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

Parent Directory | Revision Log | Patch

	revision 1171, Fri Mar 22 20:41:43 2002 UTC	revision 1831, Wed Jul 27 16:15:03 2005 UTC
#	Line 13	Line 13
13	Description:	Description:
14
15	----------------------------------------------------------------------	----------------------------------------------------------------------
16		Name: John Reppy
17		Date: 2005/07/27
18		Tag:
19		Description:
20
21		Fixed ml-yacc to work on files with non-native end-of-line
22		encodings (e.g., Windows text file on a Unix system).
23
24		----------------------------------------------------------------------
25		Name: John Reppy
26		Date: 2005/07/20
27		Tag:
28		Description:
29
30		Added changes from Dominic Evans (oldmanuk (at) gmail (dot) com)
31		to support HPUX 11.
32
33		----------------------------------------------------------------------
34		Name: John Reppy
35		Date: 2005/07/06
36		Tag:
37		Description:
38
39		Changes to the SML/NJ library.  See smlnj-lib/CHANGES for details.
40
41		----------------------------------------------------------------------
42		Name: Matthias Blume (blume (at) tti - c (dot) org)
43		Date: 2005/07/06 00:45:00 CDT
44		Tag: blume-20050706-slice-copy
45		Description:
46
47		Fixed reversed logic for deciding whether to "copy up" or "copy down"
48		in *-array-slice.sml.
49
50		----------------------------------------------------------------------
51		Name: Allen Leung (leunga (at) reservoir (dot) com)
52		Date: 2005/05/31 17:00:00 EST
53		Tag: leunga-20050531-cygwin-fault-2
54		Description:
55
56		A typo in the cygwin code fixed.
57
58		----------------------------------------------------------------------
59		Name: Allen Leung (leunga (at) reservoir (dot) com)
60		Date: 2005/05/31 16:47:00 EST
61		Tag: leunga-20050531-cygwin-fault
62		Description:
63
64		Updated Cygwin's fault/signal handling to match the Windows version.
65		Updated the export list.
66
67		----------------------------------------------------------------------
68		Name: Matthias Blume (blume (at) tti - c (dot) org)
69		Date: 2005/05/18 13:35:00 CDT
70		Tag: Release_110_54
71		Description:
72
73		New working version (110.54).  NEW BOOTFILES!
74
75		----------------------------------------------------------------------
76		Name: Matthias Blume (blume (at) tti - c (dot) org)
77		Date: 2005/05/18 11:58:00 CDT
78		Tag: blume-20050518-installer
79		Description:
80
81		Added support scripts for Mac OS X PackageMaker and modified
82		config/install.sh so that it supports re-dumping a heap image
83		after customization.
84
85		----------------------------------------------------------------------
86		Name: Matthias Blume (blume (at) tti - c (dot) org)
87		Date: 2005/05/18 10:55:00 CDT
88		Tag: blume-20050518-realdiv-noovld
89		Description:
90
91		Un-overloaded / to work around bug in overloading resolution code.
92
93		----------------------------------------------------------------------
94		Name: Matthias Blume (blume (at) tti - c (dot) org)
95		Date: 2005/05/16 23:50:00 CDT
96		Tag: blume-20050516-redump-heap
97		Description:
98
99		Added mechanism for re-creating a heap file for the interactive system
100		after configuration variables have been changed.
101
102		CM.redump_heap : string -> unit
103
104		This is much like SMLofNJ.exportML, but starting from the resulting
105		heap does not return to the caller of CM.redump_heap but
106		restarts the interactive system from scratch.  The original call of
107		CM.redump_heap does not return but ends the interactive session.
108		Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
109
110		Internally, redump_heap winds the dynamic execution context back to
111		the point where the original heap image was created and re-executes
112		the heap image generation code in the boot code.
113
114		----------------------------------------------------------------------
115		Name: Matthias Blume (blume (at) tti - c (dot) org)
116		Date: 2005/05/09 21:55:00 CDT
117		Tag: blume-20050509-word64
118		Description:
119
120		Added a hack to the existing hack known as Word64 to make fromString
121		behave correctly.  I am still not sure whether Word64.scan will work
122		as specified with respect to the interaction of radix and prefix.
123
124		----------------------------------------------------------------------
125		Name: Allen Leung (leunga (at) reservoir (dot) com)
126		Date: 2005/05/04 11:50:00 EST
127		Tag: leunga-20050504-checkgc
128		Description:
129
130		Added a gc protocol checking phase.  This phase is enabled with
131		the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
132
133		----------------------------------------------------------------------
134		Name: Matthias Blume (blume (at) tti - c (dot) org)
135		Date: 2005/05/04 10:45:00 CDT
136		Tag: blume-20050504-intinf
137		Description:
138
139		Fixed a bug in the implementation of div and mod for IntInf.
140		Thanks to Neophytos Michael for reporting the problem.
141
142		----------------------------------------------------------------------
143		Name: Matthias Blume (blume (at) tti - c (dot) org)
144		Date: 2005/05/04 10:35:00 CDT
145		Tag: blume-20050504-join
146		Description:
147
148		Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
149
150		----------------------------------------------------------------------
151		Name: Matthias Blume (blume (at) tti - c (dot) org)
152		Date: 2005/02/28 23:40:00 CST
153		Tag: blume-20050228-mVar
154		Description:
155
156		Fixed serious bug (brown paper bag variety) in new implementation of
157		structure Atom in CML.  (I had accidentally used a mailbox instead of
158		an mvar, leaving the door open for races.)
159
160		----------------------------------------------------------------------
161		Name: Matthias Blume (blume (at) tti - c (dot) org)
162		Date: 2005/02/25 15:00:00 CST
163		Tag: Release_110_53
164		Description:
165
166		New working version (110.53).  NEW BOOTFILES!
167
168		----------------------------------------------------------------------
169		Name: Matthias Blume (blume (at) tti - c (dot) org)
170		Date: 2005/02/25 14:50:00 CST
171		Tag: blume-20050225-susp
172		Description:
173
174		Brought back SMLofNJ.Susp.  The underlying suspension type is the one
175		implemented in Core -- which means that it is the same as the one used
176		by the lazy extension.
177
178		----------------------------------------------------------------------
179		Name: Matthias Blume (blume (at) tti - c (dot) org)
180		Date: 2005/02/24 16:50:00 CST
181		Tag: blume-20050224-cml-atom
182		Description:
183
184		Simpler and at the same time more general implementation of structure
185		Atom in CML.
186
187		----------------------------------------------------------------------
188		Name: Matthias Blume (blume (at) tti - c (dot) org)
189		Date: 2005/02/15 17:35:00 CST
190		Tag: blume-20050215-tools
191		Description:
192
193		Created new "tools" directory under "src" and moved "TraceDebugProf"
194		there.
195
196		----------------------------------------------------------------------
197		Name: Matthias Blume (blume (at) tti - c (dot) org)
198		Date: 2005/02/10 17:55:00 CST
199		Tag: blume-20050210-longlong
200		Description:
201
202		Implemented "long long" arguments and results for NLFFI.  (Only the
203		PPC/MacOS implementation is complete, the other backends still need to
204		be updated.)
205
206		----------------------------------------------------------------------
207		Name: Matthias Blume (blume (at) tti - c (dot) org)
208		Date: 2005/01/24 17:40:00 CST
209		Tag: blume-20050124-mlyacc
210		Description:
211
212		Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
213		problem with certain "as" patterns which previously got rendered
214		using incorrect syntax.
215
216		----------------------------------------------------------------------
217		Name: Matthias Blume (blume (at) tti - c (dot) org)
218		Date: 2005/01/18 12:00:00 CST
219		Tag: blume-20050118-profile
220		Description:
221
222		Made time profiling code (interrupt handler) in runtime system aware
223		of new array representation.
224
225		----------------------------------------------------------------------
226		Name: Matthias Blume (blume (at) tti - c (dot) org)
227		Date: 2005/01/14 18:00:00 CST
228		Tag: blume-20050114-heap2exec
229		Description:
230
231		Implemented new (but still experimental) heap2exec facility.  This is
232		tested under Mac OS X and should work under Linux (will test shortly).
233		It will probably also work on the Sparc (will test some time later).
234
235		- removed old "HACKED_STANDALONE" hack from runtime
236
237		To be able to test this, uncomment the request for "heap2asm" in
238		config/targets prior to installation.  (Notice that this is different
239		from "heap2exec" mentioned below.  Not a typo.)
240
241		To perform an actual test, run the command
242
243		$ bin/heap2exec heapfile execfile
244
245		(You can put heap2exec on your shell's path.)
246
247		For example, run
248
249		$ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
250
251		This will create a standalone executable called "mly" which you can
252		then invoke directly as a command.
253
254		----------------------------------------------------------------------
255		Name: Matthias Blume (blume (at) tti - c (dot) org)
256		Date: 2005/01/07 11:44:00 CST
257		Tag: blume-20050107-mlstring
258		Description:
259
260		fixed off-by-one error in ML_STRING macro (globals.c)
261
262		----------------------------------------------------------------------
263		Name: Matthias Blume (blume (at) tti - c (dot) org)
264		Date: 2004/12/23 18:00:00 CST
265		Tag: blume-20041223-santa
266		Description:
267
268		Made ml-build script "smarter" (but only very little).
269
270		----------------------------------------------------------------------
271		Name: Matthias Blume (blume (at) tti - c (dot) org)
272		Date: 2004/12/21 15:05:00 CST
273		Tag: blume-20041221-longlong
274		Description:
275
276		* Implemented access to signed and unsigned long long data in NLFFI.
277		(The parameter-passing part of the picture has not complete. But
278		data structure access seems to work.)
279
280		* Fixed CM's incorrect assumption that the PPC is little-endian.
281		(On the Mac, it is big-endian.  And that's currently our only
282		PPC platform.)
283
284		----------------------------------------------------------------------
285		Name: Matthias Blume (blume (at) tti - c (dot) org)
286		Date: 2004/12/21 12:50:00 CST
287		Tag: blume-20041221-memory
288		Description:
289
290		Some cleanup in the $c/memory.cm library: separated some concerns by
291		moving allocation code and memory access code each into their own
292		files.
293
294		----------------------------------------------------------------------
295		Name: Allen Leung (leunga (at) reservoir (dot) com)
296		Date: 2004/12/17 16:12:00 EST
297		Tag: leunga-20041217-cygwin-smlnj-home
298		Description:
299
300		The Unix I/O library of SML/NJ on cygwin does not understand
301		Windows style pathname, so problems arise when SMLNJ_HOME is set
302		to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
303		to a POSIX pathname on cygwin.
304
305		----------------------------------------------------------------------
306		Name: Matthias Blume (blume (at) tti - c (dot) org)
307		Date: 2004/12/16 13:35:00 CST
308		Tag: Release_110_52
309		Description:
310
311		Last-minute changes incorporated into 110.52.  Release tag moved.
312
313		The changes:
314		- HashString.hashString' -> HashString.hashSubstring
315		- bug fix in UnivariateStats
316
317		----------------------------------------------------------------------
318		Name: Matthias Blume (blume (at) tti - c (dot) org)
319		Date: 2004/12/15 23:40:00 CST
320		Tag: blume-20041215-hashSubstring
321		Description:
322
323		- HashString.hashString' -> HashString.hashSubstring
324		- corresponding changes in atom.sml
325		- "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
326		added some comments
327
328		----------------------------------------------------------------------
329		Name: Matthias Blume (blume (at) tti - c (dot) org)
330		Date: 2004/12/15 15:30:00 CST
331		Tag: (Release_110_52)
332		Description:
333
334		New working version (110.52).  NEW BOOTFILES!
335
336		----------------------------------------------------------------------
337		Name: Matthias Blume (blume (at) tti - c (dot) org)
338		Date: 2004/12/15 12:45:00 CST
339		Tag: blume-20041215-spaces
340		Description:
341
342		More on the space problem (this time for Win32).
343
344		----------------------------------------------------------------------
345		Name: Matthias Blume (blume (at) tti - c (dot) org)
346		Date: 2004/12/14 17:30:00 CST
347		Tag: blume-20041214-spaces
348		Description:
349
350		Hacked some of the scripts (in particular: the installer) to cope with
351		spaces in filenames a bit better.  But beware: the current "solution"
352		is likely still full of bugs and inherently incomplete.  (We need to
353		do away with those shell scripts for a comprehensive solution.)
354
355		----------------------------------------------------------------------
356		Name: Matthias Blume (blume (at) tti - c (dot) org)
357		Date: 2004/12/13 14:45:00 CST
358		Tag: blume-20041213-ml-makedepend
359		Description:
360
361		Fixed bug in code for ml-makedepend.
362
363		----------------------------------------------------------------------
364		Name: Matthias Blume (blume (at) tti - c (dot) org)
365		Date: 2004/12/09 16:30:00 CST
366		Tag: blume-20041209-statistics
367		Description:
368
369		Added two simple but potentially useful statistics modules to SML/NJ Library.
370		(See CHANGES file there.)
