/)
238		Date: 2005/11/16 10:25:00 CST
239		Tag: blume-22051116-btrace
240		Description:
241
242		Fixed problem with bogus exception message when using back-trace
243		facility.
244
245		----------------------------------------------------------------------
246		Name: Matthias Blume (http://www.tti-c.org/blume/)
247		Date: 2005/11/15 22:35:00 CST
248		Tag: blume-22051115-array2copy
249		Description:
250
251		Added simple implementation of Array2.copy.  (Warning: mostly untested.)
252
253		----------------------------------------------------------------------
254		Name: Matthias Blume (http://www.tti-c.org/blume/)
255		Date: 2005/11/15 21:35:00 CST
256		Tag: blume-22051115-versiontool
257		Description:
258
259		Reversed change to src/system/smlnj/internals/versiontool.cm.  This
260		file gets loaded as a tool -- by the equivalent of CM.make during the
261		run of CMB.make.  Thus, CMB's path configuration is meaningless for
262		it.  Instead, it has the status of "user code", so it should use
263		$/basis.cm to refer to the Basis library.  (At least that's true for
264		the purpose of bootstrapping the previous change.  In the future it
265		might make sense to have versiontool.cm refer to
266		$smlnj/basis/basis.cm, i.e., the version of the Basis that the
267		compiler itself uses.)
268
269		Also patched src/system/testml to have it activate those extra anchor
270		bindings in config/extrapathconfig.
271
272		----------------------------------------------------------------------
273		Name: Matthias Blume (http://www.tti-c.org/blume/)
274		Date: 2005/11/15 16:35:00 CST
275		Tag: blume-22051115-libraries
276		Description:
277
278		This change affects the way the following libraries are tied into
279		the system:
280
281		$/basis.cm
282		$/smlnj-lib.cm
283		$/pp-lib.cm
284		$/controls-lib.cm
285		$/html-lib.cm
286		$/ml-yacc-lib.cm
287
288		These libraries are now internally (as seen from the source code
289		of the implementation itself) known by the following names:
290
291		$smlnj/basis/basis.cm
292		$smlnj/smlnj-lib/smlnj-lib.cm
293		$smlnj/smlnj-lib/pp-lib.cm
294		$smlnj/smlnj-lib/controls-lib.cm
295		$smlnj/smlnj-lib/html-lib.cm
296		$smlnj/ml-yacc/ml-yacc-lib.cm
297
298		This makes it possible to work with code that requires different
299		versions of these libraries, and which refers to these libraries using
300		their "default" names (i.e., the first set of names above).  In other
301		words, one can un-define or re-define those default names without
302		compromising the proper functioning of the compiler itself.
303
304		A similar procedure had already been performed for several of the
305		MLRISC libraries that are linked into the compiler.  I did some
306		cleanup on this code.
307
308		A new file in the config directory (named extrapathconfig) is
309		responsible for setting up path anchors that the compiler itself does
310		not need, but that are typically required by user code.
311
312		----------------------------------------------------------------------
313		Name: Matthias Blume
314		Date: 2005/11/07 17:05:00 CST
315		Tag: blume-20051107-slice
316		Description:
317
318		Fixed erroneous out-of-bounds test in the "update" function of
319		various *ArraySlice modules.  (Thanks to Vesa A. Norrman for pointing
320		out the problem.)
321
322		Pushed some Basis changes through ML-Lex, CML, and eXene.
323
324		----------------------------------------------------------------------
325		Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
326		Date: 2005/11/07
327		Description:
328		Fixed a Basis incompatibility: The depreciated function Substring.all
329		was removed (use Substring.full instead).
330
331		----------------------------------------------------------------------
332		Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
333		Date: 2005/11/05
334		Description:
335		Tweaked interval set API in SML/NJ library; see the CHANGES file for details.
336
337		----------------------------------------------------------------------
338		Name: Matthias Blume (http://www.tti-c.org/blume/)
339		Date: 2005/11/02 16:25:00 CST
340		Tag: blume-20051102-hex
341		Description:
342
343		* Runtime system bootstrap code now accepts hex digits in BOOTLIST
344		in either upper case or lower case format.
345		* Pushed changes to names of Pack<N>{Big,Little} structures through
346		CML and eXene.
347
348		----------------------------------------------------------------------
349		Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
350		Date: 2005/11/02
351		Description:
352		Fixed a Basis incompatibility: Pack<N>{Big,Little} structures should
353		be called PackWord<N>{Big,Little}.
354
355		----------------------------------------------------------------------
356		Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
357		Date: 2005/10/28
358		Description:
359		Fixed a minor Basis incompatibility: hex digits should be upper case.
360
361		----------------------------------------------------------------------
362		Name: Matthias Blume (blume (at) tti - c (dot) org)
363		Date: 2005/10/25 16:00:00 CDT
364		Tag: Release_110_56
365		Description:
366
367		New working version (110.56).  NEW BOOTFILES!
368
369		----------------------------------------------------------------------
370		Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
371		Date: 2005/10/25
372		Description:
373		Added interval sets to utility library (signatures INTERVAL_DOMAIN
374		and INTERVAL_SET, and functor IntervalSetFn).
375
376		----------------------------------------------------------------------
377		Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
378		Date: 2005/10/14
379		Description:
380		Add Zhong Shao's fix for datatype equality functions.
381
382		----------------------------------------------------------------------
383		Name: Allen Leung (leunga (at) reservoir (dot) com)
384		Date: 2005/10/14 12:30:00 EST
385		Tag: leunga-20051014-x86-spill
386		Description:
387
388		Bug fix a bug found by Carl Hauser.
389		There was a typo in the reload code for FCMP in x86SpillInstr.sml.
390
391		----------------------------------------------------------------------
392		Name: Allen Leung (leunga (at) reservoir (dot) com)
393		Date: 2005/10/14 11:15:00 EST
394		Tag: leunga-20051014-x86-asm
395		Description:
396
397		Removed some debugging code in file x86Asm.sml.
398		The function emit_operand was printing out debugging output.
399
400		----------------------------------------------------------------------
401		Name: John Reppy
402		Date: 2005/07/27
403		Tag:
404		Description:
405
406		Fixed ml-lex to recognize "\r" as representing carriage return.
407
408		----------------------------------------------------------------------
409		Name: John Reppy
410		Date: 2005/07/27
411		Tag:
412		Description:
413
414		Fixed ml-yacc to work on files with non-native end-of-line
415		encodings (e.g., Windows text file on a Unix system).
416
417		----------------------------------------------------------------------
418		Name: John Reppy
419		Date: 2005/07/20
420		Tag:
421		Description:
422
423		Added changes from Dominic Evans (oldmanuk (at) gmail (dot) com)
424		to support HPUX 11.
425
426		----------------------------------------------------------------------
427		Name: John Reppy
428		Date: 2005/07/06
429		Tag:
430		Description:
431
432		Changes to the SML/NJ library.  See smlnj-lib/CHANGES for details.
433
434		----------------------------------------------------------------------
435		Name: Matthias Blume (blume (at) tti - c (dot) org)
436		Date: 2005/07/06 00:45:00 CDT
437		Tag: blume-20050706-slice-copy
438		Description:
439
440		Fixed reversed logic for deciding whether to "copy up" or "copy down"
441		in *-array-slice.sml.
442
443		----------------------------------------------------------------------
444		Name: Allen Leung (leunga (at) reservoir (dot) com)
445		Date: 2005/05/31 17:00:00 EST
446		Tag: leunga-20050531-cygwin-fault-2
447		Description:
448
449		A typo in the cygwin code fixed.
450
451		----------------------------------------------------------------------
452		Name: Allen Leung (leunga (at) reservoir (dot) com)
453		Date: 2005/05/31 16:47:00 EST
454		Tag: leunga-20050531-cygwin-fault
455		Description:
456
457		Updated Cygwin's fault/signal handling to match the Windows version.
458		Updated the export list.
459
460		----------------------------------------------------------------------
461		Name: Matthias Blume (blume (at) tti - c (dot) org)
462		Date: 2005/05/18 13:35:00 CDT
463		Tag: Release_110_54
464		Description:
465
466		New working version (110.54).  NEW BOOTFILES!
467
468		----------------------------------------------------------------------
469		Name: Matthias Blume (blume (at) tti - c (dot) org)
470		Date: 2005/05/18 11:58:00 CDT
471		Tag: blume-20050518-installer
472		Description:
473
474		Added support scripts for Mac OS X PackageMaker and modified
475		config/install.sh so that it supports re-dumping a heap image
476		after customization.
477
478		----------------------------------------------------------------------
479		Name: Matthias Blume (blume (at) tti - c (dot) org)
480		Date: 2005/05/18 10:55:00 CDT
481		Tag: blume-20050518-realdiv-noovld
482		Description:
483
484		Un-overloaded / to work around bug in overloading resolution code.
485
486		----------------------------------------------------------------------
487		Name: Matthias Blume (blume (at) tti - c (dot) org)
488		Date: 2005/05/16 23:50:00 CDT
489		Tag: blume-20050516-redump-heap
490		Description:
491
492		Added mechanism for re-creating a heap file for the interactive system
493		after configuration variables have been changed.
494
495		CM.redump_heap : string -> unit
496
497		This is much like SMLofNJ.exportML, but starting from the resulting
498		heap does not return to the caller of CM.redump_heap but
499		restarts the interactive system from scratch.  The original call of
500		CM.redump_heap does not return but ends the interactive session.
501		Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
502
503		Internally, redump_heap winds the dynamic execution context back to
504		the point where the original heap image was created and re-executes
505		the heap image generation code in the boot code.
506
507		----------------------------------------------------------------------
508		Name: Matthias Blume (blume (at) tti - c (dot) org)
509		Date: 2005/05/09 21:55:00 CDT
510		Tag: blume-20050509-word64
511		Description:
512
513		Added a hack to the existing hack known as Word64 to make fromString
514		behave correctly.  I am still not sure whether Word64.scan will work
515		as specified with respect to the interaction of radix and prefix.
516
517		----------------------------------------------------------------------
518		Name: Allen Leung (leunga (at) reservoir (dot) com)
519		Date: 2005/05/04 11:50:00 EST
520		Tag: leunga-20050504-checkgc
521		Description:
522
523		Added a gc protocol checking phase.  This phase is enabled with
524		the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
525
526		----------------------------------------------------------------------
527		Name: Matthias Blume (blume (at) tti - c (dot) org)
528		Date: 2005/05/04 10:45:00 CDT
529		Tag: blume-20050504-intinf
530		Description:
531
532		Fixed a bug in the implementation of div and mod for IntInf.
533		Thanks to Neophytos Michael for reporting the problem.
534
535		----------------------------------------------------------------------
536		Name: Matthias Blume (blume (at) tti - c (dot) org)
537		Date: 2005/05/04 10:35:00 CDT
538		Tag: blume-20050504-join
539		Description:
540
541		Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
542
543		----------------------------------------------------------------------
544		Name: Matthias Blume (blume (at) tti - c (dot) org)
545		Date: 2005/02/28 23:40:00 CST
546		Tag: blume-20050228-mVar
547		Description:
548
549		Fixed serious bug (brown paper bag variety) in new implementation of
550		structure Atom in CML.  (I had accidentally used a mailbox instead of
551		an mvar, leaving the door open for races.)
552
553		----------------------------------------------------------------------
554		Name: Matthias Blume (blume (at) tti - c (dot) org)
555		Date: 2005/02/25 15:00:00 CST
556		Tag: Release_110_53
557	Description:	Description:
558
559		New working version (110.53).  NEW BOOTFILES!
560
561		----------------------------------------------------------------------
562		Name: Matthias Blume (blume (at) tti - c (dot) org)
563		Date: 2005/02/25 14:50:00 CST
564		Tag: blume-20050225-susp
565		Description:
566
567		Brought back SMLofNJ.Susp.  The underlying suspension type is the one
568		implemented in Core -- which means that it is the same as the one used
569		by the lazy extension.