371
372		----------------------------------------------------------------------
373		Name: Matthias Blume (blume (at) tti - c (dot) org)
374		Date: 2004/12/01 16:50:00 CST
375		Tag: blume-20041201-atom
376		Description:
377
378		smlnj-lib:
379
380		Added function HashString.hashString' for substrings.
381		Hand-inlined CharVector.foldl into HashString (for speed).
382		Modified implementation of structure Atom to avoid extracting
383		strings from substrings unless necessary.
384		(Also see CHANGES file for smlnj-lib.)
385
386		----------------------------------------------------------------------
387		Name: Matthias Blume (blume (at) tti - c (dot) org)
388		Date: 2004/11/24 22:15:00 CST
389		Tag: blume-20041124-cml
390		Description:
391
392		Made sure CML compiles when Position = Int64.
393
394		----------------------------------------------------------------------
395		Name: Matthias Blume (blume (at) tti - c (dot) org)
396		Date: 2004/11/24 14:30:00 CST
397		Tag: blume-20041124-position
398		Description:
399
400		The compiler can now be compiled in a mode that makes structure Position
401		equal to Int64.  The default, however, is unchanged (Position = Int31)
402		for the time being.
403
404		To enable 64-bit positions, use the following procedure:
405
406		1. Start sml
407		2. Autoload $smlnj/cmb.cm (if not already autoloaded)
408		3. Type
409		#set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
410		4. Run CMB.make() as usual.
411
412		This is barely tested.  The only test so far was a little SML program
413		counting the number of characters in an 8-gigabyte file by
414		reading it character-by-character.  That test was successful.
415
416		In support of 64-bit positions, a number of new functions have been
417		added to the runtime system.
418
419		----------------------------------------------------------------------
420		Name: Matthias Blume (blume (at) tti - c (dot) org)
421		Date: 2004/11/23 14:45:00 CST
422		Tag: blume-20041123-useFile
423		Description:
424
425		Fixed a problem with unhelpful error messages related to problems with
426		.cm- or .sml files that appear as part of the sml command line.
427
428		----------------------------------------------------------------------
429		Name: Matthias Blume (blume (at) tti - c (dot) org)
430		Date: 2004/11/18 15:40:00 CST
431		Tag: Release_110_51
432		Description:
433
434		New working version (110.51).  NEW BOOTFILES!
435
436		----------------------------------------------------------------------
437		Name: Matthias Blume (blume (at) tti - c (dot) org)
438		Date: 2004/11/18 15:35:00 CST
439		Tag: <none>
440		Description:
441
442		Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
443		Joemann).
444
445		----------------------------------------------------------------------
446		Name: Allen Leung (leunga (at) reservoir (dot) com)
447		Date: 2004/11/17 16:05:21 EST 2004
448		Tag: leunga-20041117-mlrisc-live-kill
449		Description:
450
451		Added support for MLTree constructs LIVE and KILL
452		to all the architectures.
453
454		----------------------------------------------------------------------
455		Name: Matthias Blume (blume (at) tti - c (dot) org)
456		Date: 2004/11/13 00:20:00 CST
457		Tag: blume-20041113-versiontool
458		Description:
459
460		- Stripped down the versiontool: It now only handles the version number.
461		The date string is generated at bootstrap time (during makeml).
462
463		- In a previous commit, fixed a minor issue with how polyequal is being
464		translated.  In particular, the code now "looks through" abstractions.
465		This results in slightly fewer polyEqual warnings and hopefully slightly
466		more efficient code.  Important examples for where this matters are
467		the new int64 and word64 types.
468
469		----------------------------------------------------------------------
470		Name: Matthias Blume (blume (at) tti - c (dot) org)
471		Date: 2004/11/12 00:30:00 CST
472		Tag: blume-20041112-int64
473		Description:
474
475		Structure Int64 fully hooked in.  (The implementation is not very
476		efficient, though.)
477
478		----------------------------------------------------------------------
479		Name: Matthias Blume (blume (at) tti - c (dot) org)
480		Date: 2004/11/11 17:30:00 CST
481		Tag: blume-20041111-more64
482		Description:
483
484		All the pieces of Word64 are now there, with the exception of the
485		conversions from and to LargeWord. (Eventually these need to be identities,
486		but for the time being they don't even make sense because LargeWord is
487		32-bit wide.)
488
489		Also started to add similar support for Int64, but major pieces of that
490		are still missing.
491
492		----------------------------------------------------------------------
493		Name: Matthias Blume (blume (at) tti - c (dot) org)
494		Date: 2004/11/11 00:15:00 CST
495		Tag: blume-20041111-word64
496		Description:
497
498		Structure Word64 is now (almost) complete, word literals and patterns
499		seem to work.  There are a few odd pieces missing.  In particular,
500		I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
501		at the moment.
502
503		Making Word64 official would mean that LargeWord becomes Word64.  But
504		this requires extreme care because most word-word conversions have to
505		go through LargeWord, so making a mistake means loss of efficiency or
506		worse.  Eventually there will be a solution similar to (but actually
507		simpler than) what I did with IntInf.
508
509		----------------------------------------------------------------------
510		Name: Matthias Blume (blume (at) tti - c (dot) org)
511		Date: 2004/11/10 18:12:00 CST
512		Tag: blume-20041110-64bit
513		Description:
514
515		More 64-bit hacking (but still not even half-way there yet).
516		Also, some assorted improvements to the handling of 8-bit words.
517
518		----------------------------------------------------------------------
519		Name: Matthias Blume (blume (at) tti - c (dot) org)
520		Date: 2004/11/09 17:50:00 CST
521		Tag: <none>
522		Description:
523
524		Started adding some infrastructure for supporting 64-bit int- and
525		word-types.  (Still in its very early stages.)
526
527		----------------------------------------------------------------------
528		Name: Matthias Blume (blume (at) tti - c (dot) org)
529		Date: 2004/10/28 10:45:00 CDT
530		Tag: Release_110_50
531		Description:
532
533		New working version (110.50).  NEW BOOTFILES!
534
535		=====================
536
537		Also:
538
539		- Changend config/srcarchiveurl from a file just
540		containing the URL string into a file containing
541		shell script code.  The code has access to the $VERSION variable.
542		- Made corresponding changes to config/install.sh and config/unpack.
543		- Default contents of config/srcarchiveurl uses $VERSION and
544		normally does not have to be edited to reflect a version change.
545
546		(As a result, a version change can be done by just editing
547		config/version, the rest is now automatic.)
548
549		----------------------------------------------------------------------
550		Name: Matthias Blume (blume (at) tti - c (dot) org)
551		Date: 2004/10/27 17:50:00 CDT
552		Tag: blume-20041027-btrace-msg
553		Description:
554
555		BackTrace.monitor now also reports the source of the exception that
556		triggered the trace.
557
558		----------------------------------------------------------------------
559		Name: Matthias Blume (blume (at) tti - c (dot) org)
560		Date: 2004/10/27 17:20:00 CDT
561		Tag: blume-20041027-x86-c-calls
562		Description:
563
564		This is the HISTORY entry for two earlier commits, both concerning
565		the x86 c-calls code in MLRISC:
566
567		- added a missing LOAD in the code that deals with struct arguments
568		- made sure the caller does not add the wrong number of bytes to the
569		stack pointer after a call of a function returning a struct
570		(the callee already pops the implicit argument which points to
571		the space reserved for the result)
572
573		----------------------------------------------------------------------
574		Name: Allen Leung (leunga (at) reservoir (dot) com)
575		Date: 2004/10/24 14:00:00 EST
576		Tag: leunga-20041024-x86-gas-fucomip
577		Description:
578
579		John discovered a bug in the syntax of fucomip.
580		The opcodes FU?COMIP? have been changed to
581
582		fu?comip? %st(i), %st
583
584		----------------------------------------------------------------------
585		Name: Matthias Blume (blume (at) tti - c (dot) org)
586		Date: 2004/10/20 15:06:00 CDT
587		Tag: blume-20041020-standalone-backtrace
588		Description:
589
590		Added a mechanism for getting back-trace information from standalone
591		programs.  Here is how it works:
592
593		1. The part of the program from which you want to get backtrace
594		information (usually the whole program) should be wrapped with
595		BackTrace.monitor.  This is a (unit->'a)->'a function, and your
596		main program could be modified from something like
597
598		fun main (pgm, args) = ...
599
600		to
601
602		fun main (pgm, args) = BackTrace.monitor (fn () => ...)
603
604		2. To be able to access BackTrace.monitor, you have to add
605		library $smlnj-tdp/plugins.cm to the .cm file that contains your
606		main function.
607
608		3. Remove all compiled code (i.e., all the .cm/ subdirectories that
609		CM might have created in the past for your project).
610
611		4. Build the system using this command line:
612
613		ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
614		myprog.cm MyProg.main myprog
615
616		instead of the usual
617
618		ml-build myprog.cm MyProg.main myprog
619
620		I changed a library name:
621
622		$/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
623
624		New libraries:
625
626		$smlnj-tdp/back-trace.cm
627		-- when loaded causes the back-trace plugin to be installed
628		$smlnj-tdp/coverage.cm
629		-- when loaded causes the coverage plugin to be installed
630
631		----------------------------------------------------------------------
632		Name: Matthias Blume (blume (at) tti - c (dot) org)
633		Date: 2004/10/18 16:45:00 CDT
634		Tag: blume-20041018-groupowner
635		Description:
636
637		Added an "obsolete" warning for the "group owner" syntax to CM's parser.
638
639		Eliminated group owner specs from .cm files throughout the source tree.
640
641		----------------------------------------------------------------------
642		Name: Matthias Blume (blume (at) tti - c (dot) org)
643		Date: 2004/10/15 15:45:00 CDT
644		Tag: blume-20041015-coverage
645		Description:
646
647		* Test coverage tool added!
648		* Further reorganization of tracing-, debugging-, and profiling support:
649
650		- moved original BTImp -- now called BackTrace -- into a separate
651		library called $/trace-debug-profile.cm
652		- eliminated all mentions of BTrace from SMLofNJ.Internals
653		- only the instrumentation mechanism is now left in the compiler proper
654		- BackTrace module is a plugin which is NOT plugged in by default
655		- Coverage module is another such plugin
656
657		To get the benefits of any of these plugin modules, the code in
658		question must be compiled with tdp instrumentation turned on.  This
659		can be done by setting SMLofNJ.Internals.TDP.mode to true.
660		(The ref cell is also controlled via the -Ctdp.instrument=... switch.)
661
662		Plugins are selected at link time.  (Pre-compiled instrumented code
663		can be re-loaded with different plugins in effect.)  When an
664		instrumented module is linked, whatever plugins are at that time
665		enabled will come into effect for that module.
666
667		To enable the back-trace plugin, load library $/trace-debug-profile.cm
668		and invoke BackTrace.install() (e.g., from the interactive prompt).
669		To enable the coverage plugin, load the same library and invoke
670		Coverage.install().
671
672		Back-traces are generated automatically on uncaught exceptions and
673		when the code in question explicitly invokes BackTrace.trigger().
674
675		Coverage (and execution frequency-) information must be queried
676		explicitly by calling Coverage.not_covered and Coverage.hot_spots.
677
678		----------------------------------------------------------------------
679		Name: Matthias Blume (blume (at) tti - c (dot) org)
680		Date: 2004/10/14 17:40:00 CDT
681		Tag: blume-20041014-tdp-core
682		Description:
683
684		Snapshot of a significant overhaul of how the trace/debug/profile support
685		is hooked into the system (specifically: Core and SMLofNJ.Internals).
686
687		----------------------------------------------------------------------
688		Name: Matthias Blume (blume (at) tti - c (dot) org)
689		Date: 2004/10/13 16:34:00 CDT
690		Tag: blume-20041013-tdp
691		Description:
692
693		Some rationalization of names:
694
695		structure BTrace -> structure TDPInstrument
696		etc.
697
698		This is is preparation of using the original back-trace
699		instrumentation for other purposes.  "TDP" stands for
700		Trace/Debug/Profile.
701
702		The control flag controlling whether instrumentation is on or off is now
703		registered under a different name, so instead of running sml as
704
705		sml -Cinstrument.btrace-mode=true
706
707		one has to say
708
709		sml -Ctdp.instrument=true
710
711		----------------------------------------------------------------------
712		Name: Matthias Blume (blume (at) tti - c (dot) org)
713		Date: 2004/10/11 16:37:00 CDT
714		Tag: blume-20041011-regions
715		Description:
716
717		Made some minor modifications to elabcore.sml to have source regions
718		be propagated more tightly -- resulting in better (i.e., smaller)
719		regions being reported in error- and debug messages.