570
571		----------------------------------------------------------------------
572		Name: Matthias Blume (blume (at) tti - c (dot) org)
573		Date: 2005/02/24 16:50:00 CST
574		Tag: blume-20050224-cml-atom
575		Description:
576
577		Simpler and at the same time more general implementation of structure
578		Atom in CML.
579
580		----------------------------------------------------------------------
581		Name: Matthias Blume (blume (at) tti - c (dot) org)
582		Date: 2005/02/15 17:35:00 CST
583		Tag: blume-20050215-tools
584		Description:
585
586		Created new "tools" directory under "src" and moved "TraceDebugProf"
587		there.
588
589		----------------------------------------------------------------------
590		Name: Matthias Blume (blume (at) tti - c (dot) org)
591		Date: 2005/02/10 17:55:00 CST
592		Tag: blume-20050210-longlong
593		Description:
594
595		Implemented "long long" arguments and results for NLFFI.  (Only the
596		PPC/MacOS implementation is complete, the other backends still need to
597		be updated.)
598
599		----------------------------------------------------------------------
600		Name: Matthias Blume (blume (at) tti - c (dot) org)
601		Date: 2005/01/24 17:40:00 CST
602		Tag: blume-20050124-mlyacc
603		Description:
604
605		Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
606		problem with certain "as" patterns which previously got rendered
607		using incorrect syntax.
608
609		----------------------------------------------------------------------
610		Name: Matthias Blume (blume (at) tti - c (dot) org)
611		Date: 2005/01/18 12:00:00 CST
612		Tag: blume-20050118-profile
613		Description:
614
615		Made time profiling code (interrupt handler) in runtime system aware
616		of new array representation.
617
618		----------------------------------------------------------------------
619		Name: Matthias Blume (blume (at) tti - c (dot) org)
620		Date: 2005/01/14 18:00:00 CST
621		Tag: blume-20050114-heap2exec
622		Description:
623
624		Implemented new (but still experimental) heap2exec facility.  This is
625		tested under Mac OS X and should work under Linux (will test shortly).
626		It will probably also work on the Sparc (will test some time later).
627
628		- removed old "HACKED_STANDALONE" hack from runtime
629
630		To be able to test this, uncomment the request for "heap2asm" in
631		config/targets prior to installation.  (Notice that this is different
632		from "heap2exec" mentioned below.  Not a typo.)
633
634		To perform an actual test, run the command
635
636		$ bin/heap2exec heapfile execfile
637
638		(You can put heap2exec on your shell's path.)
639
640		For example, run
641
642		$ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
643
644		This will create a standalone executable called "mly" which you can
645		then invoke directly as a command.
646
647		----------------------------------------------------------------------
648		Name: Matthias Blume (blume (at) tti - c (dot) org)
649		Date: 2005/01/07 11:44:00 CST
650		Tag: blume-20050107-mlstring
651		Description:
652
653		fixed off-by-one error in ML_STRING macro (globals.c)
654
655		----------------------------------------------------------------------
656		Name: Matthias Blume (blume (at) tti - c (dot) org)
657		Date: 2004/12/23 18:00:00 CST
658		Tag: blume-20041223-santa
659		Description:
660
661		Made ml-build script "smarter" (but only very little).
662
663		----------------------------------------------------------------------
664		Name: Matthias Blume (blume (at) tti - c (dot) org)
665		Date: 2004/12/21 15:05:00 CST
666		Tag: blume-20041221-longlong
667		Description:
668
669		* Implemented access to signed and unsigned long long data in NLFFI.
670		(The parameter-passing part of the picture has not complete. But
671		data structure access seems to work.)
672
673		* Fixed CM's incorrect assumption that the PPC is little-endian.
674		(On the Mac, it is big-endian.  And that's currently our only
675		PPC platform.)
676
677		----------------------------------------------------------------------
678		Name: Matthias Blume (blume (at) tti - c (dot) org)
679		Date: 2004/12/21 12:50:00 CST
680		Tag: blume-20041221-memory
681		Description:
682
683		Some cleanup in the $c/memory.cm library: separated some concerns by
684		moving allocation code and memory access code each into their own
685		files.
686
687		----------------------------------------------------------------------
688		Name: Allen Leung (leunga (at) reservoir (dot) com)
689		Date: 2004/12/17 16:12:00 EST
690		Tag: leunga-20041217-cygwin-smlnj-home
691		Description:
692
693		The Unix I/O library of SML/NJ on cygwin does not understand
694		Windows style pathname, so problems arise when SMLNJ_HOME is set
695		to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
696		to a POSIX pathname on cygwin.
697
698		----------------------------------------------------------------------
699		Name: Matthias Blume (blume (at) tti - c (dot) org)
700		Date: 2004/12/16 13:35:00 CST
701		Tag: Release_110_52
702		Description:
703
704		Last-minute changes incorporated into 110.52.  Release tag moved.
705
706		The changes:
707		- HashString.hashString' -> HashString.hashSubstring
708		- bug fix in UnivariateStats
709
710		----------------------------------------------------------------------
711		Name: Matthias Blume (blume (at) tti - c (dot) org)
712		Date: 2004/12/15 23:40:00 CST
713		Tag: blume-20041215-hashSubstring
714		Description:
715
716		- HashString.hashString' -> HashString.hashSubstring
717		- corresponding changes in atom.sml
718		- "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
719		added some comments
720
721		----------------------------------------------------------------------
722		Name: Matthias Blume (blume (at) tti - c (dot) org)
723		Date: 2004/12/15 15:30:00 CST
724		Tag: (Release_110_52)
725		Description:
726
727		New working version (110.52).  NEW BOOTFILES!
728
729		----------------------------------------------------------------------
730		Name: Matthias Blume (blume (at) tti - c (dot) org)
731		Date: 2004/12/15 12:45:00 CST
732		Tag: blume-20041215-spaces
733		Description:
734
735		More on the space problem (this time for Win32).
736
737		----------------------------------------------------------------------
738		Name: Matthias Blume (blume (at) tti - c (dot) org)
739		Date: 2004/12/14 17:30:00 CST
740		Tag: blume-20041214-spaces
741		Description:
742
743		Hacked some of the scripts (in particular: the installer) to cope with
744		spaces in filenames a bit better.  But beware: the current "solution"
745		is likely still full of bugs and inherently incomplete.  (We need to
746		do away with those shell scripts for a comprehensive solution.)
747
748		----------------------------------------------------------------------
749		Name: Matthias Blume (blume (at) tti - c (dot) org)
750		Date: 2004/12/13 14:45:00 CST
751		Tag: blume-20041213-ml-makedepend
752		Description:
753
754		Fixed bug in code for ml-makedepend.
755
756		----------------------------------------------------------------------
757		Name: Matthias Blume (blume (at) tti - c (dot) org)
758		Date: 2004/12/09 16:30:00 CST
759		Tag: blume-20041209-statistics
760		Description:
761
762		Added two simple but potentially useful statistics modules to SML/NJ Library.
763		(See CHANGES file there.)
764
765		----------------------------------------------------------------------
766		Name: Matthias Blume (blume (at) tti - c (dot) org)
767		Date: 2004/12/01 16:50:00 CST
768		Tag: blume-20041201-atom
769		Description:
770
771		smlnj-lib:
772
773		Added function HashString.hashString' for substrings.
774		Hand-inlined CharVector.foldl into HashString (for speed).
775		Modified implementation of structure Atom to avoid extracting
776		strings from substrings unless necessary.
777		(Also see CHANGES file for smlnj-lib.)
778
779		----------------------------------------------------------------------
780		Name: Matthias Blume (blume (at) tti - c (dot) org)
781		Date: 2004/11/24 22:15:00 CST
782		Tag: blume-20041124-cml
783		Description:
784
785		Made sure CML compiles when Position = Int64.
786
787		----------------------------------------------------------------------
788		Name: Matthias Blume (blume (at) tti - c (dot) org)
789		Date: 2004/11/24 14:30:00 CST
790		Tag: blume-20041124-position
791		Description:
792
793		The compiler can now be compiled in a mode that makes structure Position
794		equal to Int64.  The default, however, is unchanged (Position = Int31)
795		for the time being.
796
797		To enable 64-bit positions, use the following procedure:
798
799		1. Start sml
800		2. Autoload $smlnj/cmb.cm (if not already autoloaded)
801		3. Type
802		#set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
803		4. Run CMB.make() as usual.
804
805		This is barely tested.  The only test so far was a little SML program
806		counting the number of characters in an 8-gigabyte file by
807		reading it character-by-character.  That test was successful.
808
809		In support of 64-bit positions, a number of new functions have been
810		added to the runtime system.
811
812		----------------------------------------------------------------------
813		Name: Matthias Blume (blume (at) tti - c (dot) org)
814		Date: 2004/11/23 14:45:00 CST
815		Tag: blume-20041123-useFile
816		Description:
817
818		Fixed a problem with unhelpful error messages related to problems with
819		.cm- or .sml files that appear as part of the sml command line.
820
821		----------------------------------------------------------------------
822		Name: Matthias Blume (blume (at) tti - c (dot) org)
823		Date: 2004/11/18 15:40:00 CST
824		Tag: Release_110_51
825		Description:
826
827		New working version (110.51).  NEW BOOTFILES!
828
829		----------------------------------------------------------------------
830		Name: Matthias Blume (blume (at) tti - c (dot) org)
831		Date: 2004/11/18 15:35:00 CST
832		Tag: <none>
833		Description:
834
835		Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
836		Joemann).
837
838		----------------------------------------------------------------------
839		Name: Allen Leung (leunga (at) reservoir (dot) com)
840		Date: 2004/11/17 16:05:21 EST 2004
841		Tag: leunga-20041117-mlrisc-live-kill
842		Description:
843
844		Added support for MLTree constructs LIVE and KILL
845		to all the architectures.
846
847		----------------------------------------------------------------------
848		Name: Matthias Blume (blume (at) tti - c (dot) org)
849		Date: 2004/11/13 00:20:00 CST
850		Tag: blume-20041113-versiontool
851		Description:
852
853		- Stripped down the versiontool: It now only handles the version number.
854		The date string is generated at bootstrap time (during makeml).
855
856		- In a previous commit, fixed a minor issue with how polyequal is being
857		translated.  In particular, the code now "looks through" abstractions.
858		This results in slightly fewer polyEqual warnings and hopefully slightly
859		more efficient code.  Important examples for where this matters are
860		the new int64 and word64 types.
861
862		----------------------------------------------------------------------
863		Name: Matthias Blume (blume (at) tti - c (dot) org)
864		Date: 2004/11/12 00:30:00 CST
865		Tag: blume-20041112-int64
866		Description:
867
868		Structure Int64 fully hooked in.  (The implementation is not very
869		efficient, though.)
870
871		----------------------------------------------------------------------
872		Name: Matthias Blume (blume (at) tti - c (dot) org)
873		Date: 2004/11/11 17:30:00 CST
874		Tag: blume-20041111-more64
875		Description:
876
877		All the pieces of Word64 are now there, with the exception of the
878		conversions from and to LargeWord. (Eventually these need to be identities,
879		but for the time being they don't even make sense because LargeWord is
880		32-bit wide.)
881
882		Also started to add similar support for Int64, but major pieces of that
883		are still missing.
884
885		----------------------------------------------------------------------
886		Name: Matthias Blume (blume (at) tti - c (dot) org)
887		Date: 2004/11/11 00:15:00 CST
888		Tag: blume-20041111-word64
889		Description:
890
891		Structure Word64 is now (almost) complete, word literals and patterns
892		seem to work.  There are a few odd pieces missing.  In particular,
893		I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
894		at the moment.
895
896		Making Word64 official would mean that LargeWord becomes Word64.  But
897		this requires extreme care because most word-word conversions have to
898		go through LargeWord, so making a mistake means loss of efficiency or
899		worse.  Eventually there will be a solution similar to (but actually
900		simpler than) what I did with IntInf.
901
902		----------------------------------------------------------------------
903		Name: Matthias Blume (blume (at) tti - c (dot) org)
904		Date: 2004/11/10 18:12:00 CST
905		Tag: blume-20041110-64bit
906		Description:
907
908		More 64-bit hacking (but still not even half-way there yet).
909		Also, some assorted improvements to the handling of 8-bit words.
910
911		----------------------------------------------------------------------
912		Name: Matthias Blume (blume (at) tti - c (dot) org)
913		Date: 2004/11/09 17:50:00 CST
914		Tag: <none>
915		Description:
916
917		Started adding some infrastructure for supporting 64-bit int- and
918		word-types.  (Still in its very early stages.)
919
920		----------------------------------------------------------------------
921		Name: Matthias Blume (blume (at) tti - c (dot) org)
922		Date: 2004/10/28 10:45:00 CDT
923		Tag: Release_110_50
924		Description:
925
926		New working version (110.50).  NEW BOOTFILES!
927
928		=====================
929
930		Also:
931
932		- Changend config/srcarchiveurl from a file just
933		containing the URL string into a file containing
934		shell script code.  The code has access to the $VERSION variable.
935		- Made corresponding changes to config/install.sh and config/unpack.
936		- Default contents of config/srcarchiveurl uses $VERSION and
937		normally does not have to be edited to reflect a version change.
938
939		(As a result, a version change can be done by just editing
940		config/version, the rest is now automatic.)
941
942		----------------------------------------------------------------------
943		Name: Matthias Blume (blume (at) tti - c (dot) org)
944		Date: 2004/10/27 17:50:00 CDT
945		Tag: blume-20041027-btrace-msg
946		Description:
947
948		BackTrace.monitor now also reports the source of the exception that
949		triggered the trace.
950
951		----------------------------------------------------------------------
952		Name: Matthias Blume (blume (at) tti - c (dot) org)
953		Date: 2004/10/27 17:20:00 CDT
954		Tag: blume-20041027-x86-c-calls
955		Description:
956
957		This is the HISTORY entry for two earlier commits, both concerning
958		the x86 c-calls code in MLRISC:
959
960		- added a missing LOAD in the code that deals with struct arguments
961		- made sure the caller does not add the wrong number of bytes to the
962		stack pointer after a call of a function returning a struct
963		(the callee already pops the implicit argument which points to
964		the space reserved for the result)
965
966		----------------------------------------------------------------------
967		Name: Allen Leung (leunga (at) reservoir (dot) com)
968		Date: 2004/10/24 14:00:00 EST
969		Tag: leunga-20041024-x86-gas-fucomip
970		Description:
971
972		John discovered a bug in the syntax of fucomip.
973		The opcodes FU?COMIP? have been changed to
974
975		fu?comip? %st(i), %st
976
977		----------------------------------------------------------------------
978		Name: Matthias Blume (blume (at) tti - c (dot) org)
979		Date: 2004/10/20 15:06:00 CDT
980		Tag: blume-20041020-standalone-backtrace
981		Description:
982
983		Added a mechanism for getting back-trace information from standalone
984		programs.  Here is how it works:
985
986		1. The part of the program from which you want to get backtrace
987		information (usually the whole program) should be wrapped with
988		BackTrace.monitor.  This is a (unit->'a)->'a function, and your
989		main program could be modified from something like
990
991		fun main (pgm, args) = ...
992
993		to
994
995		fun main (pgm, args) = BackTrace.monitor (fn () => ...)
996
997		2. To be able to access BackTrace.monitor, you have to add
998		library $smlnj-tdp/plugins.cm to the .cm file that contains your
999		main function.
1000
1001		3. Remove all compiled code (i.e., all the .cm/ subdirectories that
1002		CM might have created in the past for your project).
1003
1004		4. Build the system using this command line:
1005
1006		ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
1007		myprog.cm MyProg.main myprog
1008
1009		instead of the usual
1010
1011		ml-build myprog.cm MyProg.main myprog
1012
1013		I changed a library name:
1014
1015		$/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
1016
1017		New libraries:
1018
1019		$smlnj-tdp/back-trace.cm
1020		-- when loaded causes the back-trace plugin to be installed
1021		$smlnj-tdp/coverage.cm
1022		-- when loaded causes the coverage plugin to be installed
1023
1024		----------------------------------------------------------------------
1025		Name: Matthias Blume (blume (at) tti - c (dot) org)
1026		Date: 2004/10/18 16:45:00 CDT
1027		Tag: blume-20041018-groupowner
1028		Description:
1029
1030		Added an "obsolete" warning for the "group owner" syntax to CM's parser.
1031
1032		Eliminated group owner specs from .cm files throughout the source tree.
1033
1034		----------------------------------------------------------------------
1035		Name: Matthias Blume (blume (at) tti - c (dot) org)
1036		Date: 2004/10/15 15:45:00 CDT
1037		Tag: blume-20041015-coverage
1038		Description:
1039
1040		* Test coverage tool added!
1041		* Further reorganization of tracing-, debugging-, and profiling support:
1042
1043		- moved original BTImp -- now called BackTrace -- into a separate
1044		library called $/trace-debug-profile.cm
1045		- eliminated all mentions of BTrace from SMLofNJ.Internals
1046		- only the instrumentation mechanism is now left in the compiler proper
1047		- BackTrace module is a plugin which is NOT plugged in by default
1048		- Coverage module is another such plugin
1049
1050		To get the benefits of any of these plugin modules, the code in
1051		question must be compiled with tdp instrumentation turned on.  This
1052		can be done by setting SMLofNJ.Internals.TDP.mode to true.
1053		(The ref cell is also controlled via the -Ctdp.instrument=... switch.)
1054
1055		Plugins are selected at link time.  (Pre-compiled instrumented code
1056		can be re-loaded with different plugins in effect.)  When an
1057		instrumented module is linked, whatever plugins are at that time
1058		enabled will come into effect for that module.
1059
1060		To enable the back-trace plugin, load library $/trace-debug-profile.cm
1061		and invoke BackTrace.install() (e.g., from the interactive prompt).
1062		To enable the coverage plugin, load the same library and invoke
1063		Coverage.install().
1064
1065		Back-traces are generated automatically on uncaught exceptions and
1066		when the code in question explicitly invokes BackTrace.trigger().
1067
1068		Coverage (and execution frequency-) information must be queried
1069		explicitly by calling Coverage.not_covered and Coverage.hot_spots.
1070
1071		----------------------------------------------------------------------
1072		Name: Matthias Blume (blume (at) tti - c (dot) org)
1073		Date: 2004/10/14 17:40:00 CDT
1074		Tag: blume-20041014-tdp-core
1075		Description:
1076
1077		Snapshot of a significant overhaul of how the trace/debug/profile support
1078		is hooked into the system (specifically: Core and SMLofNJ.Internals).
1079
1080		----------------------------------------------------------------------
1081		Name: Matthias Blume (blume (at) tti - c (dot) org)
1082		Date: 2004/10/13 16:34:00 CDT
1083		Tag: blume-20041013-tdp
1084		Description:
1085
1086		Some rationalization of names:
1087
1088		structure BTrace -> structure TDPInstrument
1089		etc.
1090
1091		This is is preparation of using the original back-trace
1092		instrumentation for other purposes.  "TDP" stands for
1093		Trace/Debug/Profile.
1094
1095		The control flag controlling whether instrumentation is on or off is now
1096		registered under a different name, so instead of running sml as
1097
1098		sml -Cinstrument.btrace-mode=true
1099
1100		one has to say
1101
1102		sml -Ctdp.instrument=true
1103
1104		----------------------------------------------------------------------
1105		Name: Matthias Blume (blume (at) tti - c (dot) org)
1106		Date: 2004/10/11 16:37:00 CDT
1107		Tag: blume-20041011-regions
1108		Description:
1109
1110		Made some minor modifications to elabcore.sml to have source regions
1111		be propagated more tightly -- resulting in better (i.e., smaller)
1112		regions being reported in error- and debug messages.
1113
1114		----------------------------------------------------------------------
1115		Name: Matthias Blume (blume (at) tti - c (dot) org)
1116		Date: 2004/10/08 22:50:00 CDT
1117		Tag: blume-20041008-cmkw
1118		Description:
1119
1120		Fixed handling of keywords in .cm files: After seeing "is" the lexer
1121		treats subsequent occurrences of "group", "library", "source", "is",
1122		"*", and "-" as ordinary identifiers rather than keywords.
1123
1124		Most seriously, this fixes a problem with CM's "shell" tool.  The tool
1125		is supposed to accept a tool argument called "source", but this did
1126		not work because of the clash with the keyword.
1127
1128		----------------------------------------------------------------------
1129		Name: Matthias Blume (blume (at) tti - c (dot) org)
1130		Date: 2004/10/07 16:00:00 CDT
1131		Tag: blume-20041007-cleanup
1132		Description:
1133
1134		Assorted cleanup work:
1135
1136		- got rid of intstrmap in favor of using the library's
1137		hash table implementation
1138		- threw out most of the pathnames stuff, as it was not used anyway
1139		- simplified tokentable implementation
1140		- fixed some minor spelling errors
1141
1142		----------------------------------------------------------------------
1143		Name: Matthias Blume (blume (at) tti - c (dot) org)
1144		Date: 2004/10/06 15:15:15 CDT
1145		Tag: blume-20041006-handler
1146		Description:
1147
1148		Cleaned up the absyn to reflect the invariant that HANDLE always
1149		carries a FNexp as part of the type definition.  This eliminates some
1150		superfluous sanity checks at runtime down the road.
1151
1152		Some minor cleanup of the btrace code.
1153
1154		----------------------------------------------------------------------
1155		Name: Matthias Blume (blume (at) tti - c (dot) org)
1156		Date: 2004/10/01 10:20:30 CDT
1157		Tag: blume-20041001-slave
1158		Description:
1159
1160		Added hack to make slave mode work in the presence of the version
1161		tool.  (Still, since the master does two passes over the code for
1162		CMB.make, the release number gets bumped twice when slaves are
1163		attached.  I don't know if this is worth fixing...)
1164
1165		----------------------------------------------------------------------
1166		Name: Matthias Blume (blume (at) tti - c (dot) org)
1167		Date: 2004/09/30 10:55:00 CDT
1168		Tag: blume-20040930-version
1169		Description:
1170
1171		* Moved the "version" magic into its own little library under
1172		src/system/smlnj/internal.  This avoids expensive reconstruction of
1173		a stable src/compiler/core.cm.
1174
1175		* At the same time, structure CompilerVersion is now known as
1176		structure SMLNJVersion.
1177
1178		* Arranged for the version tool to NOT kick in when rebuilding the system
1179		(makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
1180		Also, loading the versiontool does not work when rebuilding the system
1181		because CM is not properly initialized at that time.
1182
1183		----------------------------------------------------------------------
1184		Name: Matthias Blume (blume (at) tti - c (dot) org)
1185		Date: 2004/09/29 14:00:00 CDT
1186		Tag: blume-20040929-autoversion
1187		Description:
1188
1189		Implemented some CM magic to have
1190		file src/compiler/TopLevel/main/version.sml
1191		generated automagically.
1192		The version is taken from two files: config/version and config/release.
1193		The first is expected to contain a two-part version number such as 110.49.
1194		The second should contain a single number, but it may be missing.
1195
1196		If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
1197		time the version tool is loaded (which is the first time you say CMB.make),
1198		then the tool will increment the value stored in config/release every
1199		time CMB.make is invoked.