720
721		----------------------------------------------------------------------
722		Name: Matthias Blume (blume (at) tti - c (dot) org)
723		Date: 2004/10/08 22:50:00 CDT
724		Tag: blume-20041008-cmkw
725		Description:
726
727		Fixed handling of keywords in .cm files: After seeing "is" the lexer
728		treats subsequent occurrences of "group", "library", "source", "is",
729		"*", and "-" as ordinary identifiers rather than keywords.
730
731		Most seriously, this fixes a problem with CM's "shell" tool.  The tool
732		is supposed to accept a tool argument called "source", but this did
733		not work because of the clash with the keyword.
734
735		----------------------------------------------------------------------
736		Name: Matthias Blume (blume (at) tti - c (dot) org)
737		Date: 2004/10/07 16:00:00 CDT
738		Tag: blume-20041007-cleanup
739		Description:
740
741		Assorted cleanup work:
742
743		- got rid of intstrmap in favor of using the library's
744		hash table implementation
745		- threw out most of the pathnames stuff, as it was not used anyway
746		- simplified tokentable implementation
747		- fixed some minor spelling errors
748
749		----------------------------------------------------------------------
750		Name: Matthias Blume (blume (at) tti - c (dot) org)
751		Date: 2004/10/06 15:15:15 CDT
752		Tag: blume-20041006-handler
753		Description:
754
755		Cleaned up the absyn to reflect the invariant that HANDLE always
756		carries a FNexp as part of the type definition.  This eliminates some
757		superfluous sanity checks at runtime down the road.
758
759		Some minor cleanup of the btrace code.
760
761		----------------------------------------------------------------------
762		Name: Matthias Blume (blume (at) tti - c (dot) org)
763		Date: 2004/10/01 10:20:30 CDT
764		Tag: blume-20041001-slave
765		Description:
766
767		Added hack to make slave mode work in the presence of the version
768		tool.  (Still, since the master does two passes over the code for
769		CMB.make, the release number gets bumped twice when slaves are
770		attached.  I don't know if this is worth fixing...)
771
772		----------------------------------------------------------------------
773		Name: Matthias Blume (blume (at) tti - c (dot) org)
774		Date: 2004/09/30 10:55:00 CDT
775		Tag: blume-20040930-version
776		Description:
777
778		* Moved the "version" magic into its own little library under
779		src/system/smlnj/internal.  This avoids expensive reconstruction of
780		a stable src/compiler/core.cm.
781
782		* At the same time, structure CompilerVersion is now known as
783		structure SMLNJVersion.
784
785		* Arranged for the version tool to NOT kick in when rebuilding the system
786		(makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
787		Also, loading the versiontool does not work when rebuilding the system
788		because CM is not properly initialized at that time.
789
790		----------------------------------------------------------------------
791		Name: Matthias Blume (blume (at) tti - c (dot) org)
792		Date: 2004/09/29 14:00:00 CDT
793		Tag: blume-20040929-autoversion
794		Description:
795
796		Implemented some CM magic to have
797		file src/compiler/TopLevel/main/version.sml
798		generated automagically.
799		The version is taken from two files: config/version and config/release.
800		The first is expected to contain a two-part version number such as 110.49.
801		The second should contain a single number, but it may be missing.
802
803		If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
804		time the version tool is loaded (which is the first time you say CMB.make),
805		then the tool will increment the value stored in config/release every
806		time CMB.make is invoked.
807
808		The binfile format is now insensitive to anything beyond the first
809		two components of a version number, so bumping the release does not render
810		binfiles incompatible.  Auto-bumping can be used to keep track of versions
811		during development without invalidating existing binfiles.
812
813		In any case, every CMB.make updates the date information in version.sml.
814		(This is the date that is printed in the banner.)
815
816		----------------------------------------------------------------------
817		Name: Matthias Blume (blume (at) tti - c (dot) org)
818		Date: 2004/09/28 10:53:00 CDT
819		Tag: blume-20040928-controls
820		Description:
821
822		Some cleanup of the controls code.
823
824		----------------------------------------------------------------------
825		Name: Matthias Blume (blume (at) tti - c (dot) org)
826		Date: 2004/09/27 22:08:00 CDT
827		Tag: blume-20040927-controls
828		Description:
829
830		Added two pieces of functionality to the Controls interface:
831
832		1.   val save'restore: 'a control -> unit -> unit
833
834		grabs the current value of the control in stage 1 and restores it
835		in stage 2.
836
837		2.   val set' : 'a control * 'a -> unit -> unit
838
839		stores the given value into the control in stage 2 (i.e., delayed)
840		but does all error checking in stage 1.
841		(This is for string controls that need to do parse their argument --
842		something that might fail.  In some cases, notably in CM, one
843		already knows the intended argument but wants to delay the actual
844		assignment until a time when error recovery would be more difficult.)
845
846		Changed the handling of controls in tool arguments to classes "sml" and
847		"lazysml":
848		- use Controls.save'restore as a more robust way of restoring the
849		old value (in particular: without having to re-parse the string)
850		- use controls to handle the "overload" keyword in the init group
851		(I believe this change actually fixes a long-standing obscure bug.)
852
853		----------------------------------------------------------------------
854		Name: Matthias Blume (blume (at) tti - c (dot) org)
855		Date: 2004/09/27 17:00:00 CDT
856		Tag: blume-20040927-lazysml
857		Description:
858
859		Added a new tool class called "lazysml" to CM's tool chest.  The only
860		difference to "sml" is that compilation is done with Control.lazysml
861		set to true.  A source of class "lazysml" is automatically recognized
862		by a file name suffix of ".lml".
863
864		In addition to the above feature, the original class "sml" now also
865		supports a tool argument "lazy" which has the same effect.  As a
866		result, the following three lines are equivalent:
867
868		foo.sml : lazysml
869		foo.sml : sml (lazy)
870		foo.sml (lazy)
871
872		The setting goes into effect both during parsing and during
873		compilation.  The original setting is restored right after parsing and
874		after compilation, respectively.
875
876		In addition to all the above, there is also a general mechanism to set
877		ANY of the "controls" that are available at the command line via
878		"-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
879		apply as well. (In fact, "lazy" is implemented as a special case of
880		the general mechanism.)
881
882		The .cm file syntax uses a new keyword tool argument called "with".
883		There are several ways of indicating the desired settings:
884
885		foo.sml (with:parser.quotations=true)
886		foo.sml (with:(name:parser.quotations value:true))
887		foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
888		foo.sml (with:(name1=value1 name2=value2 ...))
889		foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
890
891		etc.
892
893		Another possible abbreviation is to leave out the =v or value:v part
894		if the name refers to a boolean control (in which case the value is
895		taken to be true).  Thus, one could get lazy sml also by saying:
896
897		foo.sml (with:parser.lazy-keyword=true)
898		foo.sml (with:parser.lazy-keyword)
899		foo.sml (with:(name:parser.lazy-keyword value:true))
900		foo.sml (with:(name:parser.lazy-keyword))
901
902		----------------------------------------------------------------------
903		Name: Matthias Blume (blume (at) tti - c (dot) org)
904		Date: 2004/09/24 16:22:00 CDT
905		Tag: blume-20040924-ppc-long-branch
906		Description:
907
908		Turned message about "emiting long form of branch" off by default.
909		Added a control flag to turn it back on when desired.
910
911		----------------------------------------------------------------------
912		Name: Matthias Blume (blume (at) tti - c (dot) org)
913		Date: 2004/09/24 16:05:00 CDT
914		Tag: blume-20040924-rounding
915		Description:
916
917		Applied patch for setting rounding modes under Mac OS X.  Thanks to
918		Melissa O'Neill for providing the code!
919
920		----------------------------------------------------------------------
921		Name: Matthias Blume (blume (at) tti - c (dot) org)
922		Date: 2004/09/23 17:30:00 CDT
923		Tag: blume-20040923-envvars
924		Description:
925
926		1. Changed definition of type ControlRegistry.registry_tree to
927		include control_info (i.e., the name of the controlling environment
928		variable).
929
930		2. Added command-line flags -e and -E to print the names of environment
931		variables that can be used to control internal settings.
932		(This uses the new API mentioned in 1.)
933
934		----------------------------------------------------------------------
935		Name: Matthias Blume (blume (at) tti - c (dot) org)
936		Date: 2004/09/13 16:50:00 CDT
937		Tag: Release_110_49
938		Description:
939
940		New working version (110.49).  NEW BOOTFILES!
941
942		----------------------------------------------------------------------
943		Name: Matthias Blume
944		Date: 2004/09/13 16:20:00 CDT
945		Tag: blume-20040913-config-mlrisc
946		Description:
947
948		Put target "mlrisc" back into the default list.
949		(There is no harm in having it, and some users have expressed their
950		wish to have "mlrisc" included by default.)
951
952		----------------------------------------------------------------------
953		Name: John Reppy
954		Date: 2004/09/13
955		Tag: jhr-20040913-signals
956		Description:
957
958		Fixed the signal masking code to properly nest mask/unmask operations
959		on a per-signal basis.
960
961		----------------------------------------------------------------------
962		Name: Matthias Blume (blume (at) tti - c (dot) org)
963		Date: 2004/09/08 13:20:00 CDT
964		Tag: blume-20040908-heap-magic
965		Description:
966
967		Bumped the heap macig number to 0x09082004 to account for the changed
968		layout of the ML frame under MacOS X.
969
970		----------------------------------------------------------------------
971		Name: Allen Leung (leunga (at) reservoir (dot) com)
972		Date: 2004/09/03 11:26:00 EST
973		Tag: leunga-20040903-cygwin-install
974		Description:
975
976		Added a patch to _arch-n-opsys to enable the Cygwin runtime.
977		The Cygwin runtime is turned on by setting the environment
978		variable SMLNJ_CYGWIN_RUNTIME to 1.
979
980		----------------------------------------------------------------------
981		Name: Matthias Blume (blume (at) tti - c (dot) org)
982		Date: 2004/08/31 17:15:00 CDT
983		Tag: blume-20040831-core
984		Description:
985
986		Added some exports to src/compiler/core.cm upon request by J. Joemann.
987
988		----------------------------------------------------------------------
989		Name: Matthias Blume (blume (at) tti - c (dot) org)
990		Date: 2004/08/30 17:55:00 CDT
991		Tag: blume-20040830-installer
992		Description:
993
994		Upon request by Johannes Joemann:
995
996		- improved ML code of installer to fall back to coping when renaming
997		fails (i.e., when source and target are on different file systems);
998		the code compiles but has yet to be tested in anger
999		- removed mlrisc from list of default targets (config/targets)
1000
1001		----------------------------------------------------------------------
1002		Name: Matthias Blume (blume (at) tti - c (dot) org)
1003		Date: 2004/08/27 17:20:00 CDT
1004		Tag: blume-20040827-ptreql
1005		Description:
1006
1007		Added ptreql primop to structure InlineT (upon request from Larry
1008		Paulson).
1009
1010		----------------------------------------------------------------------
1011		Name: Allen Leung (leunga (at) reservoir (dot) com)
1012		Date: 2004/08/15 21:21:00 EST
1013		Tag: leunga-110_48-udgraph
1014		Description:
1015
1016		Another bug fix from Carl Hauser:
1017
1018		diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
1019		> 48c48
1020		> <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
1021		> e::L)
1022		> ---
1023		> >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
1024		> Without this, any deletion of an edge in an undirected graph does severe
1025		> violence to the graph.
1026
1027		----------------------------------------------------------------------
1028		Name: Allen Leung (leunga (at) reservoir (dot) com)
1029		Date: 2004/08/10 23:35:00 EST
1030		Tag: leunga-110_48-ppc
1031		Description:
1032
1033		The IBM/MacOS syntax switch on PPC was incorrectly swapped.
1034
1035		----------------------------------------------------------------------
1036		Name: Matthias Blume (blume (at) tti - c (dot) org)
1037		Date: 2004/08/10 12:00:00 CDT
1038		Tag: Release_110_48
1039		Description:
1040
1041		New working version (110.48).  NEW BOOTFILES!
1042
1043		----------------------------------------------------------------------
1044		Name: Allen Leung (leunga (at) reservoir (dot) com)
1045		Date: 2004/08/09 12:21:00 EST
1046		Tag: leunga-110_47-dijsktra
1047		Description:
1048
1049		Bug fix from Carl Hauser:
1050
1051		single_source_shortest_paths in dijkstra.sml was observed to get wrong
1052		answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1053
1054		The problem is that following the expression A.update(dist,s,Num.zero)
1055		it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1056
1057		----------------------------------------------------------------------
1058		Name: Matthias Blume (blume (at) tti - c (dot) org)
1059		Date: 2004/08/06 18:10:00 CDT
1060		Tag: blume-20040806-cmdline
1061		Description:
1062
1063		Fiddled with handling of command-line options:
1064
1065		* sml now quits after processing the command line
1066		if -H, -S, -h<n>, or -s<n> appears as the last
1067		command-line argument
1068		* a new option -q terminates the session when encountered on
1069		the command line; subsequent arguments will be ignored
1070		* bug fixes: short (erroneous) arguments are no longer ignored
1071		completely
1072
1073		----------------------------------------------------------------------
1074		Name: Allen Leung (leunga (at) reservoir (dot) com)
1075		Date: 2004/08/04 18:17:00 EST
1076		Tag: leunga-110_47-ppc-ibm-asm
1077		Description:
1078
1079		- Added minimal IBM assembly syntax support for PowerPC.
1080
1081		- Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1082		been altered in the runtime.  We need an automatic way to keep the file
1083		in sync.
1084
1085		----------------------------------------------------------------------
1086		Name: Matthias Blume (blume (at) tti - c (dot) org)
1087		Date: 2004/08/04 14:00:00 CDT
1088		Tag: Release_110_47
1089		Description:
1090
1091		New working version (110.47).  NEW BOOTFILES!
1092
1093		----------------------------------------------------------------------
1094		Name: Matthias Blume (blume (at) tti - c (dot) org)
1095		Date: 2004/08/03 14:25:00 CDT
1096		Tag: blume-20040803-callingconv
1097		Description:
1098
1099		Added low-level support for choosing C calling conventions by
1100		twiddling the type of rawccall.  (See
1101		src/compiler/Semant/types/cproto.sml for details.)
1102
1103		----------------------------------------------------------------------
1104		Name: Matthias Blume (blume (at) tti - c (dot) org)
1105		Date: 2004/08/02 15:55:00 CDT
1106		Tag: blume-20040802-backout
1107		Description:
1108
1109		Backed out of change to win32-filesys.c.  The earlier patch to
1110		get_file_time caused CM to produce files with the wrong time stamp.
1111
1112		----------------------------------------------------------------------
1113		Name: Matthias Blume (blume (at) tti - c (dot) org)
1114		Date: 2004/08/02 14:45:00 CDT
1115		Tag: blume-20040802-nlffi-win32
1116		Description:
1117
1118		Added NLFFI support for Win32, adapted from a patch provided by David
1119		Hansel.  This is currently completely untested.  Also, the issue
1120		concerning stdcall vs. ccall is still unresolved.
1121
1122		----------------------------------------------------------------------
1123		Name: Matthias Blume (blume (at) tti - c (dot) org)
1124		Date: 2004/07/30 17:55:00 CDT
1125		Tag: blume-20040730-various
1126		Description:
1127
1128		Gearing up towards 110.47...
1129
1130		- various minor bugfixes to ml-nlffigen
1131		- a beginning of a manual for nlffi
1132
1133		- eliminated 'export name=value' in config/install.sh as this does
1134		not work with certain versions of /bin/sh
1135		(Thanks to David King at Motorola for catching this.)
1136
1137		- several bugfixes provided or suggested by David Hansel at Reactive Systems:
1138		- added a test for tm==NULL to gmtime.c and localtime.c
1139		- applied patch for incorrect GetFileTime under win32
1140		- toSeconds -> toMilliseconds in Win32/win32-process.sml
1141
1142		----------------------------------------------------------------------
1143		Name: Matthias Blume (blume (at) tti - c (dot) org)
1144		Date: 2004/07/21 18:20:00 CDT
1145		Tag: blume-20040721-nlffigen
1146		Description:
1147
1148		- Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1149		for a typedef to an incomplete type, but leave out the "typ" member.
1150		(This is just for consistency.)
1151		- Started to produce what is supposed to become better (i.e., comprehensive)
1152		documentation of what ml-nlffigen does and produces.
1153
1154		----------------------------------------------------------------------
1155		Name: Matthias Blume (blume (at) tti - c (dot) org)
1156		Date: 2004/07/14 16:25:00 CDT
1157		Tag: blume-20040714-union
1158		Description:
1159
1160		Added C_UNION to c-calls/c-types.sml and updated the machinery
1161		(ml-nlffigen, cproto.sml) that conveys C function interface
1162		information to the code generator.
1163
1164		However, the actual architecture-specific implementation of function
1165		arguments and results that are C unions is still not implemented.
1166
1167		----------------------------------------------------------------------
1168		Name: Allen Leung (leunga (at) reservoir (dot) com)
1169		Date: 2004/07/14 14:38:00 EST
1170		Tag: leunga-110_46_1-ppc-lwzu
1171		Description:
1172
1173		Added these instructions to the PowerPC architecture:
1174		LBZU(X), LHZU(X), LWZU(X),
1175		STWU(X), STFDU, STFSU
1176
1177		etc...
1178
1179		Note: I haven't added their instruction encoding into the description.
1180
1181		----------------------------------------------------------------------
1182		Name: Allen Leung (leunga (at) reservoir (dot) com)
1183		Date: 2004/07/13 15:04:00 EST
1184		Tag: leunga-110_46_1-ppc-lwarx
1185		Description:
1186
1187		Added the two instructions LWARX and STWCX to the PowerPC
1188		instruction set.
1189
1190		A (untested) rewrite of loop-structure.sml.  The old version
1191		is completely broken.
1192
1193		----------------------------------------------------------------------
1194		Name: Matthias Blume (blume (at) tti - c (dot) org)
1195		Date: 2004/07/13 13:50:00 CDT
1196		Tag: blume-20040713-nlffi
1197		Description:
1198
1199		- use paramAlloc to report c-calls with too many arguments
1200		(for PPC version where parameter area is pre-allocated)
1201		- added ccall_maxargspace to machspec (to implement the above)
1202		- made "make" commend in CM's "make" tool configurable
1203		- added option (default: on) for passing the name of the SML/NJ's "bin"
1204		directory to "make"; the call looks like this:
1205
1206		make <options> SMLNJ_BINDIR=<dir> <target>
1207
1208		This can be used by the Makefile to, e.g., pick the "right" version
1209		of ml-nlffigen.
1210		- minor code tweaks
1211
1212		----------------------------------------------------------------------
1213		Name: Matthias Blume (blume (at) tti - c (dot) org)
1214		Date: 2004/07/12 22:50:00 CDT
1215		Tag: blume-110_46_1-macosx-nlffi
1216		Description:
1217
1218		NLFFI under Mac OS X now working (sort of).  This is largely untested,
1219		though.
1220
1221		Note:  1. You have to make a new, clean build of the runtime system.
1222		2. There are new BOOTFILES, you have to use them!
1223		(Doing the bootstrap process yourself would be *very* painful!
1224		If you absolutely have to do it, build the system under
1225		a different architecture and then cross-compile.)
1226
1227		Version bumped to 110.46.1 to account for runtime data format changes.
1228
1229		----------------------------------------------------------------------
1230		Name: Matthias Blume (blume (at) tti - c (dot) org)
1231		Date: 2004/06/18 14:30:00 CDT
1232		Tag: blume-20040618-unix
1233		Description:
1234
1235		Changed the implementation of structure Unix so that the same stream
1236		is returned every time one of the {text,bin}{In,Out}streamOf functions
1237		is invoked on the same proc.  This is not what the spec currently
1238		says -- although IMO it arguably should.  (See discussion below.)
1239
1240		----------------------------------------------------------------------
1241		Name: Matthias Blume (blume (at) tti - c (dot) org)
1242		Date: 2004/06/17 18:15:00 CDT
1243		Tag: Release_110_46
1244		Description:
1245
1246		New working version (110.46).  NEW BOOTFILES!
1247
1248		----------------------------------------------------------------------
1249		Name: Matthias Blume (blume (at) tti - c (dot) org)
1250		Date: 2004/06/17 17:20:00 CDT
1251		Tag: blume-20040617-timer-unix
1252		Description:
1253
1254		Changed the interface of structures Timer and Unix to match the most
1255		recent Basis spec.
1256
1257		In the case of Unix there still seems to be an open/weird issue:
1258
1259		The {text,bin}{In,Out}streamOf functions are supposed to create
1260		fresh streams whenever they are called -- as opposed to have them
1261		return the same stream every time.  This design is supposed to
1262		prevent space leaks caused by proc values hanging on to streams.
1263
1264		The reap function, on the other hand, is supposed to close the
1265		streams.  This cannot be done without having a handle on the
1266		stream in proc after all...
1267
1268		I took the liberty to implement the following stopgap solution:
1269
1270		The proc value hangs on to the most recently created stream(s).
1271		Reap closes those.  If either or both of the two streams hadn't
1272		been created at all yet, then reap will close the corresponding
1273		file descriptors directly.
1274
1275		PS: I don't understand the original space leak argument anymore.  If
1276		a proc hangs on to the imperative stream, then I/O operations on those
1277		will advance the state of the cached stream and avoid the space leak.
1278
1279		----------------------------------------------------------------------
1280		Name: Matthias Blume (blume (at) tti - c (dot) org)
1281		Date: 2004/05/28 16:45:00 CDT
1282		Tag: blume-20040528-basis
1283		Description:
1284
1285		Added signature PACK_REAL and exported functor PrimIO.
1286
1287		----------------------------------------------------------------------
1288		Name: Matthias Blume (blume (at) tti - c (dot) org)
1289		Date: 2004/05/25 16:00:00 CDT
1290		Tag: blume-20040525-group-owner
1291		Description:
1292
1293		CM now ignores (but still accepts) the "owner" information in group
1294		descriptions.  The owner of a group is its next enclosing
1295		library. Each group must have a unique owner.  (There is a virtual
1296		"toplevel" library that owns groups which are not nested within a real
1297		library.)  Previously, each group had to explicitly declare its owner,
1298		and CM would check that such a declaration is correct.  The new scheme
1299		is to have CM check that for each group there is precisely one owning
1300		library.
1301
1302		The advantage of the new scheme is that the programmer no longer needs
1303		to maintain the somewhat annoying owner information.  The downside is
1304		that CM cannot enforce the ownership rule across multiple runs of
1305		CM.make.  Fortunately, enclosing the same group in two different
1306		libraries A and B which are not part of the same program does not
1307		cause real problems.
1308
1309		----------------------------------------------------------------------
1310		Name: Matthias Blume (blume (at) tti - c (dot) org)
1311		Date: 2004/05/20 16:00:00 CDT
1312		Tag: blume-20040520-win32
1313		Description:
1314
1315		Made the win32 version work again.  (Strangely, a misplaced comma had
1316		slipped into win32-process.c which prevented the runtime from being
1317		compiled correctly.)
1318
1319		Also, included a minor addition to ml-build.bat analogous to what was
1320		done in blume-20040519-ml-build.
1321
1322		----------------------------------------------------------------------
1323		Name: Matthias Blume (blume (at) tti - c (dot) org)
1324		Date: 2004/05/19 22:10:00 CDT
1325		Tag: blume-20040519-ml-build
1326		Description:
1327
1328		Arranged for ml-build to clean up after itself a little bit better.
1329		The script generates a temporary SML source file and compiles it using
1330		CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1331		gets rid of those at the end, so they don't accumulate under .cm.
1332
1333		This required a minor change to install.sh because the name of the
1334		metadata directory (default: .cm) is actually configurable at
1335		installation time.
1336
1337		----------------------------------------------------------------------
1338		Name: Matthias Blume (blume (at) tti - c (dot) org)
1339		Date: 2004/05/18 15:50:00 CDT
1340		Tag: blume-20040518-mkreader
1341		Description:
1342
1343		Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1344		implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1345
1346		----------------------------------------------------------------------
1347		Name: Matthias Blume (blume (at) tti - c (dot) org)
1348		Date: 2004/05/11 14:35:00 CDT
1349		Tag: blume-20040511-win32sock
1350		Description:
1351
1352		Added previously missing support for many socket-related functions
1353		under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1354		for the voluminous patch!
1355
1356		(I have not tested this patch under win32 yet.)
1357
1358		Here is David's e-mail:
1359
1360		Hi,
1361
1362		Attached to this email you find a diff against sml/nj 110.45
1363		that will enable socket support under Windows.
1364
1365		To apply the patch (using unix or cygwin)
1366		1) gunzip runtime.diff.gz
1367		2) "cd" into "src/runtime" in the source tree of a fresh
1368		110.45 installation.
1369		3) patch -p 1 < [your/path/to]runtime.diff
1370
1371		The code compiles fine but has NOT yet been extensively tested.
1372		I only ran a few tests for basic socket client functionality
1373		(which worked fine).  Especially the functions that use ioctl
1374		are not tested at all and might not work (see below).