1200
1201		The binfile format is now insensitive to anything beyond the first
1202		two components of a version number, so bumping the release does not render
1203		binfiles incompatible.  Auto-bumping can be used to keep track of versions
1204		during development without invalidating existing binfiles.
1205
1206		In any case, every CMB.make updates the date information in version.sml.
1207		(This is the date that is printed in the banner.)
1208
1209		----------------------------------------------------------------------
1210		Name: Matthias Blume (blume (at) tti - c (dot) org)
1211		Date: 2004/09/28 10:53:00 CDT
1212		Tag: blume-20040928-controls
1213		Description:
1214
1215		Some cleanup of the controls code.
1216
1217		----------------------------------------------------------------------
1218		Name: Matthias Blume (blume (at) tti - c (dot) org)
1219		Date: 2004/09/27 22:08:00 CDT
1220		Tag: blume-20040927-controls
1221		Description:
1222
1223		Added two pieces of functionality to the Controls interface:
1224
1225		1.   val save'restore: 'a control -> unit -> unit
1226
1227		grabs the current value of the control in stage 1 and restores it
1228		in stage 2.
1229
1230		2.   val set' : 'a control * 'a -> unit -> unit
1231
1232		stores the given value into the control in stage 2 (i.e., delayed)
1233		but does all error checking in stage 1.
1234		(This is for string controls that need to do parse their argument --
1235		something that might fail.  In some cases, notably in CM, one
1236		already knows the intended argument but wants to delay the actual
1237		assignment until a time when error recovery would be more difficult.)
1238
1239		Changed the handling of controls in tool arguments to classes "sml" and
1240		"lazysml":
1241		- use Controls.save'restore as a more robust way of restoring the
1242		old value (in particular: without having to re-parse the string)
1243		- use controls to handle the "overload" keyword in the init group
1244		(I believe this change actually fixes a long-standing obscure bug.)
1245
1246		----------------------------------------------------------------------
1247		Name: Matthias Blume (blume (at) tti - c (dot) org)
1248		Date: 2004/09/27 17:00:00 CDT
1249		Tag: blume-20040927-lazysml
1250		Description:
1251
1252		Added a new tool class called "lazysml" to CM's tool chest.  The only
1253		difference to "sml" is that compilation is done with Control.lazysml
1254		set to true.  A source of class "lazysml" is automatically recognized
1255		by a file name suffix of ".lml".
1256
1257		In addition to the above feature, the original class "sml" now also
1258		supports a tool argument "lazy" which has the same effect.  As a
1259		result, the following three lines are equivalent:
1260
1261		foo.sml : lazysml
1262		foo.sml : sml (lazy)
1263		foo.sml (lazy)
1264
1265		The setting goes into effect both during parsing and during
1266		compilation.  The original setting is restored right after parsing and
1267		after compilation, respectively.
1268
1269		In addition to all the above, there is also a general mechanism to set
1270		ANY of the "controls" that are available at the command line via
1271		"-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
1272		apply as well. (In fact, "lazy" is implemented as a special case of
1273		the general mechanism.)
1274
1275		The .cm file syntax uses a new keyword tool argument called "with".
1276		There are several ways of indicating the desired settings:
1277
1278		foo.sml (with:parser.quotations=true)
1279		foo.sml (with:(name:parser.quotations value:true))
1280		foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
1281		foo.sml (with:(name1=value1 name2=value2 ...))
1282		foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
1283
1284		etc.
1285
1286		Another possible abbreviation is to leave out the =v or value:v part
1287		if the name refers to a boolean control (in which case the value is
1288		taken to be true).  Thus, one could get lazy sml also by saying:
1289
1290		foo.sml (with:parser.lazy-keyword=true)
1291		foo.sml (with:parser.lazy-keyword)
1292		foo.sml (with:(name:parser.lazy-keyword value:true))
1293		foo.sml (with:(name:parser.lazy-keyword))
1294
1295		----------------------------------------------------------------------
1296		Name: Matthias Blume (blume (at) tti - c (dot) org)
1297		Date: 2004/09/24 16:22:00 CDT
1298		Tag: blume-20040924-ppc-long-branch
1299		Description:
1300
1301		Turned message about "emiting long form of branch" off by default.
1302		Added a control flag to turn it back on when desired.
1303
1304		----------------------------------------------------------------------
1305		Name: Matthias Blume (blume (at) tti - c (dot) org)
1306		Date: 2004/09/24 16:05:00 CDT
1307		Tag: blume-20040924-rounding
1308		Description:
1309
1310		Applied patch for setting rounding modes under Mac OS X.  Thanks to
1311		Melissa O'Neill for providing the code!
1312
1313		----------------------------------------------------------------------
1314		Name: Matthias Blume (blume (at) tti - c (dot) org)
1315		Date: 2004/09/23 17:30:00 CDT
1316		Tag: blume-20040923-envvars
1317		Description:
1318
1319		1. Changed definition of type ControlRegistry.registry_tree to
1320		include control_info (i.e., the name of the controlling environment
1321		variable).
1322
1323		2. Added command-line flags -e and -E to print the names of environment
1324		variables that can be used to control internal settings.
1325		(This uses the new API mentioned in 1.)
1326
1327		----------------------------------------------------------------------
1328		Name: Matthias Blume (blume (at) tti - c (dot) org)
1329		Date: 2004/09/13 16:50:00 CDT
1330		Tag: Release_110_49
1331		Description:
1332
1333		New working version (110.49).  NEW BOOTFILES!
1334
1335		----------------------------------------------------------------------
1336		Name: Matthias Blume
1337		Date: 2004/09/13 16:20:00 CDT
1338		Tag: blume-20040913-config-mlrisc
1339		Description:
1340
1341		Put target "mlrisc" back into the default list.
1342		(There is no harm in having it, and some users have expressed their
1343		wish to have "mlrisc" included by default.)
1344
1345		----------------------------------------------------------------------
1346		Name: John Reppy
1347		Date: 2004/09/13
1348		Tag: jhr-20040913-signals
1349		Description:
1350
1351		Fixed the signal masking code to properly nest mask/unmask operations
1352		on a per-signal basis.
1353
1354		----------------------------------------------------------------------
1355		Name: Matthias Blume (blume (at) tti - c (dot) org)
1356		Date: 2004/09/08 13:20:00 CDT
1357		Tag: blume-20040908-heap-magic
1358		Description:
1359
1360		Bumped the heap macig number to 0x09082004 to account for the changed
1361		layout of the ML frame under MacOS X.
1362
1363		----------------------------------------------------------------------
1364		Name: Allen Leung (leunga (at) reservoir (dot) com)
1365		Date: 2004/09/03 11:26:00 EST
1366		Tag: leunga-20040903-cygwin-install
1367		Description:
1368
1369		Added a patch to _arch-n-opsys to enable the Cygwin runtime.
1370		The Cygwin runtime is turned on by setting the environment
1371		variable SMLNJ_CYGWIN_RUNTIME to 1.
1372
1373		----------------------------------------------------------------------
1374		Name: Matthias Blume (blume (at) tti - c (dot) org)
1375		Date: 2004/08/31 17:15:00 CDT
1376		Tag: blume-20040831-core
1377		Description:
1378
1379		Added some exports to src/compiler/core.cm upon request by J. Joemann.
1380
1381		----------------------------------------------------------------------
1382		Name: Matthias Blume (blume (at) tti - c (dot) org)
1383		Date: 2004/08/30 17:55:00 CDT
1384		Tag: blume-20040830-installer
1385		Description:
1386
1387		Upon request by Johannes Joemann:
1388
1389		- improved ML code of installer to fall back to coping when renaming
1390		fails (i.e., when source and target are on different file systems);
1391		the code compiles but has yet to be tested in anger
1392		- removed mlrisc from list of default targets (config/targets)
1393
1394		----------------------------------------------------------------------
1395		Name: Matthias Blume (blume (at) tti - c (dot) org)
1396		Date: 2004/08/27 17:20:00 CDT
1397		Tag: blume-20040827-ptreql
1398		Description:
1399
1400		Added ptreql primop to structure InlineT (upon request from Larry
1401		Paulson).
1402
1403		----------------------------------------------------------------------
1404		Name: Allen Leung (leunga (at) reservoir (dot) com)
1405		Date: 2004/08/15 21:21:00 EST
1406		Tag: leunga-110_48-udgraph
1407		Description:
1408
1409		Another bug fix from Carl Hauser:
1410
1411		diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
1412		> 48c48
1413		> <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
1414		> e::L)
1415		> ---
1416		> >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
1417		> Without this, any deletion of an edge in an undirected graph does severe
1418		> violence to the graph.
1419
1420		----------------------------------------------------------------------
1421		Name: Allen Leung (leunga (at) reservoir (dot) com)
1422		Date: 2004/08/10 23:35:00 EST
1423		Tag: leunga-110_48-ppc
1424		Description:
1425
1426		The IBM/MacOS syntax switch on PPC was incorrectly swapped.
1427
1428		----------------------------------------------------------------------
1429		Name: Matthias Blume (blume (at) tti - c (dot) org)
1430		Date: 2004/08/10 12:00:00 CDT
1431		Tag: Release_110_48
1432		Description:
1433
1434		New working version (110.48).  NEW BOOTFILES!
1435
1436		----------------------------------------------------------------------
1437		Name: Allen Leung (leunga (at) reservoir (dot) com)
1438		Date: 2004/08/09 12:21:00 EST
1439		Tag: leunga-110_47-dijsktra
1440		Description:
1441
1442		Bug fix from Carl Hauser:
1443
1444		single_source_shortest_paths in dijkstra.sml was observed to get wrong
1445		answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1446
1447		The problem is that following the expression A.update(dist,s,Num.zero)
1448		it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1449
1450		----------------------------------------------------------------------
1451		Name: Matthias Blume (blume (at) tti - c (dot) org)
1452		Date: 2004/08/06 18:10:00 CDT
1453		Tag: blume-20040806-cmdline
1454		Description:
1455
1456		Fiddled with handling of command-line options:
1457
1458		* sml now quits after processing the command line
1459		if -H, -S, -h<n>, or -s<n> appears as the last
1460		command-line argument
1461		* a new option -q terminates the session when encountered on
1462		the command line; subsequent arguments will be ignored
1463		* bug fixes: short (erroneous) arguments are no longer ignored
1464		completely
1465
1466		----------------------------------------------------------------------
1467		Name: Allen Leung (leunga (at) reservoir (dot) com)
1468		Date: 2004/08/04 18:17:00 EST
1469		Tag: leunga-110_47-ppc-ibm-asm
1470		Description:
1471
1472		- Added minimal IBM assembly syntax support for PowerPC.
1473
1474		- Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1475		been altered in the runtime.  We need an automatic way to keep the file
1476		in sync.
1477
1478		----------------------------------------------------------------------
1479		Name: Matthias Blume (blume (at) tti - c (dot) org)
1480		Date: 2004/08/04 14:00:00 CDT
1481		Tag: Release_110_47
1482		Description:
1483
1484		New working version (110.47).  NEW BOOTFILES!
1485
1486		----------------------------------------------------------------------
1487		Name: Matthias Blume (blume (at) tti - c (dot) org)
1488		Date: 2004/08/03 14:25:00 CDT
1489		Tag: blume-20040803-callingconv
1490		Description:
1491
1492		Added low-level support for choosing C calling conventions by
1493		twiddling the type of rawccall.  (See
1494		src/compiler/Semant/types/cproto.sml for details.)
1495
1496		----------------------------------------------------------------------
1497		Name: Matthias Blume (blume (at) tti - c (dot) org)
1498		Date: 2004/08/02 15:55:00 CDT
1499		Tag: blume-20040802-backout
1500		Description:
1501
1502		Backed out of change to win32-filesys.c.  The earlier patch to
1503		get_file_time caused CM to produce files with the wrong time stamp.
1504
1505		----------------------------------------------------------------------
1506		Name: Matthias Blume (blume (at) tti - c (dot) org)
1507		Date: 2004/08/02 14:45:00 CDT
1508		Tag: blume-20040802-nlffi-win32
1509		Description:
1510
1511		Added NLFFI support for Win32, adapted from a patch provided by David
1512		Hansel.  This is currently completely untested.  Also, the issue
1513		concerning stdcall vs. ccall is still unresolved.
1514
1515		----------------------------------------------------------------------
1516		Name: Matthias Blume (blume (at) tti - c (dot) org)
1517		Date: 2004/07/30 17:55:00 CDT
1518		Tag: blume-20040730-various
1519		Description:
1520
1521		Gearing up towards 110.47...
1522
1523		- various minor bugfixes to ml-nlffigen
1524		- a beginning of a manual for nlffi
1525
1526		- eliminated 'export name=value' in config/install.sh as this does
1527		not work with certain versions of /bin/sh
1528		(Thanks to David King at Motorola for catching this.)
1529
1530		- several bugfixes provided or suggested by David Hansel at Reactive Systems:
1531		- added a test for tm==NULL to gmtime.c and localtime.c
1532		- applied patch for incorrect GetFileTime under win32
1533		- toSeconds -> toMilliseconds in Win32/win32-process.sml
1534
1535		----------------------------------------------------------------------
1536		Name: Matthias Blume (blume (at) tti - c (dot) org)
1537		Date: 2004/07/21 18:20:00 CDT
1538		Tag: blume-20040721-nlffigen
1539		Description:
1540
1541		- Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1542		for a typedef to an incomplete type, but leave out the "typ" member.
1543		(This is just for consistency.)
1544		- Started to produce what is supposed to become better (i.e., comprehensive)
1545		documentation of what ml-nlffigen does and produces.
1546
1547		----------------------------------------------------------------------
1548		Name: Matthias Blume (blume (at) tti - c (dot) org)
1549		Date: 2004/07/14 16:25:00 CDT
1550		Tag: blume-20040714-union
1551		Description:
1552
1553		Added C_UNION to c-calls/c-types.sml and updated the machinery
1554		(ml-nlffigen, cproto.sml) that conveys C function interface
1555		information to the code generator.
1556
1557		However, the actual architecture-specific implementation of function
1558		arguments and results that are C unions is still not implemented.
1559
1560		----------------------------------------------------------------------
1561		Name: Allen Leung (leunga (at) reservoir (dot) com)
1562		Date: 2004/07/14 14:38:00 EST
1563		Tag: leunga-110_46_1-ppc-lwzu
1564		Description:
1565
1566		Added these instructions to the PowerPC architecture:
1567		LBZU(X), LHZU(X), LWZU(X),
1568		STWU(X), STFDU, STFSU
1569
1570		etc...
1571
1572		Note: I haven't added their instruction encoding into the description.
1573
1574		----------------------------------------------------------------------
1575		Name: Allen Leung (leunga (at) reservoir (dot) com)
1576		Date: 2004/07/13 15:04:00 EST
1577		Tag: leunga-110_46_1-ppc-lwarx
1578		Description:
1579
1580		Added the two instructions LWARX and STWCX to the PowerPC
1581		instruction set.
1582
1583		A (untested) rewrite of loop-structure.sml.  The old version
1584		is completely broken.
1585
1586		----------------------------------------------------------------------
1587		Name: Matthias Blume (blume (at) tti - c (dot) org)
1588		Date: 2004/07/13 13:50:00 CDT
1589		Tag: blume-20040713-nlffi
1590		Description:
1591
1592		- use paramAlloc to report c-calls with too many arguments
1593		(for PPC version where parameter area is pre-allocated)
1594		- added ccall_maxargspace to machspec (to implement the above)
1595		- made "make" commend in CM's "make" tool configurable
1596		- added option (default: on) for passing the name of the SML/NJ's "bin"
1597		directory to "make"; the call looks like this:
1598
1599		make <options> SMLNJ_BINDIR=<dir> <target>
1600
1601		This can be used by the Makefile to, e.g., pick the "right" version
1602		of ml-nlffigen.
1603		- minor code tweaks
1604
1605		----------------------------------------------------------------------
1606		Name: Matthias Blume (blume (at) tti - c (dot) org)
1607		Date: 2004/07/12 22:50:00 CDT
1608		Tag: blume-110_46_1-macosx-nlffi
1609		Description:
1610
1611		NLFFI under Mac OS X now working (sort of).  This is largely untested,
1612		though.
1613
1614		Note:  1. You have to make a new, clean build of the runtime system.
1615		2. There are new BOOTFILES, you have to use them!
1616		(Doing the bootstrap process yourself would be *very* painful!
1617		If you absolutely have to do it, build the system under
1618		a different architecture and then cross-compile.)
1619
1620		Version bumped to 110.46.1 to account for runtime data format changes.
1621
1622		----------------------------------------------------------------------
1623		Name: Matthias Blume (blume (at) tti - c (dot) org)
1624		Date: 2004/06/18 14:30:00 CDT
1625		Tag: blume-20040618-unix
1626		Description:
1627
1628		Changed the implementation of structure Unix so that the same stream
1629		is returned every time one of the {text,bin}{In,Out}streamOf functions
1630		is invoked on the same proc.  This is not what the spec currently
1631		says -- although IMO it arguably should.  (See discussion below.)
1632
1633		----------------------------------------------------------------------
1634		Name: Matthias Blume (blume (at) tti - c (dot) org)
1635		Date: 2004/06/17 18:15:00 CDT
1636		Tag: Release_110_46
1637		Description:
1638
1639		New working version (110.46).  NEW BOOTFILES!
1640
1641		----------------------------------------------------------------------
1642		Name: Matthias Blume (blume (at) tti - c (dot) org)
1643		Date: 2004/06/17 17:20:00 CDT
1644		Tag: blume-20040617-timer-unix
1645		Description:
1646
1647		Changed the interface of structures Timer and Unix to match the most
1648		recent Basis spec.
1649
1650		In the case of Unix there still seems to be an open/weird issue:
1651
1652		The {text,bin}{In,Out}streamOf functions are supposed to create
1653		fresh streams whenever they are called -- as opposed to have them
1654		return the same stream every time.  This design is supposed to
1655		prevent space leaks caused by proc values hanging on to streams.
1656
1657		The reap function, on the other hand, is supposed to close the
1658		streams.  This cannot be done without having a handle on the
1659		stream in proc after all...
1660
1661		I took the liberty to implement the following stopgap solution:
1662
1663		The proc value hangs on to the most recently created stream(s).
1664		Reap closes those.  If either or both of the two streams hadn't
1665		been created at all yet, then reap will close the corresponding
1666		file descriptors directly.
1667
1668		PS: I don't understand the original space leak argument anymore.  If
1669		a proc hangs on to the imperative stream, then I/O operations on those
1670		will advance the state of the cached stream and avoid the space leak.
1671
1672		----------------------------------------------------------------------
1673		Name: Matthias Blume (blume (at) tti - c (dot) org)
1674		Date: 2004/05/28 16:45:00 CDT
1675		Tag: blume-20040528-basis
1676		Description:
1677
1678		Added signature PACK_REAL and exported functor PrimIO.
1679
1680		----------------------------------------------------------------------
1681		Name: Matthias Blume (blume (at) tti - c (dot) org)
1682		Date: 2004/05/25 16:00:00 CDT
1683		Tag: blume-20040525-group-owner
1684		Description:
1685
1686		CM now ignores (but still accepts) the "owner" information in group
1687		descriptions.  The owner of a group is its next enclosing
1688		library. Each group must have a unique owner.  (There is a virtual
1689		"toplevel" library that owns groups which are not nested within a real
1690		library.)  Previously, each group had to explicitly declare its owner,
1691		and CM would check that such a declaration is correct.  The new scheme
1692		is to have CM check that for each group there is precisely one owning
1693		library.
1694
1695		The advantage of the new scheme is that the programmer no longer needs
1696		to maintain the somewhat annoying owner information.  The downside is
1697		that CM cannot enforce the ownership rule across multiple runs of
1698		CM.make.  Fortunately, enclosing the same group in two different
1699		libraries A and B which are not part of the same program does not
1700		cause real problems.
1701
1702		----------------------------------------------------------------------
1703		Name: Matthias Blume (blume (at) tti - c (dot) org)
1704		Date: 2004/05/20 16:00:00 CDT
1705		Tag: blume-20040520-win32
1706		Description:
1707
1708		Made the win32 version work again.  (Strangely, a misplaced comma had
1709		slipped into win32-process.c which prevented the runtime from being
1710		compiled correctly.)
1711
1712		Also, included a minor addition to ml-build.bat analogous to what was
1713		done in blume-20040519-ml-build.
1714
1715		----------------------------------------------------------------------
1716		Name: Matthias Blume (blume (at) tti - c (dot) org)
1717		Date: 2004/05/19 22:10:00 CDT
1718		Tag: blume-20040519-ml-build
1719		Description:
1720
1721		Arranged for ml-build to clean up after itself a little bit better.
1722		The script generates a temporary SML source file and compiles it using
1723		CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1724		gets rid of those at the end, so they don't accumulate under .cm.
1725
1726		This required a minor change to install.sh because the name of the
1727		metadata directory (default: .cm) is actually configurable at
1728		installation time.
1729
1730		----------------------------------------------------------------------
1731		Name: Matthias Blume (blume (at) tti - c (dot) org)
1732		Date: 2004/05/18 15:50:00 CDT
1733		Tag: blume-20040518-mkreader
1734		Description:
1735
1736		Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1737		implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1738
1739		----------------------------------------------------------------------
1740		Name: Matthias Blume (blume (at) tti - c (dot) org)
1741		Date: 2004/05/11 14:35:00 CDT
1742		Tag: blume-20040511-win32sock
1743		Description:
1744
1745		Added previously missing support for many socket-related functions
1746		under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1747		for the voluminous patch!
1748
1749		(I have not tested this patch under win32 yet.)
1750
1751		Here is David's e-mail:
1752
1753		Hi,
1754
1755		Attached to this email you find a diff against sml/nj 110.45
1756		that will enable socket support under Windows.
1757
1758		To apply the patch (using unix or cygwin)
1759		1) gunzip runtime.diff.gz
1760		2) "cd" into "src/runtime" in the source tree of a fresh
1761		110.45 installation.
1762		3) patch -p 1 < [your/path/to]runtime.diff
1763
1764		The code compiles fine but has NOT yet been extensively tested.
1765		I only ran a few tests for basic socket client functionality
1766		(which worked fine).  Especially the functions that use ioctl
1767		are not tested at all and might not work (see below).
1768
1769		I implemented this since we want to move to a newer version of sml/nj
1770		but need socket support in order to use it.  This is the first time I
1771		even had a look at the sml/nj source,  so please review my changes
1772		before making this part of the distribution!  Here are a few issues
1773		that I think might be better for someone to solve who is more
1774		familiar with the sml/nj source (and socket programming):
1775
1776		- getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1777		exception since I could not figure out what the windows equivalent
1778		of these functions is
1779
1780		- In sockets-osdep.h there are a some #include statements that are
1781		only used in a few files that include sockets-osdep.h
1782
1783		- In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1784		does not process its return value since I don't know how to
1785		report an error upwards.
1786
1787		- It would probably be good to have a call to WSACleanup() when
1788		the library is unloaded (if there is such a possibility).
1789		Otherwise I think Windows will take care of this automatically
1790		when the process finishes.
1791
1792		- I used ioctlsocket() as a replacement for ioctl() but I have
1793		no idea if that is actually the proper replacement on Windows.
1794
1795		- All these issues are marked in the code by "FIXME" comments.
1796
1797		We use sml/nj extensively in our products and are quite happy
1798		with it.  I hope this contribution will help you.