1375
1376		I implemented this since we want to move to a newer version of sml/nj
1377		but need socket support in order to use it.  This is the first time I
1378		even had a look at the sml/nj source,  so please review my changes
1379		before making this part of the distribution!  Here are a few issues
1380		that I think might be better for someone to solve who is more
1381		familiar with the sml/nj source (and socket programming):
1382
1383		- getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1384		exception since I could not figure out what the windows equivalent
1385		of these functions is
1386
1387		- In sockets-osdep.h there are a some #include statements that are
1388		only used in a few files that include sockets-osdep.h
1389
1390		- In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1391		does not process its return value since I don't know how to
1392		report an error upwards.
1393
1394		- It would probably be good to have a call to WSACleanup() when
1395		the library is unloaded (if there is such a possibility).
1396		Otherwise I think Windows will take care of this automatically
1397		when the process finishes.
1398
1399		- I used ioctlsocket() as a replacement for ioctl() but I have
1400		no idea if that is actually the proper replacement on Windows.
1401
1402		- All these issues are marked in the code by "FIXME" comments.
1403
1404		We use sml/nj extensively in our products and are quite happy
1405		with it.  I hope this contribution will help you.
1406
1407		Keep up the good work!
1408
1409		David
1410
1411		----------------------------------------------------------------------
1412		Name: Matthias Blume (blume (at) tti - c (dot) org)
1413		Date: 2004/05/11 14:20:00 CDT
1414		Tag: blume-20040511-installml
1415		Description:
1416
1417		Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1418		the patch.)
1419
1420		----------------------------------------------------------------------
1421		Name: Matthias Blume (blume (at) tti - c (dot) org)
1422		Date: 2004/05/11 14:05:00 CDT
1423		Tag: blume-20040511-nlffi-netbsd
1424		Description:
1425
1426		Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1427		the patch.)
1428
1429		----------------------------------------------------------------------
1430		Name: Matthias Blume (blume (at) tti - c (dot) org)
1431		Date: 2004/05/11 12:05:00 CDT
1432		Tag: blume-20040511-exports
1433		Description:
1434
1435		As per request by Adam Chlipala <adam@hcoop.net>, extended various
1436		export lists in compiler-related .cm-files.
1437
1438		----------------------------------------------------------------------
1439		Name: Matthias Blume (blume (at) tti - c (dot) org)
1440		Date: 2004/05/11 11:35:00 CDT
1441		Tag: blume-20040511-allsource
1442		Description:
1443
1444		The installer now honors the "src-smlnj" target again, although its meaning
1445		has changed from "all sources required for the compiler" to "all sources
1446		the installer knows about".  In other words, if you enable "src-smlnj"
1447		in the "targets" file, then the installer will pull in sources for
1448		everything.  (Notice that this refers to source code only.  Compiled
1449		code is still only installed for modules that were requested explicitly
1450		or which are required for other modules that were requested explicitly.)
1451
1452		----------------------------------------------------------------------
1453		Name: Matthias Blume (blume (at) tti - c (dot) org)
1454		Date: 2004/04/23 17:40:00 CDT
1455		Tag: blume-20040423-ieee-scan
1456		Description:
1457
1458		Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1459		in the exponent calculation we get INF or ZERO (depending on the mantissa
1460		and the sign of the exponent).
1461
1462		----------------------------------------------------------------------
1463		Name: Matthias Blume (blume (at) tti - c (dot) org)
1464		Date: 2004/04/23 10:40:00 CDT
1465		Tag: blume-20040423-ml-build
1466		Description:
1467
1468		The ml-build script now terminates with a non-0 status when something
1469		goes wrong.
1470
1471		----------------------------------------------------------------------
1472		Name: Matthias Blume (blume (at) tti - c (dot) org)
1473		Date: 2004/04/22 16:35:00 CDT
1474		Tag: blume-20040422-Option
1475		Description:
1476
1477		Made exception Option to be the same as exception Option.Option
1478		(as it should be).
1479
1480		----------------------------------------------------------------------
1481		Name: Allen Leung (leunga (at) reservoir (dot) com)
1482		Date: 2004/03/19 14:40:00 EST
1483		Tag: leunga-20040319-cygwin-nlffi
1484		Description:
1485
1486		Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1487		of SML/NJ.  The problem is that
1488
1489		lib = dlopen(NULL, ...)
1490		f   = dlsym(lib, "malloc");
1491
1492		does not work on Windows unless we explicitly export symbols
1493		such as 'malloc' during linking.  We fixed this by explicitly
1494		exporting the required symbols with the magic gcc incantation:
1495
1496		-Wl,--export-all cygwin.def
1497
1498		where cygwin.def is a file containing all the symbols that
1499		we wish to export.
1500
1501		I suspect this is a Windows problem and we'll have to
1502		do the same (somehow with windows compilers) when
1503		we build the native win32 version with the system
1504		calls LoadLibrary/GetProcAddress.
1505
1506		----------------------------------------------------------------------
1507		Name: Matthias Blume (blume (at) tti - c (dot) org)
1508		Date: 2004/03/04 16:35:00 CST
1509		Tag: blume-20040304-intinf-fmt
1510		Description:
1511
1512		Fixed problem with IntInf.fmt (sign would show up on the right instead
1513		of on the left for BIN, OCT, and HEX).
1514
1515		----------------------------------------------------------------------
1516		Name: Matthias Blume (blume (at) tti - c (dot) org)
1517		Date: 2004/03/04 11:25:00 CST
1518		Tag: blume-20040304-symlinks
1519		Description:
1520
1521		Fixed problem with installer script (unix only) where bin/ml-yacc and
1522		friends pointed (via symlinks) to absolute locations instead of just
1523		.run-sml.  This was reported by Vesa A Norrman.
1524
1525		----------------------------------------------------------------------
1526		Name: Matthias Blume (blume (at) tti - c (dot) org)
1527		Date: 2004/02/13 14:50:00 CST
1528		Tag: Release_110_45
1529		Description:
1530
1531		New working version (110.45). New bootfiles.
1532
1533		----------------------------------------------------------------------
1534		Name: Matthias Blume (blume (at) tti - c (dot) org)
1535		Date: 2004/01/26 15:15:15 CST
1536		Tag: blume-20040126-toplevel
1537		Description:
1538
1539		Improved handling of exceptions at the interactive toplevel.
1540
1541		----------------------------------------------------------------------
1542		Name: Matthias Blume (blume (at) tti - c (dot) org)
1543		Date: 2004/01/26 11:25:00
1544		Tag: blume-20040126-app
1545		Description:
1546
1547		Type of top-level "app" corrected.
1548		Added code for setting vp_limitPtrMask to Win32-specific runtime.
1549
1550		----------------------------------------------------------------------
1551		Name: Matthias Blume (blume (at) tti - c (dot) org)
1552		Date: 2003/11/18 17:10 CST
1553		Tag: blume-20031118-basis-fiddle
1554		Description:
1555
1556		- changed Timer interface to what might become the spec
1557		- POSIX_FLAGS -> BIT_FLAGS according to spec
1558		- some other minor discrepancies wrt. spec eliminated
1559
1560		----------------------------------------------------------------------
1561		Name: Matthias Blume (blume (at) tti - c (dot) org)
1562		Date: 2003/11/06 12:00:00 CST
1563		Tag: Release_110_44
1564		Description:
1565
1566		New working version (110.44). New bootfiles.
1567
1568		----------------------------------------------------------------------
1569		Name: Matthias Blume (blume (at) tti - c (dot) org)
1570		Date: 2003/11/04 11:50:00 CST
1571		Tag: blume-20031104-move-libraries
1572		Description:
1573
1574		Eliminated the "dont_move_libraries" directive in config/targets.
1575		(The mechanism was broken and could not be fixed easily.  Moreover,
1576		there does not seem to be any reason not to move all libraries into
1577		lib during installation.  I originally implemented this directive as a
1578		backward-compatibility feature when I first introduced the new CM.
1579		Now that things have been stable for a long time and going back to the
1580		old CM is not an option, there is no reason to keep it around.)
1581
1582		----------------------------------------------------------------------
1583		Name: Matthias Blume (blume (at) tti - c (dot) org)
1584		Date: 2003/11/03 16:00:00 CST
1585		Tag: blume-20031103-installdir
1586		Description:
1587
1588		Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1589		Richards for pointing out the problem and providing the solution.)
1590
1591		----------------------------------------------------------------------
1592		Name: Matthias Blume (blume (at) tti - c (dot) org)
1593		Date: 2003/10/01 17:05:00 CDT
1594		Tag: blume-20031001-lal-mlrisc
1595		Description:
1596
1597		MLRISC bug fix from Lal.
1598
1599		----------------------------------------------------------------------
1600		Name: Matthias Blume (blume (at) tti - c (dot) org)
1601		Date: 2003/09/30 16:10:00 CDT
1602		Tag: blume-20030930-primio-bat
1603		Description:
1604
1605		1. Added openVector, nullRd, and nullWr to PRIM_IO.
1606		2. Improved .bat files (for Win32 port) to make things work under Win95.
1607		(thanks to Aaron S. Hawley for this one)
1608
1609		----------------------------------------------------------------------
1610		Name: Matthias Blume (blume (at) tti - c (dot) org)
1611		Date: 2003/09/26 16:05:00 CDT
1612		Tag: blume-20030926-wrappriv
1613		Description:
1614
1615		Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1616
1617		----------------------------------------------------------------------
1618		Name: Matthias Blume (blume (at) tti - c (dot) org)
1619		Date: 2003/09/26 15:00:00 CDT
1620		Tag: blume-20030926-110_43_3
1621		Description:
1622
1623		- additional cleanup
1624		- version number bump, NEW BOOTFILES
1625
1626		----------------------------------------------------------------------
1627		Name: Matthias Blume (blume (at) tti - c (dot) org)
1628		Date: 2003/09/26 12:00:00 CDT
1629		Tag: blume-20030926-ppautoload
1630		Description:
1631
1632		I modified the read-eval-print loop so that the autoloader gets
1633		invoked whenever the prettyprinter tries to look up a symbol that
1634		is not currently defined in the toplevel environment but which
1635		appears in CM's autoload registry.  As a result, we see far fewer of
1636		those ?.Foo.Bar.xxx names in the prettyprinter's output.
1637
1638		In addition to this I tried to clean up some pieces of the Basis
1639		implementation (e.g., Socket, Word8Array) in order to prevent other
1640		instances of these ?.Foo.Bar.xxx names from being printed.
1641
1642		The mechanism that picks names for types still needs some work, though.
1643		(Right now it seems that if there is a type A.t which is defined to
1644		be B.u, but B is unavailable at toplevel, then A.t gets printed as
1645		"?.B.u" although the perhaps more sensible solution would be to use
1646		"A.t" in this case.  In other words, the prettyprinter should follow
1647		a chain of DEFtycs not farther than there are corresponding toplevel
1648		names in the current environment.)
1649
1650		----------------------------------------------------------------------
1651		Name: Matthias Blume (blume (at) tti - c (dot) org)
1652		Date: 2003/09/24 16:31:00 CDT
1653		Tag: blume-20030924-installer
1654		Description:
1655
1656		Another installer tweak:  All the ML code for the installer is now
1657		compiled during CMB.make and put into a little library called
1658		$smlnj/installer.cm.  The installation then simply invokes
1659
1660		sml -m $smlnj/installer.cm
1661
1662		and everything happens automagically.
1663
1664		Win32:  ML code senses value of environment variable SMLNJ_HOME.
1665		Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1666		and BINDIR.
1667
1668		The new scheme guarantees that the ML code responsible for the installation
1669		is in sync with the APIs of the main system.  Also, the installer is
1670		somewhat faster because the installer script is precompiled.
1671
1672		----------------------------------------------------------------------
1673		Name: Matthias Blume (blume (at) tti - c (dot) org)
1674		Date: 2003/09/24 15:35:00 CDT
1675		Tag: blume-20030924-synsock
1676		Description:
1677
1678		Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1679		but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1680		and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1681		implementations of non-blocking operations from CML's Socket
1682		structure.
1683
1684		----------------------------------------------------------------------
1685		Name: Matthias Blume (blume (at) tti - c (dot) org)
1686		Date: 2003/09/24 15:10:05 CDT
1687		Tag: blume-20030924-sockets
1688		Description:
1689
1690		1. Fixed SOCKET API and implementation to match Basis spec.
1691		This required changing the internal representation of sockets to one
1692		that remembers (for each socket file descriptor) whether it is currently
1693		blocking or non-blocking.  This state is maintained lazily (i.e., a system
1694		call is made only if the state actually needs to change).
1695
1696		2. OS-specific details of sockets were moved into separate files, thus
1697		making it possible to unify the bulk of the socket implementations
1698		between Unix and Win32.
1699
1700		3. CML's socket API changed accordingly.
1701		(Note that we need to remove non-blocking functions from this API
1702		since they are redundant in the case of CML!)
1703
1704		4. CML's socket implementation now makes use of non-blocking functions
1705		provided by Basis, thus removing all OS-dependent code from this part
1706		of CML.