1799
1800		Keep up the good work!
1801
1802		David
1803
1804		----------------------------------------------------------------------
1805		Name: Matthias Blume (blume (at) tti - c (dot) org)
1806		Date: 2004/05/11 14:20:00 CDT
1807		Tag: blume-20040511-installml
1808		Description:
1809
1810		Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1811		the patch.)
1812
1813		----------------------------------------------------------------------
1814		Name: Matthias Blume (blume (at) tti - c (dot) org)
1815		Date: 2004/05/11 14:05:00 CDT
1816		Tag: blume-20040511-nlffi-netbsd
1817		Description:
1818
1819		Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1820		the patch.)
1821
1822		----------------------------------------------------------------------
1823		Name: Matthias Blume (blume (at) tti - c (dot) org)
1824		Date: 2004/05/11 12:05:00 CDT
1825		Tag: blume-20040511-exports
1826		Description:
1827
1828		As per request by Adam Chlipala <adam@hcoop.net>, extended various
1829		export lists in compiler-related .cm-files.
1830
1831		----------------------------------------------------------------------
1832		Name: Matthias Blume (blume (at) tti - c (dot) org)
1833		Date: 2004/05/11 11:35:00 CDT
1834		Tag: blume-20040511-allsource
1835		Description:
1836
1837		The installer now honors the "src-smlnj" target again, although its meaning
1838		has changed from "all sources required for the compiler" to "all sources
1839		the installer knows about".  In other words, if you enable "src-smlnj"
1840		in the "targets" file, then the installer will pull in sources for
1841		everything.  (Notice that this refers to source code only.  Compiled
1842		code is still only installed for modules that were requested explicitly
1843		or which are required for other modules that were requested explicitly.)
1844
1845		----------------------------------------------------------------------
1846		Name: Matthias Blume (blume (at) tti - c (dot) org)
1847		Date: 2004/04/23 17:40:00 CDT
1848		Tag: blume-20040423-ieee-scan
1849		Description:
1850
1851		Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1852		in the exponent calculation we get INF or ZERO (depending on the mantissa
1853		and the sign of the exponent).
1854
1855		----------------------------------------------------------------------
1856		Name: Matthias Blume (blume (at) tti - c (dot) org)
1857		Date: 2004/04/23 10:40:00 CDT
1858		Tag: blume-20040423-ml-build
1859		Description:
1860
1861		The ml-build script now terminates with a non-0 status when something
1862		goes wrong.
1863
1864		----------------------------------------------------------------------
1865		Name: Matthias Blume (blume (at) tti - c (dot) org)
1866		Date: 2004/04/22 16:35:00 CDT
1867		Tag: blume-20040422-Option
1868		Description:
1869
1870		Made exception Option to be the same as exception Option.Option
1871		(as it should be).
1872
1873		----------------------------------------------------------------------
1874		Name: Allen Leung (leunga (at) reservoir (dot) com)
1875		Date: 2004/03/19 14:40:00 EST
1876		Tag: leunga-20040319-cygwin-nlffi
1877		Description:
1878
1879		Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1880		of SML/NJ.  The problem is that
1881
1882		lib = dlopen(NULL, ...)
1883		f   = dlsym(lib, "malloc");
1884
1885		does not work on Windows unless we explicitly export symbols
1886		such as 'malloc' during linking.  We fixed this by explicitly
1887		exporting the required symbols with the magic gcc incantation:
1888
1889		-Wl,--export-all cygwin.def
1890
1891		where cygwin.def is a file containing all the symbols that
1892		we wish to export.
1893
1894		I suspect this is a Windows problem and we'll have to
1895		do the same (somehow with windows compilers) when
1896		we build the native win32 version with the system
1897		calls LoadLibrary/GetProcAddress.
1898
1899		----------------------------------------------------------------------
1900		Name: Matthias Blume (blume (at) tti - c (dot) org)
1901		Date: 2004/03/04 16:35:00 CST
1902		Tag: blume-20040304-intinf-fmt
1903		Description:
1904
1905		Fixed problem with IntInf.fmt (sign would show up on the right instead
1906		of on the left for BIN, OCT, and HEX).
1907
1908		----------------------------------------------------------------------
1909		Name: Matthias Blume (blume (at) tti - c (dot) org)
1910		Date: 2004/03/04 11:25:00 CST
1911		Tag: blume-20040304-symlinks
1912		Description:
1913
1914		Fixed problem with installer script (unix only) where bin/ml-yacc and
1915		friends pointed (via symlinks) to absolute locations instead of just
1916		.run-sml.  This was reported by Vesa A Norrman.
1917
1918		----------------------------------------------------------------------
1919		Name: Matthias Blume (blume (at) tti - c (dot) org)
1920		Date: 2004/02/13 14:50:00 CST
1921		Tag: Release_110_45
1922		Description:
1923
1924		New working version (110.45). New bootfiles.
1925
1926		----------------------------------------------------------------------
1927		Name: Matthias Blume (blume (at) tti - c (dot) org)
1928		Date: 2004/01/26 15:15:15 CST
1929		Tag: blume-20040126-toplevel
1930		Description:
1931
1932		Improved handling of exceptions at the interactive toplevel.
1933
1934		----------------------------------------------------------------------
1935		Name: Matthias Blume (blume (at) tti - c (dot) org)
1936		Date: 2004/01/26 11:25:00
1937		Tag: blume-20040126-app
1938		Description:
1939
1940		Type of top-level "app" corrected.
1941		Added code for setting vp_limitPtrMask to Win32-specific runtime.
1942
1943		----------------------------------------------------------------------
1944		Name: Matthias Blume (blume (at) tti - c (dot) org)
1945		Date: 2003/11/18 17:10 CST
1946		Tag: blume-20031118-basis-fiddle
1947		Description:
1948
1949		- changed Timer interface to what might become the spec
1950		- POSIX_FLAGS -> BIT_FLAGS according to spec
1951		- some other minor discrepancies wrt. spec eliminated
1952
1953		----------------------------------------------------------------------
1954		Name: Matthias Blume (blume (at) tti - c (dot) org)
1955		Date: 2003/11/06 12:00:00 CST
1956		Tag: Release_110_44
1957		Description:
1958
1959		New working version (110.44). New bootfiles.
1960
1961		----------------------------------------------------------------------
1962		Name: Matthias Blume (blume (at) tti - c (dot) org)
1963		Date: 2003/11/04 11:50:00 CST
1964		Tag: blume-20031104-move-libraries
1965		Description:
1966
1967		Eliminated the "dont_move_libraries" directive in config/targets.
1968		(The mechanism was broken and could not be fixed easily.  Moreover,
1969		there does not seem to be any reason not to move all libraries into
1970		lib during installation.  I originally implemented this directive as a
1971		backward-compatibility feature when I first introduced the new CM.
1972		Now that things have been stable for a long time and going back to the
1973		old CM is not an option, there is no reason to keep it around.)
1974
1975		----------------------------------------------------------------------
1976		Name: Matthias Blume (blume (at) tti - c (dot) org)
1977		Date: 2003/11/03 16:00:00 CST
1978		Tag: blume-20031103-installdir
1979		Description:
1980
1981		Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1982		Richards for pointing out the problem and providing the solution.)
1983
1984		----------------------------------------------------------------------
1985		Name: Matthias Blume (blume (at) tti - c (dot) org)
1986		Date: 2003/10/01 17:05:00 CDT
1987		Tag: blume-20031001-lal-mlrisc
1988		Description:
1989
1990		MLRISC bug fix from Lal.
1991
1992		----------------------------------------------------------------------
1993		Name: Matthias Blume (blume (at) tti - c (dot) org)
1994		Date: 2003/09/30 16:10:00 CDT
1995		Tag: blume-20030930-primio-bat
1996		Description:
1997
1998		1. Added openVector, nullRd, and nullWr to PRIM_IO.
1999		2. Improved .bat files (for Win32 port) to make things work under Win95.
2000		(thanks to Aaron S. Hawley for this one)
2001
2002		----------------------------------------------------------------------
2003		Name: Matthias Blume (blume (at) tti - c (dot) org)
2004		Date: 2003/09/26 16:05:00 CDT
2005		Tag: blume-20030926-wrappriv
2006		Description:
2007
2008		Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
2009
2010		----------------------------------------------------------------------
2011		Name: Matthias Blume (blume (at) tti - c (dot) org)
2012		Date: 2003/09/26 15:00:00 CDT
2013		Tag: blume-20030926-110_43_3
2014		Description:
2015
2016		- additional cleanup
2017		- version number bump, NEW BOOTFILES
2018
2019		----------------------------------------------------------------------
2020		Name: Matthias Blume (blume (at) tti - c (dot) org)
2021		Date: 2003/09/26 12:00:00 CDT
2022		Tag: blume-20030926-ppautoload
2023		Description:
2024
2025		I modified the read-eval-print loop so that the autoloader gets
2026		invoked whenever the prettyprinter tries to look up a symbol that
2027		is not currently defined in the toplevel environment but which
2028		appears in CM's autoload registry.  As a result, we see far fewer of
2029		those ?.Foo.Bar.xxx names in the prettyprinter's output.
2030
2031		In addition to this I tried to clean up some pieces of the Basis
2032		implementation (e.g., Socket, Word8Array) in order to prevent other
2033		instances of these ?.Foo.Bar.xxx names from being printed.
2034
2035		The mechanism that picks names for types still needs some work, though.
2036		(Right now it seems that if there is a type A.t which is defined to
2037		be B.u, but B is unavailable at toplevel, then A.t gets printed as
2038		"?.B.u" although the perhaps more sensible solution would be to use
2039		"A.t" in this case.  In other words, the prettyprinter should follow
2040		a chain of DEFtycs not farther than there are corresponding toplevel
2041		names in the current environment.)
2042
2043		----------------------------------------------------------------------
2044		Name: Matthias Blume (blume (at) tti - c (dot) org)
2045		Date: 2003/09/24 16:31:00 CDT
2046		Tag: blume-20030924-installer
2047		Description:
2048
2049		Another installer tweak:  All the ML code for the installer is now
2050		compiled during CMB.make and put into a little library called
2051		$smlnj/installer.cm.  The installation then simply invokes
2052
2053		sml -m $smlnj/installer.cm
2054
2055		and everything happens automagically.
2056
2057		Win32:  ML code senses value of environment variable SMLNJ_HOME.
2058		Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
2059		and BINDIR.
2060
2061		The new scheme guarantees that the ML code responsible for the installation
2062		is in sync with the APIs of the main system.  Also, the installer is
2063		somewhat faster because the installer script is precompiled.
2064
2065		----------------------------------------------------------------------
2066		Name: Matthias Blume (blume (at) tti - c (dot) org)
2067		Date: 2003/09/24 15:35:00 CDT
2068		Tag: blume-20030924-synsock
2069		Description:
2070
2071		Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
2072		but excludes all non-blocking operations.  Defined SOCKET (in Basis)
2073		and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
2074		implementations of non-blocking operations from CML's Socket
2075		structure.
2076
2077		----------------------------------------------------------------------
2078		Name: Matthias Blume (blume (at) tti - c (dot) org)
2079		Date: 2003/09/24 15:10:05 CDT
2080		Tag: blume-20030924-sockets
2081		Description:
2082
2083		1. Fixed SOCKET API and implementation to match Basis spec.
2084		This required changing the internal representation of sockets to one
2085		that remembers (for each socket file descriptor) whether it is currently
2086		blocking or non-blocking.  This state is maintained lazily (i.e., a system
2087		call is made only if the state actually needs to change).
2088
2089		2. OS-specific details of sockets were moved into separate files, thus
2090		making it possible to unify the bulk of the socket implementations
2091		between Unix and Win32.
2092
2093		3. CML's socket API changed accordingly.