1707
1708		5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1709
1710		----------------------------------------------------------------------
1711		Name: Matthias Blume (blume (at) tti - c (dot) org)
1712		Date: 2003/09/22 12:10:00 CDT
1713		Tag: blume-20030922-110_43_2
1714		Description:
1715
1716		Made a new interim version and bootfiles for developer's bootstrapping
1717		convenience.
1718
1719		110.43.2 -- NEW BOOTFILES
1720
1721		----------------------------------------------------------------------
1722		Name: Matthias Blume (blume (at) tti - c (dot) org)
1723		Date: 2003/09/19 15:55:00 CDT
1724		Tag: blume-20030919-cmdir
1725		Description:
1726
1727		1. new-install.sh -> install.sh
1728		2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1729		3. tweaked installer so that another name instead of .cm can be chosen
1730		at install time (by setting the CM_DIR_ARC environment variable
1731		during installation); once installation is complete, the name is
1732		fixed
1733
1734		----------------------------------------------------------------------
1735		Name: Matthias Blume (blume (at) tti - c (dot) org)
1736		Date: 2003/09/18 16:00:00 CDT
1737		Tag: blume-20030918-110_43_1
1738		Description:
1739
1740		Made a new interim version and bootfiles for developer's bootstrapping
1741		convenience.
1742
1743		110.43.1 -- NEW BOOTFILES
1744
1745		----------------------------------------------------------------------
1746		Name: Matthias Blume (blume (at) tti - c (dot) org)
1747		Date: 2003/09/18 15:20:00 CDT
1748		Tag: blume-20030918-misc
1749		Description:
1750
1751		1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
1752		this seems to be controversial at the moment) and used those in
1753		Posix.ProcEnv.times.
1754
1755		2. Added Time.{from,to}Nanoseconds to Time.
1756
1757		3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
1758		For example, fromLargeInt never needs to look at more than 3 "big
1759		digits" to get its 53 bits of precision.
1760
1761		----------------------------------------------------------------------
1762		Name: Matthias Blume (blume (at) tti - c (dot) org)
1763		Date: 2003/09/17 16:30:00 CDT
1764		Tag: blume-20030917-real32-slices
1765		Description:
1766
1767		Added an entry to the primitive environment
1768		(compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
1769		added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
1770
1771		Removed some of the "magic" constants in real64.sml and replaced them
1772		with code that generates these values from their corresponding
1773		integer counterparts.
1774
1775		Made all(?) the slice-related changes to the Basis and made everything
1776		compile again...
1777
1778		----------------------------------------------------------------------
1779		Name: Matthias Blume (blume (at) tti - c (dot) org)
1780		Date: 2003/09/15 17:45:00 CDT
1781		Tag: blume-20030915-rbase
1782		Description:
1783
1784		Fixed bug in Real.fromLargeInt.
1785
1786		----------------------------------------------------------------------
1787		Name: Matthias Blume (blume (at) tti - c (dot) org)
1788		Date: 2003/09/13 18:11:00 CDT
1789		Tag: blume-20030913-libinstall
1790		Description:
1791
1792		Minor bugfix in config/libinstall (set anchor with path to
1793		standalone tool after installing it, otherwise libraries that
1794		need ml-lex or ml-yacc won't compile the first time the installer
1795		runs).
1796
1797		----------------------------------------------------------------------
1798		Name: Matthias Blume (blume (at) tti - c (dot) org)
1799		Date: 2003/09/12 11:45:00 CDT
1800		Tag: blume-20030912-various
1801		Description:
1802
1803		- fixed bug in Real.toLargeInt
1804		- fixed bug in Posix.ProcEnv.times
1805		- changed inputLine functions to return an option
1806		- minor installer improvements / bugfixes
1807		- changed default @SMLalloc parameter for x86/celeron to 64k
1808
1809		----------------------------------------------------------------------
1810		Name: Matthias Blume (blume (at) tti - c (dot) org)
1811		Date: 2003/09/09 22:00:00 CDT
1812		Tag: Release_110_43
1813		Description:
1814
1815		New working release 110.43.  New bootfiles.
1816
1817		----------------------------------------------------------------------
1818		Name: Matthias Blume (blume (at) tti - c (dot) org)
1819		Date: 2003/09/09 19:20:00 CDT
1820		Tag: blume-20030909-installer
1821		Description:
1822
1823		Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
1824		Modified config/install.bat to take advantage of it.  Also modified
1825		config/install.sh (and called it config/new-install.sh) to take advantage
1826		of it on Unix systems.  (The SML code is (supposed to be) platform-
1827		independent.)
1828
1829		The installer can now install everything under Win32
1830		as well as under *nix as long as it compiles.
1831
1832		Other changes:
1833
1834		- made CML compile again under Win32
1835		- made eXene compile under Win32 (by providing a fake structure UnixSock
1836		and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
1837		- fixed a bug in nowhere: it assumed that type OS.Process.status is the
1838		same as type int;  under Win32 it isn't
1839		- fixed some slice-related problems in the win32-specific parts of CML
1840		- added a functor argument "sameVol" to os-path-fn.sml in the Basis
1841		(under Win32, the volume name is case-insensitive, and the
1842		OS.Path code compares volume names for equality)
1843
1844		----------------------------------------------------------------------
1845		Name: Matthias Blume (blume (at) tti - c (dot) org)
1846		Date: 2003/09/08 11:55:00 CDT
1847		Tag: blume-20030908-fullpath
1848		Description:
1849
1850		Made Win32 version of OS.FileSys.fullPath return current directory
1851		when given an empty string.  This is what the spec says, and incidentally,
1852		CM depends on it.  (CM otherwise goes into an infinite loop in certain
1853		cases when presented with the name of a non-existing .cm file.)
1854
1855		----------------------------------------------------------------------
1856		Name: Matthias Blume (blume (at) tti - c (dot) org)
1857		Date: 2003/09/04 16:30:00 CDT
1858		Tag: blume-20030905-slices-etc
1859		Description:
1860
1861		1. Changed interface to vectors and arrays in Basis to match
1862		(draft) Basis spec.
1863		2. Added signatures and implementations of slices according to
1864		Basis spec.
1865		3. Edited source code throughout the system to make it compile again
1866		under 1. and 2.  (In some cases code had to be added to have it
1867		match the new signatures.)
1868		4. MLRISC should be backward-compatible: the copies of the originals
1869		of files that needed to change under 3. were retained, the .cm files
1870		check the compiler version number and use old versions when
1871		appropriate.
1872		5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
1873		dirstream -> string option (in accordance with Basis spec).
1874		6. When generating code that counts lines, ml-lex used function
1875		CharVector.foldli, taking advantage of its old interface.
1876		This has been replaced with the corresponding code from
1877		CharVectorSlice.  (html-lex must be re-lexed!)
1878		7. BitArray in smlnj-lib/Util has been extended/modified to match the
1879		new MONO_ARRAY signature.  (Do we need BitArraySlice?)
1880		8. Removed temporary additions (fromInternal, toInternal) from the
1881		(now obsolete) IntInf in smlnj-lib/Util.
1882		9. Cleaned up structure Byte.
1883		10. Added localOffset, scan, and fromString to Date (according to spec).
1884		Cleaned/corrected implementation of Date.
1885		(Still need to check for correctness; implement better canonicalizeDate.)
1886		11. Added "scan" to signature IEEE_REAL.
1887		12. Some improvements to IntInf [in particular: efficiency-hack for
1888		mod and rem when second operand is 2 (for parity checks).]
1889		13. Changed representation of type Time.time, using a single IntInf.int
1890		value counting microseconds.  This considerably simplified the
1891		implementation of structure Time.  We now support negative time
1892		values; scan and fromString handle signs.
1893		14. Functor PrimIO now takes two additional arguments (VectorSlice and
1894		ArraySlice).
1895
1896		----------------------------------------------------------------------
1897		Name: Matthias Blume (blume (at) tti - c (dot) org)
1898		Date: 2003/08/28 17:00:00 CDT
1899		Tag: blume-20030828-intinf
1900		Description:
1901
1902		This is a major update which comes with a version number bump
1903		(110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
1904		and an implementation of IntInf in the Basis.
1905
1906		There are a fairly large number of related changes and updates throughout
1907		the system:
1908
1909		Basis:
1910		- Implemented IntInf.
1911		- Made LargeInt a projection of IntInf (by filtering through INTEGER).
1912		- Added some missing Real64 operations, most notably Real.toLargeInt.
1913		- Added FixedInt as a synonym for Int32.
1914
1915		compiler:
1916		* Added support for a built-in intinf type.
1917		- literals
1918		- pattern matching
1919		- conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
1920		- overloading on literals and operations
1921
1922		This required adding a primitive type intinf, some additional
1923		primops, and implementations for several non-trivial intinf
1924		operations in Core.  (The intinf type is completely abstract
1925		to the compiler; all operations get delegated back to the Core.)
1926
1927		* Intinf equality is handled by polyequal.  However, the compiler
1928		does not print its usual warning in this case (since polyequal
1929		is the right thing to do there).
1930
1931		* Improved the organization of structure InlineT.
1932
1933		* A word about conversion primops:
1934		If conversions involving intinf do not cancel out during
1935		CPS contract, then the compiler must insert calls to Core functions.
1936		Since all core access must be resolved already during the FLINT
1937		translate phase, it would be too late a the time of CPS contract
1938		to add new Core calls.  For this reason, conversion primops
1939		for intinf carry two arguments:  1. the numeric argument that
1940		they are supposed to convert, and 2. the Core function that
1941		can help with this conversion if necessary.  If CPS contract
1942		eliminates a primop, then the associated Core function becomes
1943		dead and goes away.  Intinf conversion primops that do not get
1944		eliminated by CPS contract get rewritten into calls of their
1945		core functions by a separate, new phase.
1946
1947		interactive system:
1948		- Control.Print.intinfDepth controls max length of intinf constants
1949		being printed. (Analogous to Control.Print.stringDepth.)
1950		- Cleanup in printutil and pputil: got rid of unused stuff and
1951		duplicates; replaced some of the code with code that makes better
1952		use of library functionality.
1953
1954		CM:
1955		Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
1956		are no longer silent.
1957
1958		CKIT:
1959		Fixed mismatched uses of Int32 and LargeInt.  I always decided
1960		in favor of LargeInt -- which is now the same as IntInf.
1961		CKIT-knowledgable people should check whether this is what's
1962		intended and otherwise change things back to using Int32 or
1963		FixedInt.
1964
1965		Throughout the code:
1966		Started using IntInf.int literals and built-in operations
1967		(e.g., comparison with 0) where this seems appropriate.
1968
1969
1970		----------------------------------------------------------------------
1971		Name: Dave MacQueen (dbm@cs.uchicago.edu)
1972		Date: 2003/08/13 11:36:00 CDT
1973		Tag: dbm-20030813-mcz-merge1
1974		Description:
1975
1976		Merging changes from the mcz-branch development branch into trunk.
1977		These changes involve replacement of the emulated old prettyprinter
1978		interface with direct use of the SML/NJ Lib PP library, and fixing
1979		of a couple of bugs (895, 1186) relating to error messages.  A new
1980		prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
1981		has been added.
1982
1983		----------------------------------------------------------------------
1984		Name: Matthias Blume (blume (at) tti - c (dot) org)
1985		Date: 2003/08/11 15:45:00 CDT
1986		Tag: blume-20030811-windows
1987		Description:
1988
1989		Version number bumped to 110.42.9.  NEW BOOTFILES!!!
1990
1991		http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
1992
1993		This patch restores SML/NJ's ability to run under win32.  There are a
1994		number of changes, including fixes for several bugs that had gone
1995		unnoticed until now:
1996
1997		- uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
1998		trying to run the win32 version from within cygwin.)
1999
2000		- There are a number of simple .bat scripts that substitute for their
2001		corresponding Unix shell-scripts.  (See below.)
2002
2003		- The internals of ml-build have been modified slightly.  The main
2004		difference is that instead of calling ".link-sml" (or link-sml.bat)
2005		using OS.Process.system, the ML process delegates this task back
2006		to the script.  Otherwise problems arise in mixed environments such
2007		as Cygwin where scripts look and work like Unix scripts, but
2008		where OS.Process.system cannot run them.
2009
2010		- In CM, the srcpath pickler used native pathname syntax -- which
2011		is incorrect in the case of cross-compilation.  The new pickle format
2012		is independent of platform-specific naming conventions.
2013
2014		- Path configuration files (such as lib/pathconfig) can now choose
2015		between native and standard syntax.  Placing a line of the form
2016
2017		standard!
2018
2019		into the file causes all subsequent paths to be interpreted using
2020		CM standard pathname syntax (= Unix conventions); a line
2021
2022		native!
2023
2024		switches back to native style.  This was needed so that
2025		path config files can be written portably, see src/system/pathconfig.
2026
2027		- Runtime system:
2028
2029		- win32-filesys.c:  get_file_time and set_file_time now
2030		access modification time, not creation time.
2031
2032		- I/O code made aware of new array representation.
2033
2034		- Bug fixes in X86.prim.masm.
2035
2036		- src/system/makeml made aware of win32. (For use under cygwin
2037		and other Unix-environments for windows.)
2038
2039		- In Basis, fixed off-by-one error in win32-io.sml (function vecF)
2040		which caused BinIO.inputAll to fail consistently.
2041
2042		.bat scripts:
2043
2044		Windows .bat scripts assume that SMLNJ_HOME is defined.
2045
2046		- sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
2047		applications (sml, ml-yacc, ml-lex).
2048		- ml-build.bat: analogous to ml-build.
2049		- config\install.bat:  Analogous to config/install.sh.  This requires
2050		that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2051		(nmake etc. must be on the path, and vcvars32 must have been run.)
2052		Moreover, sources for ml-lex and ml-yacc need to exist under src,
2053		and the bootfile hierarchy must have been unpacked under
2054		sml.boot.x86-win32.
2055		The script is very primitive and does a poor job at error checking.
2056		It only installs the base system, ml-lex, and ml-yacc.  No other
2057		libraries are being installed (i.e., you get only those that
2058		are part of the compiler.)
2059		- link-sml.bat:  analogous to .link-sml, but not currently used
2060
2061		Unrelated bug fixes:
2062
2063		- ml-nlffigen now exports structures ST_* corresponding to incomplete
2064		types.
2065		- Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
2066		otherwise.)
2067
2068		----------------------------------------------------------------------
2069		Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2070		Date: 2003/06/17
2071		Tag: macqueen-20030617-bug895
2072		Description:
2073
2074		Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2075		Tag will be used for new development branch (mcz-branch) for
2076		use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2077
2078		----------------------------------------------------------------------
2079		Name: Matthias Blume (blume (at) tti - c (dot) org)
2080		Date: 2003/05/27 16:55:00 CDT
2081		Tag: blume-20030527-polyeq
2082		Description:
2083
2084		Tried to eliminated most cases of polymorphic equality.
2085
2086		----------------------------------------------------------------------
2087		Name: Matthias Blume (blume (at) tti - c (dot) org)
2088		Date: 2003/05/21 17:45:00 CDT
2089		Tag: blume-20030517-complete
2090		Description:
2091
2092		Two changes:
2093
2094		1. Added a flag for controlling whether non-exhaustive bindings will
2095		be treated as errors (default is false).
2096		2. Cleaned up the *entire* source tree so that CMB.make goes through
2097		without a single non-exhaustive match- or bind warning.
2098
2099		----------------------------------------------------------------------
2100		Name: Matthias Blume (blume (at) tti - c (dot) org)
2101		Date: 2003/05/17 10:20:00 CDT
2102		Tag: blume-20030517-absyn
2103		Description:
2104
2105		1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2106
2107		This mainly affects the quality of error messages.  However, some
2108		of the code is now more straightforward than before. (Treatment of
2109		the above four constructs in translate.sml is much simpler than
2110		the "macro-expansion" that was going on before.  Plus, the mach-
2111		compiler no longer gets invoked just to be able to compile an
2112		if-expression.)
2113
2114		2. The ErrorMsg.Error exception is now caught and absorbed by the
2115		interactive loop.
2116
2117		----------------------------------------------------------------------
2118		Name: Allen Leung
2119		Date: 2003/05/16 13:05:00 CDT
2120		Tag: leunga-20030516-cygwin-runtime
2121		Description:
2122
2123		Ported the runtime system to cygwin, which uses the unix
2124		x86-unix bin files.  Missing/buggy features:
2125
2126		o getnetbyname, getnetbyaddr: these functions seem to be missing in
2127		the Cygwin library.
2128		o Ctrl-C handling may be flaky.
2129		o Windows system calls and Windows I/O are not supported.
2130
2131		A new set of binfiles is located at:
2132
2133		http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2134
2135		This is only needed for bootstrapping the cygwin version of smlnj.
2136		Other x86 versions can use the existing binfiles.
2137
2138		----------------------------------------------------------------------
2139		Name: Matthias Blume
2140		Date: 2003/04/08 15:42:00 CDT
2141		Tag: blume-20030408-listpair
2142		Description:
2143
2144		1. Added a target 'mlrisc' to installer.
2145
2146		2. Added missing elements to structure ListPair.
2147
2148		----------------------------------------------------------------------
2149		Name: Allen Leung
2150		Date: 2003/01/07 10:40:00 EST
2151		Tag: leunga-20030107-int-rem
2152		Description:
2153
2154		Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2155		The arguments to the SUBL instruction were swapped.
2156
2157		----------------------------------------------------------------------
2158		Name: Matthias Blume
2159		Date: 2002/12/12 16:25:00 EST
2160		Tag: blume-20021212-risc-ra
2161		Description:
2162
2163		Fixed a serious bug in the rewrite code for FP spilling/reloading that
2164		sent the RA into an infinite loop when floating point registers get
2165		spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2166		110.37 and 110.38.)
2167		There was another set of potential problems related to the handling of
2168		MLRISC annotations (but those did not yet cause real problems, apparently).
2169
2170		----------------------------------------------------------------------
2171		Name: Matthias Blume
2172		Date: 2002/12/06 22:40:00 EST
2173		Tag: blume-20021206-cm-fileid
2174		Description:
2175
2176		Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2177		This fixes the problem of CM getting confused by files that suddenly
2178		change their identity (e.g., by getting unlinked and recreated by some
2179		text editor such as vi).  There might be a better/cheaper/cleaner way
2180		of doing this, but for now this will have to do.
2181
2182		----------------------------------------------------------------------
2183		Name: Matthias Blume
2184		Date: 2002/10/28 09:50:00 EST
2185		Tag: blume-20021028-typecheck
2186		Description:
2187
2188		Exported structure Typecheck from $smlnj/viscomp/core.cm.
2189
2190		----------------------------------------------------------------------
2191		Name: Matthias Blume
2192		Date: 2002/10/17 09:10:00 EDT
2193		Tag: Release_110_42
2194		Description:
2195
2196		In good old tradition, there has been a slight hiccup so that we have
2197		to patch 110.42 after the fact.   The old release tag has been replaced
2198		(see below).
2199
2200		The change solves a problem with two competing approaches the
2201		configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2202		each other's way.
2203
2204		This change only affects the runtime system code and the installer script.
2205		(No new bootfiles.)
2206
2207		----------------------------------------------------------------------
2208		Name: Matthias Blume
2209		Date: 2002/10/16 12:00:00 EDT
2210		Tag: Release_110_42_removed
2211		Description:
2212
2213		New working release.  New bootfiles.
2214
2215		----------------------------------------------------------------------
2216		Name: Matthias Blume
2217		Date: 2002/10/10 13:10:00 EDT
2218		Tag: blume-20021010-ppc-divs
2219		Description:
2220
2221		The mltree operator DIVS must be implemented with an overflow check on
2222		the PPC because the hardware indicates divide-by-zero using "overflow" as
2223		well.
2224
2225		----------------------------------------------------------------------
2226		Name: Matthias Blume
2227		Date: 2002/07/23 11:45:00
2228		Tag: blume-20020723-smlnj-home
2229		Description:
2230
2231		Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2232		then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2233		CM_PATHCONFIG is also set), the path configuration file is assumed
2234		to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2235		the entire tree to some other place and everything will "just work".
2236
2237		(Companion commands such as ml-build and ml-makedepend also sense this
2238		variable.)
2239
2240		----------------------------------------------------------------------
2241		Name: Matthias Blume
2242		Date: 2002/07/12 21:19:00 EDT
2243		Tag: blume-20020712-liveness
2244		Description:
2245
2246		Exported two useful "step" functions from liveness module (MLRISC).
2247
2248		----------------------------------------------------------------------
2249		Name: Matthias Blume
2250		Date: 2002/07/05 16:00 EDT
2251		Tag: Release_110_41
2252		Description:
2253
2254		New working release.  New bootfiles.
2255
2256		----------------------------------------------------------------------
2257		Name: Matthias Blume
2258		Date: 2002/07/05 10:25:00 EDT
2259		Tag: blume-20020705-btimp
2260		Description:
2261
2262		Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2263		other clients can set up backtracing support.
2264
2265		----------------------------------------------------------------------
2266		Name: Matthias Blume
2267		Date: 2002/06/25 17:23:00 EDT
2268		Tag: blume-20020625-fpmax
2269		Description:
2270
2271		Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2272		case.  (The sense of the isNaN test was reversed -- which made min and
2273		max always return their first argument.)
2274
2275		----------------------------------------------------------------------
2276		Name: Matthias Blume
2277		Date: 2002/06/11
2278		Tag: blume-20020611-unixpath
2279		Description:
2280
2281		Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2282
2283		----------------------------------------------------------------------
2284		Name: Matthias Blume
2285		Date: 2002/06/10 16:35:00 EDT
2286		Tag: blume-20020610-ieeereal
2287		Description:
2288
2289		I back-ported my implementation of IEEEReal.fromString from the
2290		idlbasis-devel branch so that we can test it.
2291
2292		Another small change is that ppDec tries to give more information
2293		than just "<sig>" in the case of functors.  However, this code is
2294		broken in some mysterious way if the functor's body's signature
2295		has not been declared by ascription but gets inferred from the
2296		implementation.  This needs fixing...
2297
2298		----------------------------------------------------------------------
2299		Name: Matthias Blume
2300		Date: 2002/05/31
2301		Tag: blume-20020531-btrace-mode
2302		Description:
2303
2304		Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2305		the wayside when I switched over to using Controls everywhere.)
2306
2307		----------------------------------------------------------------------
2308		Name: Lal George
2309		Date: 2002/05/23 12:21:40 EDT
2310		Tag: george-20020523-visual-labels
2311		Description:
2312
2313		Labels are now displayed in the graphical output to make
2314		the fall-through and target blocks obvious.
2315
2316		----------------------------------------------------------------------
2317		Name: Matthias Blume
2318		Date: 2002/05/22 11:03:00 EDT
2319		Tag: blume-20020522-shrink
2320		Description:
2321
2322		John tweaked yesterday's fix for 1131 to handle an out-of-memory
2323		situation that comes up when allocating huge arrays.
2324
2325		----------------------------------------------------------------------
2326		Name: Matthias Blume
2327		Date: 2002/05/21 16:00:00 EDT
2328		Tag: Release_110_40
2329		Description:
2330
2331		New working release (110.40).  New bootfiles.
2332
2333		[Also: John Reppy fixed GC bug 1131.]
2334
2335		----------------------------------------------------------------------
2336		Name: Matthias Blume
2337		Date: 2002/05/21 12:35:00 EDT
2338		Tag: blume-20020521-cmdoc
2339		Description:
2340
2341		CM documentation update.
2342
2343		----------------------------------------------------------------------
2344		Name: Matthias Blume
2345		Date: 2002/05/21 10:55:00 EDT
2346		Tag: blume-20020521-misc
2347		Description:
2348
2349		- John tweaked runtime to be silent on heap export (except when
2350		GC messages are on).
2351		- I added a few more things (cross-compiling versions of CMB) to
2352		config/preloads (as suggestions).
2353
2354		----------------------------------------------------------------------
2355		Name: Matthias Blume
2356		Date: 2002/05/20 22:25:00 EDT
2357		Tag: blume-20020520-controls
2358		Description:
2359
2360		- Added ControlUtil structure to control-lib.cm.
2361		- Use it throughout.
2362		- Used Controls facility to define MLRISC controls (as opposed to
2363		registering MLRISC control ref cells with Controls after the fact)
2364		- Fixed messed-up controls priorities.
2365
2366		* Removed again all the stuff from config/preloads that one wouldn't
2367		be able to preload at the time the initial heap image is built.
2368		(Many libraries, e.g., CML, do not exist yet at this time.  The
2369		only libraries that can be preloaded via config/preloads are those
2370		that come bundled with the bootfiles.)
2371
2372		----------------------------------------------------------------------
2373		Name: Matthias Blume
2374		Date: 2002/05/20 10:59:00 EDT
2375		Tag: blume-20020520-preloads
2376		Description:
2377
2378		Added a lot of commented-out suggestions for things to be included
2379		in config/preloads.
2380
2381		----------------------------------------------------------------------
2382		Name: Allen Leung
2383		Date: 2002/05/18 14:20:00 EDT
2384		Tag: leunga-20020518-mdl
2385		Description:
2386
2387		o Made the mdl tool stuff compile and run again.
2388		o I've disabled all the stuff that depends on RTL specifications; they
2389		are all badly broken anyway.
2390
2391		----------------------------------------------------------------------
2392		Name: Matthias Blume
2393		Date: 2002/05/17 16:49:00 EDT
2394		Tag: blume-20020517-controls
2395		Description:
2396
2397		0. John Reppy made several modifications to the SML/NJ library.