2094		(Note that we need to remove non-blocking functions from this API
2095		since they are redundant in the case of CML!)
2096
2097		4. CML's socket implementation now makes use of non-blocking functions
2098		provided by Basis, thus removing all OS-dependent code from this part
2099		of CML.
2100
2101		5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
2102
2103		----------------------------------------------------------------------
2104		Name: Matthias Blume (blume (at) tti - c (dot) org)
2105		Date: 2003/09/22 12:10:00 CDT
2106		Tag: blume-20030922-110_43_2
2107		Description:
2108
2109		Made a new interim version and bootfiles for developer's bootstrapping
2110		convenience.
2111
2112		110.43.2 -- NEW BOOTFILES
2113
2114		----------------------------------------------------------------------
2115		Name: Matthias Blume (blume (at) tti - c (dot) org)
2116		Date: 2003/09/19 15:55:00 CDT
2117		Tag: blume-20030919-cmdir
2118		Description:
2119
2120		1. new-install.sh -> install.sh
2121		2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
2122		3. tweaked installer so that another name instead of .cm can be chosen
2123		at install time (by setting the CM_DIR_ARC environment variable
2124		during installation); once installation is complete, the name is
2125		fixed
2126
2127		----------------------------------------------------------------------
2128		Name: Matthias Blume (blume (at) tti - c (dot) org)
2129		Date: 2003/09/18 16:00:00 CDT
2130		Tag: blume-20030918-110_43_1
2131		Description:
2132
2133		Made a new interim version and bootfiles for developer's bootstrapping
2134		convenience.
2135
2136		110.43.1 -- NEW BOOTFILES
2137
2138		----------------------------------------------------------------------
2139		Name: Matthias Blume (blume (at) tti - c (dot) org)
2140		Date: 2003/09/18 15:20:00 CDT
2141		Tag: blume-20030918-misc
2142		Description:
2143
2144		1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
2145		this seems to be controversial at the moment) and used those in
2146		Posix.ProcEnv.times.
2147
2148		2. Added Time.{from,to}Nanoseconds to Time.
2149
2150		3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
2151		For example, fromLargeInt never needs to look at more than 3 "big
2152		digits" to get its 53 bits of precision.
2153
2154		----------------------------------------------------------------------
2155		Name: Matthias Blume (blume (at) tti - c (dot) org)
2156		Date: 2003/09/17 16:30:00 CDT
2157		Tag: blume-20030917-real32-slices
2158		Description:
2159
2160		Added an entry to the primitive environment
2161		(compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
2162		added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
2163
2164		Removed some of the "magic" constants in real64.sml and replaced them
2165		with code that generates these values from their corresponding
2166		integer counterparts.
2167
2168		Made all(?) the slice-related changes to the Basis and made everything
2169		compile again...
2170
2171		----------------------------------------------------------------------
2172		Name: Matthias Blume (blume (at) tti - c (dot) org)
2173		Date: 2003/09/15 17:45:00 CDT
2174		Tag: blume-20030915-rbase
2175		Description:
2176
2177		Fixed bug in Real.fromLargeInt.
2178
2179		----------------------------------------------------------------------
2180		Name: Matthias Blume (blume (at) tti - c (dot) org)
2181		Date: 2003/09/13 18:11:00 CDT
2182		Tag: blume-20030913-libinstall
2183		Description:
2184
2185		Minor bugfix in config/libinstall (set anchor with path to
2186		standalone tool after installing it, otherwise libraries that
2187		need ml-lex or ml-yacc won't compile the first time the installer
2188		runs).
2189
2190		----------------------------------------------------------------------
2191		Name: Matthias Blume (blume (at) tti - c (dot) org)
2192		Date: 2003/09/12 11:45:00 CDT
2193		Tag: blume-20030912-various
2194		Description:
2195
2196		- fixed bug in Real.toLargeInt
2197		- fixed bug in Posix.ProcEnv.times
2198		- changed inputLine functions to return an option
2199		- minor installer improvements / bugfixes
2200		- changed default @SMLalloc parameter for x86/celeron to 64k
2201
2202		----------------------------------------------------------------------
2203		Name: Matthias Blume (blume (at) tti - c (dot) org)
2204		Date: 2003/09/09 22:00:00 CDT
2205		Tag: Release_110_43
2206		Description:
2207
2208		New working release 110.43.  New bootfiles.
2209
2210		----------------------------------------------------------------------
2211		Name: Matthias Blume (blume (at) tti - c (dot) org)
2212		Date: 2003/09/09 19:20:00 CDT
2213		Tag: blume-20030909-installer
2214		Description:
2215
2216		Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
2217		Modified config/install.bat to take advantage of it.  Also modified
2218		config/install.sh (and called it config/new-install.sh) to take advantage
2219		of it on Unix systems.  (The SML code is (supposed to be) platform-
2220		independent.)
2221
2222		The installer can now install everything under Win32
2223		as well as under *nix as long as it compiles.
2224
2225		Other changes:
2226
2227		- made CML compile again under Win32
2228		- made eXene compile under Win32 (by providing a fake structure UnixSock
2229		and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
2230		- fixed a bug in nowhere: it assumed that type OS.Process.status is the
2231		same as type int;  under Win32 it isn't
2232		- fixed some slice-related problems in the win32-specific parts of CML
2233		- added a functor argument "sameVol" to os-path-fn.sml in the Basis
2234		(under Win32, the volume name is case-insensitive, and the
2235		OS.Path code compares volume names for equality)
2236
2237		----------------------------------------------------------------------
2238		Name: Matthias Blume (blume (at) tti - c (dot) org)
2239		Date: 2003/09/08 11:55:00 CDT
2240		Tag: blume-20030908-fullpath
2241		Description:
2242
2243		Made Win32 version of OS.FileSys.fullPath return current directory
2244		when given an empty string.  This is what the spec says, and incidentally,
2245		CM depends on it.  (CM otherwise goes into an infinite loop in certain
2246		cases when presented with the name of a non-existing .cm file.)
2247
2248		----------------------------------------------------------------------
2249		Name: Matthias Blume (blume (at) tti - c (dot) org)
2250		Date: 2003/09/04 16:30:00 CDT
2251		Tag: blume-20030905-slices-etc
2252		Description:
2253
2254		1. Changed interface to vectors and arrays in Basis to match
2255		(draft) Basis spec.
2256		2. Added signatures and implementations of slices according to
2257		Basis spec.
2258		3. Edited source code throughout the system to make it compile again
2259		under 1. and 2.  (In some cases code had to be added to have it
2260		match the new signatures.)
2261		4. MLRISC should be backward-compatible: the copies of the originals
2262		of files that needed to change under 3. were retained, the .cm files
2263		check the compiler version number and use old versions when
2264		appropriate.
2265		5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
2266		dirstream -> string option (in accordance with Basis spec).
2267		6. When generating code that counts lines, ml-lex used function
2268		CharVector.foldli, taking advantage of its old interface.
2269		This has been replaced with the corresponding code from
2270		CharVectorSlice.  (html-lex must be re-lexed!)
2271		7. BitArray in smlnj-lib/Util has been extended/modified to match the
2272		new MONO_ARRAY signature.  (Do we need BitArraySlice?)
2273		8. Removed temporary additions (fromInternal, toInternal) from the
2274		(now obsolete) IntInf in smlnj-lib/Util.
2275		9. Cleaned up structure Byte.
2276		10. Added localOffset, scan, and fromString to Date (according to spec).
2277		Cleaned/corrected implementation of Date.
2278		(Still need to check for correctness; implement better canonicalizeDate.)
2279		11. Added "scan" to signature IEEE_REAL.
2280		12. Some improvements to IntInf [in particular: efficiency-hack for
2281		mod and rem when second operand is 2 (for parity checks).]
2282		13. Changed representation of type Time.time, using a single IntInf.int
2283		value counting microseconds.  This considerably simplified the
2284		implementation of structure Time.  We now support negative time
2285		values; scan and fromString handle signs.
2286		14. Functor PrimIO now takes two additional arguments (VectorSlice and
2287		ArraySlice).
2288
2289		----------------------------------------------------------------------
2290		Name: Matthias Blume (blume (at) tti - c (dot) org)
2291		Date: 2003/08/28 17:00:00 CDT
2292		Tag: blume-20030828-intinf
2293		Description:
2294
2295		This is a major update which comes with a version number bump
2296		(110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
2297		and an implementation of IntInf in the Basis.
2298
2299		There are a fairly large number of related changes and updates throughout
2300		the system:
2301
2302		Basis:
2303		- Implemented IntInf.
2304		- Made LargeInt a projection of IntInf (by filtering through INTEGER).
2305		- Added some missing Real64 operations, most notably Real.toLargeInt.
2306		- Added FixedInt as a synonym for Int32.
2307
2308		compiler:
2309		* Added support for a built-in intinf type.
2310		- literals
2311		- pattern matching
2312		- conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
2313		- overloading on literals and operations
2314
2315		This required adding a primitive type intinf, some additional
2316		primops, and implementations for several non-trivial intinf
2317		operations in Core.  (The intinf type is completely abstract
2318		to the compiler; all operations get delegated back to the Core.)
2319
2320		* Intinf equality is handled by polyequal.  However, the compiler
2321		does not print its usual warning in this case (since polyequal
2322		is the right thing to do there).
2323
2324		* Improved the organization of structure InlineT.
2325
2326		* A word about conversion primops:
2327		If conversions involving intinf do not cancel out during
2328		CPS contract, then the compiler must insert calls to Core functions.
2329		Since all core access must be resolved already during the FLINT
2330		translate phase, it would be too late a the time of CPS contract
2331		to add new Core calls.  For this reason, conversion primops
2332		for intinf carry two arguments:  1. the numeric argument that
2333		they are supposed to convert, and 2. the Core function that
2334		can help with this conversion if necessary.  If CPS contract
2335		eliminates a primop, then the associated Core function becomes
2336		dead and goes away.  Intinf conversion primops that do not get
2337		eliminated by CPS contract get rewritten into calls of their
2338		core functions by a separate, new phase.
2339
2340		interactive system:
2341		- Control.Print.intinfDepth controls max length of intinf constants
2342		being printed. (Analogous to Control.Print.stringDepth.)
2343		- Cleanup in printutil and pputil: got rid of unused stuff and
2344		duplicates; replaced some of the code with code that makes better
2345		use of library functionality.
2346
2347		CM:
2348		Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
2349		are no longer silent.
2350
2351		CKIT:
2352		Fixed mismatched uses of Int32 and LargeInt.  I always decided
2353		in favor of LargeInt -- which is now the same as IntInf.
2354		CKIT-knowledgable people should check whether this is what's
2355		intended and otherwise change things back to using Int32 or
2356		FixedInt.
2357
2358		Throughout the code:
2359		Started using IntInf.int literals and built-in operations
2360		(e.g., comparison with 0) where this seems appropriate.
2361
2362
2363		----------------------------------------------------------------------
2364		Name: Dave MacQueen (dbm@cs.uchicago.edu)
2365		Date: 2003/08/13 11:36:00 CDT
2366		Tag: dbm-20030813-mcz-merge1
2367		Description:
2368
2369		Merging changes from the mcz-branch development branch into trunk.
2370		These changes involve replacement of the emulated old prettyprinter
2371		interface with direct use of the SML/NJ Lib PP library, and fixing
2372		of a couple of bugs (895, 1186) relating to error messages.  A new
2373		prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
2374		has been added.
2375
2376		----------------------------------------------------------------------
2377		Name: Matthias Blume (blume (at) tti - c (dot) org)
2378		Date: 2003/08/11 15:45:00 CDT
2379		Tag: blume-20030811-windows
2380		Description:
2381
2382		Version number bumped to 110.42.9.  NEW BOOTFILES!!!
2383
2384		http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
2385
2386		This patch restores SML/NJ's ability to run under win32.  There are a
2387		number of changes, including fixes for several bugs that had gone
2388		unnoticed until now:
2389
2390		- uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
2391		trying to run the win32 version from within cygwin.)