2398		In particular, there is a shiny new controls-lib.cm.
2399
2400		1. Pushed new controls interface through compiler so that everything
2401		compiles again.
2402
2403		2. Added FormatComb and FORMAT_COMB to the CML version of the
2404		SML/NJ library (so that CML compiles again).
2405
2406		3. Modified init scripts because XXX_DEFAULT environment variables
2407		are no longer with us.  (Boot-time initialization is now done
2408		using the same environment variables that are also used for
2409		startup-time initialization of controls.)
2410
2411		----------------------------------------------------------------------
2412		Name: Lal George
2413		Date: 2002/05/15 09:20:10 EDT
2414		Tag: george-20020515-pseudo-op-decls
2415		Description:
2416
2417		All pseudo-ops emitted before the first segment declaration
2418		such as TEXT, DATA, and BSS directives are assumed to be global
2419		declarations and are emitted first in the assembly file. This is
2420		useful in a number of situations where one has pseudo-ops that are not
2421		specific to any segment, and also works around the constraint that one
2422		cannot have client pseudo-ops in the TEXT segment.
2423
2424		Because no segment is associated with these declarations it is
2425		an error to allocate any space or objects before the first segment
2426		directive and an exception will be raised. However, we cannot make
2427		this check for client pseudo-ops.
2428
2429		These top level declarations are a field in the CFG graph_info.
2430		In theory you can continue to add to this field after the CFG has been
2431		built -- provided you know what you are doing;-)
2432
2433		----------------------------------------------------------------------
2434		Name: Matthias Blume
2435		Date: 2002/05/13 16:40:00 EDT
2436		Tag: blume-20020513-pp-etc
2437		Description:
2438
2439		A few minor bugfixes:
2440
2441		- Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2442		(Bogus printouts for redefined bindings still occur.  Compiler
2443		bug should no longer occur now.  We need to redo the prettyprinter
2444		from scratch.)
2445
2446		- CM pathname printer now also adds escape sequences for ( and )
2447
2448		- commend and docu fixes for ml-nlffi
2449
2450		----------------------------------------------------------------------
2451		Name: Matthias Blume
2452		Date: 2002/05/10 16:40:00 EDT
2453		Tag: blume-20020510-erg-textio
2454		Description:
2455
2456		Applied the following bugfix provided by Emden Gansner:
2457
2458		Output is corrupted when outputSubstr is used rather than output.
2459
2460		The problem occurs when a substring
2461
2462		ss = (s, dataStart, dataLen)
2463
2464		where dataStart > 0, fills a stream buffer with avail bytes left.
2465		avail bytes of s, starting at index dataStart, are copied into the
2466		buffer, the buffer is flushed, and then the remaining dataLen-avail
2467		bytes of ss are copied into the beginning of the buffer. Instead of
2468		starting this copy at index dataStart+avail in s, the current code
2469		starts the copy at index avail.
2470
2471		Fix:
2472		In text-io-fn.sml, change line 695 from
2473		val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2474		to
2475		val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2476
2477		----------------------------------------------------------------------
2478		Name: Matthias Blume
2479		Date: 2002/04/12 13:55:00 EDT
2480		Tag: blume-20020412-assyntax
2481		Description:
2482
2483		1. Grabbed newer assyntax.h from the XFree86 project.
2484		2. Fiddled with how to compile X86.prim.asm without warnings.
2485		3. (Very) Minor cleanup in CM.
2486
2487		----------------------------------------------------------------------
2488		Name: Matthias Blume
2489		Date: 2002/04/01 (no joke!) 17:07:00 EST
2490		Tag: blume-20020401-x86div
2491		Description:
2492
2493		Added full support for div/mod/rem/quot on the x86, using the machine
2494		instruction's two results (without clumsily recomputing the remainder)
2495		directly where appropriate.
2496
2497		Some more extensive power-of-two support was added to the x86 instruction
2498		selector (avoiding expensive divs, mods, and muls where they can be
2499		replaced with cheaper shifts and masks).  However, this sort of thing
2500		ought to be done earlier, e.g., within the CPS optimizer so that
2501		all architectures benefit from it.
2502
2503		The compiler compiles to a fixed point, but changes might be somewhat
2504		fragile nevertheless.  Please, report any strange things that you might
2505		see wrt. div/mod/quot/rem...
2506
2507		----------------------------------------------------------------------
2508		Name: Matthias Blume
2509		Date: 2002/03/29 17:22:00
2510		Tag: blume-20020329-div
2511		Description:
2512
2513		Fixed my broken div/mod logic.  Unfortunately, this means that the
2514		inline code for div/mod now has one more comparison than before.
2515		Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2516		The problem was with quotient = 0, because that alone does not tell
2517		us which way the rounding went.  One then has to look at whether
2518		remainder and divisor have the same sign...  :(
2519
2520		Anyway, I replaced the bootfiles with fresh ones...
2521
2522		----------------------------------------------------------------------
2523		Name: Matthias Blume
2524		Date: 2002/03/29 14:10:00 EST
2525		Tag: blume-20020329-inlprims
2526		Description:
2527
2528		NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2529
2530		Primops have changed. This means that the bin/boot-file formats have
2531		changed as well.
2532
2533		To make sure that there is no confusion, I made a new version.
2534
2535
2536		CHANGES:
2537
2538		* removed REMT from mltree (remainder should never overflow).
2539
2540		* added primops to deal with divisions of all flavors to the frontend
2541
2542		* handled these primops all the way through so they map to their respective
2543		MLRISC support
2544
2545		* used these primops in the implementation of Int, Int32, Word, Word32
2546
2547		* removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2548
2549		* parameterized INLMIN, INLMAX, and INLABS by a numkind
2550
2551		* translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2552		floating point
2553
2554		* used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2555		and Real (but Real.abs maps to a separate floating-point-only primop)
2556
2557
2558		TODO items:
2559
2560		* Hacked Alpha32 instruction selection, disabling the selection of REMx
2561		instructions because the machine instruction encoder cannot handle
2562		them.  (Hppa, PPC, and Sparc instruction selection did not handle
2563		REM in the first place, and REM is supported by the x86 machine coder.)
2564
2565		* Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2566		selection phase.  (The two can be streamlined because the hardware
2567		delivers both quotient and remainder at the same time anyway.)
2568
2569		* Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2570		(Currently the behavior is inconsistent both across architectures and
2571		wrt. the draft Basis spec.)
2572
2573		* Word8 should eventually be handled natively, too.
2574
2575		* There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2576		as if there currently is no execution path that could trigger it in
2577		SML/NJ.  (The assumptions underlying functions arith and promotable do not
2578		hold for things like multiplication and division.)
2579
2580		----------------------------------------------------------------------
2581		Name: Matthias Blume
2582		Date: 2002/03/27 16:27:00 EST
2583		Tag: blume-20020327-mlrisc-divisions
2584		Description:
2585
2586		Added support for all four division operations (ML's div, mod, quot,
2587		and rem) to MLRISC.  In the course of doing so, I also rationalized
2588		the naming (no more annoying switch-around of DIV and QUOT), by
2589		parameterizing the operation by div_rounding_mode (which can be either
2590		DIV_TO_ZERO or DIV_TO_NEGINF).
2591
2592		The generic MLTreeGen functor takes care of compiling all four
2593		operations down to only round-to-zero div.
2594
2595		Missing pieces:
2596
2597		* Doing something smarter than relying on MLTreeGen on architectures
2598		like, e.g., the x86 where hardware division delivers both quotient and
2599		remainder at the same time.  With this, the implementation of the
2600		round-to-neginf operations could be further streamlined.
2601
2602		* Remove inlining support for div/mod/rem from the frontend and replace it
2603		with primops that get carried through to the backend.  Do this for all
2604		int and word types.
2605
2606		----------------------------------------------------------------------
2607		Name: Matthias Blume
2608		Date: 2002/03/25 17:25:00 EST
2609		Tag: blume-20020325-divmod
2610		Description:
2611
2612		I improved (hopefully without breaking them) the implementation of Int.div,
2613		Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2614		advantage of the following observations:
2615
2616		Let  q = x quot y      r = x rem y
2617		d = x div  y      m = x mod y
2618
2619		where "quot" is the round-to-zero version of integer division that
2620		hardware usually provides.  Then we have:
2621
2622		r = x - q * y        where neither the * nor the - will overflow
2623		d = if q >= 0 orelse x = q * y then q else q - 1
2624		where neither the * nor the - will overflow
2625		m = if q >= 0 orelse r = 0 then r else r + y
2626		where the + will not overflow
2627
2628		This results in substantial simplification of the generated code.
2629		The following table shows the number of CFG nodes and edges generated
2630		for
2631		fun f (x, y) = x OPER y
2632		(* with OPER \in div, mod, quot, rem *)
2633
2634
2635		OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2636		--------------------------------------------------------
2637		div |         24 |         39 |         12 |         16
2638		mod |         41 |         71 |         12 |         16
2639		quot |          8 |         10 |          8 |         10
2640		rem |         10 |         14 |          8 |         10
2641
2642
2643		----------------------------------------------------------------------
2644		Name: Matthias Blume
2645		Date: 2002/03/25 22:06:00 EST
2646		Tag: blume-20020325-cprotobug
2647		Description:
2648
2649		Fixed a bug in cproto (c prototype decoder).
2650
2651		----------------------------------------------------------------------
2652		Name: Matthias Blume
2653		Date: 2002/03/25 16:00:00 EST
2654		Tag: blume-20020325-raw-primops
2655		Description:
2656
2657		I did some cleanup to Allen's new primop code and
2658		replaced yesterday's bootfiles with new ones.
2659		(But they are stored in the same place.)
2660
2661		----------------------------------------------------------------------
2662		Name: Matthias Blume
2663		Date: 2002/03/24 22:40:00 EST
2664		Tag: blume-20020324-bootfiles
2665		Description:
2666
2667		Made the bootfiles that Allen asked for.
2668
2669		----------------------------------------------------------------------
2670		Name: Allen Leung
2671		Date: 2002/03/23 15:50:00 EST
2672		Tag: leunga-20020323-flint-cps-rcc-primops
2673		Description:
2674
2675		1. Changes to FLINT primops:
2676
2677		(* make a call to a C-function;
2678		* The primop carries C function prototype information and specifies
2679		* which of its (ML-) arguments are floating point. C prototype
2680		* information is for use by the backend, ML information is for
2681		* use by the CPS converter. *)
2682		| RAW_CCALL of { c_proto: CTypes.c_proto,
2683		ml_args: ccall_type list,
2684		ml_res_opt: ccall_type option,
2685		reentrant : bool
2686		} option
2687		(* Allocate uninitialized storage on the heap.
2688		* The record is meant to hold short-lived C objects, i.e., they
2689		* are not ML pointers.  With the tag, the representation is
2690		* the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2691		*)
2692		| RAW_RECORD of {tag:bool,sz:int}
2693		and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2694
2695		2.  These CPS primops are now overloaded:
2696
2697		rawload of {kind:numkind}
2698		rawstore of {kind:numkind}
2699
2700		The one argument form is:
2701
2702		rawload {kind} address
2703
2704		The two argument form is:
2705
2706		rawload {kind} [ml object, byte-offset]
2707
2708		3. RAW_CCALL/RCC now takes two extra arguments:
2709
2710		a. The first is whether the C call is reentrant, i.e., whether
2711		ML state should be saved and restored.
2712		b. The second argument is a string argument specifying the name of
2713		library and the C function.
2714
2715		These things are currently not handled in the code generator, yet.
2716
2717		4. In CProto,
2718
2719		An encoding type of "bool" means "ml object" and is mapped into
2720		C prototype of PTR.  Note that "bool" is different than "string",
2721		even though "string" is also mapped into PTR, because "bool"
2722		is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2723
2724		5. Pickler/unpicker
2725
2726		Changed to handle RAW_RECORD and newest RAW_CCALL
2727
2728		6. MLRiscGen,
2729
2730		1. Changed to handle the new rawload/rawstore/rawrecord operators.
2731		2. Code for handling C Calls has been moved to a new module CPSCCalls,
2732		in the file CodeGen/cpscompile/cps-c-calls.sml
2733
2734		7. Added the conditional move operator
2735
2736		condmove of branch
2737
2738		to cps.  Generation of this is still buggy so it is currently
2739		disabled.
2740
2741		----------------------------------------------------------------------
2742	Name: Lal George	Name: Lal George
2743	Date: 2002/03/22 14:18:25 EST	Date: 2002/03/22 14:18:25 EST
2744	Tag: george-20020322-cps-branch-prob	Tag: george-20020322-cps-branch-prob

---

Colored Diff Long Colored Diff Unidiff Context Diff Side by Side

Legend: Removed from v.1171   changed lines   Added in v.1831

---

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