2392
2393		- There are a number of simple .bat scripts that substitute for their
2394		corresponding Unix shell-scripts.  (See below.)
2395
2396		- The internals of ml-build have been modified slightly.  The main
2397		difference is that instead of calling ".link-sml" (or link-sml.bat)
2398		using OS.Process.system, the ML process delegates this task back
2399		to the script.  Otherwise problems arise in mixed environments such
2400		as Cygwin where scripts look and work like Unix scripts, but
2401		where OS.Process.system cannot run them.
2402
2403		- In CM, the srcpath pickler used native pathname syntax -- which
2404		is incorrect in the case of cross-compilation.  The new pickle format
2405		is independent of platform-specific naming conventions.
2406
2407		- Path configuration files (such as lib/pathconfig) can now choose
2408		between native and standard syntax.  Placing a line of the form
2409
2410		standard!
2411
2412		into the file causes all subsequent paths to be interpreted using
2413		CM standard pathname syntax (= Unix conventions); a line
2414
2415		native!
2416
2417		switches back to native style.  This was needed so that
2418		path config files can be written portably, see src/system/pathconfig.
2419
2420		- Runtime system:
2421
2422		- win32-filesys.c:  get_file_time and set_file_time now
2423		access modification time, not creation time.
2424
2425		- I/O code made aware of new array representation.
2426
2427		- Bug fixes in X86.prim.masm.
2428
2429		- src/system/makeml made aware of win32. (For use under cygwin
2430		and other Unix-environments for windows.)
2431
2432		- In Basis, fixed off-by-one error in win32-io.sml (function vecF)
2433		which caused BinIO.inputAll to fail consistently.
2434
2435		.bat scripts:
2436
2437		Windows .bat scripts assume that SMLNJ_HOME is defined.
2438
2439		- sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
2440		applications (sml, ml-yacc, ml-lex).
2441		- ml-build.bat: analogous to ml-build.
2442		- config\install.bat:  Analogous to config/install.sh.  This requires
2443		that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2444		(nmake etc. must be on the path, and vcvars32 must have been run.)
2445		Moreover, sources for ml-lex and ml-yacc need to exist under src,
2446		and the bootfile hierarchy must have been unpacked under
2447		sml.boot.x86-win32.
2448		The script is very primitive and does a poor job at error checking.
2449		It only installs the base system, ml-lex, and ml-yacc.  No other
2450		libraries are being installed (i.e., you get only those that
2451		are part of the compiler.)
2452		- link-sml.bat:  analogous to .link-sml, but not currently used
2453
2454		Unrelated bug fixes:
2455
2456		- ml-nlffigen now exports structures ST_* corresponding to incomplete
2457		types.
2458		- Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
2459		otherwise.)
2460
2461		----------------------------------------------------------------------
2462		Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2463		Date: 2003/06/17
2464		Tag: macqueen-20030617-bug895
2465		Description:
2466
2467		Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2468		Tag will be used for new development branch (mcz-branch) for
2469		use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2470
2471		----------------------------------------------------------------------
2472		Name: Matthias Blume (blume (at) tti - c (dot) org)
2473		Date: 2003/05/27 16:55:00 CDT
2474		Tag: blume-20030527-polyeq
2475		Description:
2476
2477		Tried to eliminated most cases of polymorphic equality.
2478
2479		----------------------------------------------------------------------
2480		Name: Matthias Blume (blume (at) tti - c (dot) org)
2481		Date: 2003/05/21 17:45:00 CDT
2482		Tag: blume-20030517-complete
2483		Description:
2484
2485		Two changes:
2486
2487		1. Added a flag for controlling whether non-exhaustive bindings will
2488		be treated as errors (default is false).
2489		2. Cleaned up the *entire* source tree so that CMB.make goes through
2490		without a single non-exhaustive match- or bind warning.
2491
2492		----------------------------------------------------------------------
2493		Name: Matthias Blume (blume (at) tti - c (dot) org)
2494		Date: 2003/05/17 10:20:00 CDT
2495		Tag: blume-20030517-absyn
2496		Description:
2497
2498		1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2499
2500		This mainly affects the quality of error messages.  However, some
2501		of the code is now more straightforward than before. (Treatment of
2502		the above four constructs in translate.sml is much simpler than
2503		the "macro-expansion" that was going on before.  Plus, the mach-
2504		compiler no longer gets invoked just to be able to compile an
2505		if-expression.)
2506
2507		2. The ErrorMsg.Error exception is now caught and absorbed by the
2508		interactive loop.
2509
2510		----------------------------------------------------------------------
2511		Name: Allen Leung
2512		Date: 2003/05/16 13:05:00 CDT
2513		Tag: leunga-20030516-cygwin-runtime
2514		Description:
2515
2516		Ported the runtime system to cygwin, which uses the unix
2517		x86-unix bin files.  Missing/buggy features:
2518
2519		o getnetbyname, getnetbyaddr: these functions seem to be missing in
2520		the Cygwin library.
2521		o Ctrl-C handling may be flaky.
2522		o Windows system calls and Windows I/O are not supported.
2523
2524		A new set of binfiles is located at:
2525
2526		http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2527
2528		This is only needed for bootstrapping the cygwin version of smlnj.
2529		Other x86 versions can use the existing binfiles.
2530
2531		----------------------------------------------------------------------
2532		Name: Matthias Blume
2533		Date: 2003/04/08 15:42:00 CDT
2534		Tag: blume-20030408-listpair
2535		Description:
2536
2537		1. Added a target 'mlrisc' to installer.
2538
2539		2. Added missing elements to structure ListPair.
2540
2541		----------------------------------------------------------------------
2542		Name: Allen Leung
2543		Date: 2003/01/07 10:40:00 EST
2544		Tag: leunga-20030107-int-rem
2545		Description:
2546
2547		Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2548		The arguments to the SUBL instruction were swapped.
2549
2550		----------------------------------------------------------------------
2551		Name: Matthias Blume
2552		Date: 2002/12/12 16:25:00 EST
2553		Tag: blume-20021212-risc-ra
2554		Description:
2555
2556		Fixed a serious bug in the rewrite code for FP spilling/reloading that
2557		sent the RA into an infinite loop when floating point registers get
2558		spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2559		110.37 and 110.38.)
2560		There was another set of potential problems related to the handling of
2561		MLRISC annotations (but those did not yet cause real problems, apparently).
2562
2563		----------------------------------------------------------------------
2564		Name: Matthias Blume
2565		Date: 2002/12/06 22:40:00 EST
2566		Tag: blume-20021206-cm-fileid
2567		Description:
2568
2569		Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2570		This fixes the problem of CM getting confused by files that suddenly
2571		change their identity (e.g., by getting unlinked and recreated by some
2572		text editor such as vi).  There might be a better/cheaper/cleaner way
2573		of doing this, but for now this will have to do.
2574
2575		----------------------------------------------------------------------
2576		Name: Matthias Blume
2577		Date: 2002/10/28 09:50:00 EST
2578		Tag: blume-20021028-typecheck
2579		Description:
2580
2581		Exported structure Typecheck from $smlnj/viscomp/core.cm.
2582
2583		----------------------------------------------------------------------
2584		Name: Matthias Blume
2585		Date: 2002/10/17 09:10:00 EDT
2586		Tag: Release_110_42
2587		Description:
2588
2589		In good old tradition, there has been a slight hiccup so that we have
2590		to patch 110.42 after the fact.   The old release tag has been replaced
2591		(see below).
2592
2593		The change solves a problem with two competing approaches the
2594		configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2595		each other's way.
2596
2597		This change only affects the runtime system code and the installer script.
2598		(No new bootfiles.)
2599
2600		----------------------------------------------------------------------
2601		Name: Matthias Blume
2602		Date: 2002/10/16 12:00:00 EDT
2603		Tag: Release_110_42_removed
2604		Description:
2605
2606		New working release.  New bootfiles.
2607
2608		----------------------------------------------------------------------
2609		Name: Matthias Blume
2610		Date: 2002/10/10 13:10:00 EDT
2611		Tag: blume-20021010-ppc-divs
2612		Description:
2613
2614		The mltree operator DIVS must be implemented with an overflow check on
2615		the PPC because the hardware indicates divide-by-zero using "overflow" as
2616		well.
2617
2618		----------------------------------------------------------------------
2619		Name: Matthias Blume
2620		Date: 2002/07/23 11:45:00
2621		Tag: blume-20020723-smlnj-home
2622		Description:
2623
2624		Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2625		then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2626		CM_PATHCONFIG is also set), the path configuration file is assumed
2627		to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2628		the entire tree to some other place and everything will "just work".
2629
2630		(Companion commands such as ml-build and ml-makedepend also sense this
2631		variable.)
2632
2633		----------------------------------------------------------------------
2634		Name: Matthias Blume
2635		Date: 2002/07/12 21:19:00 EDT
2636		Tag: blume-20020712-liveness
2637		Description:
2638
2639		Exported two useful "step" functions from liveness module (MLRISC).
2640
2641		----------------------------------------------------------------------
2642		Name: Matthias Blume
2643		Date: 2002/07/05 16:00 EDT
2644		Tag: Release_110_41
2645		Description:
2646
2647		New working release.  New bootfiles.
2648
2649		----------------------------------------------------------------------
2650		Name: Matthias Blume
2651		Date: 2002/07/05 10:25:00 EDT
2652		Tag: blume-20020705-btimp
2653		Description:
2654
2655		Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2656		other clients can set up backtracing support.
2657
2658		----------------------------------------------------------------------
2659		Name: Matthias Blume
2660		Date: 2002/06/25 17:23:00 EDT
2661		Tag: blume-20020625-fpmax
2662		Description:
2663
2664		Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2665		case.  (The sense of the isNaN test was reversed -- which made min and
2666		max always return their first argument.)
2667
2668		----------------------------------------------------------------------
2669		Name: Matthias Blume
2670		Date: 2002/06/11
2671		Tag: blume-20020611-unixpath
2672		Description:
2673
2674		Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2675
2676		----------------------------------------------------------------------
2677		Name: Matthias Blume
2678		Date: 2002/06/10 16:35:00 EDT
2679		Tag: blume-20020610-ieeereal
2680		Description:
2681
2682		I back-ported my implementation of IEEEReal.fromString from the
2683		idlbasis-devel branch so that we can test it.
2684
2685		Another small change is that ppDec tries to give more information
2686		than just "<sig>" in the case of functors.  However, this code is
2687		broken in some mysterious way if the functor's body's signature
2688		has not been declared by ascription but gets inferred from the
2689		implementation.  This needs fixing...
2690
2691		----------------------------------------------------------------------
2692		Name: Matthias Blume
2693		Date: 2002/05/31
2694		Tag: blume-20020531-btrace-mode
2695		Description:
2696
2697		Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2698		the wayside when I switched over to using Controls everywhere.)
2699
2700		----------------------------------------------------------------------
2701		Name: Lal George
2702		Date: 2002/05/23 12:21:40 EDT
2703		Tag: george-20020523-visual-labels
2704		Description:
2705
2706		Labels are now displayed in the graphical output to make
2707		the fall-through and target blocks obvious.
2708
2709	----------------------------------------------------------------------	----------------------------------------------------------------------
2710	Name: Matthias Blume	Name: Matthias Blume
2711	Date: 2002/05/22 11:03:00 EDT	Date: 2002/05/22 11:03:00 EDT

---

Colored Diff Long Colored Diff Unidiff Context Diff Side by Side

Legend: Removed from v.1220   changed lines   Added in v.2066

---

root@smlnj-gforge.cs.uchicago.edu	ViewVC Help
Powered by ViewVC 1.0.0