Diff of /sml/trunk/HISTORY

Parent Directory | Revision Log | Patch

	revision 1446, Mon Jan 26 17:23:06 2004 UTC	revision 1701, Wed Nov 17 21:07:35 2004 UTC
#	Line 13	Line 13
13	Description:	Description:
14
15	----------------------------------------------------------------------	----------------------------------------------------------------------
16		Name: Allen Leung (leunga (at) reservoir (dot) com)
17		Date: 2004/11/17 16:05:21 EST 2004
18		Tag: leunga-20041117-mlrisc-live-kill
19		Description:
20
21		Added support for MLTree constructs LIVE and KILL
22		to all the architectures.
23
24		----------------------------------------------------------------------
25		Name: Matthias Blume (blume (at) tti - c (dot) org)
26		Date: 2004/11/13 00:20:00 CST
27		Tag: blume-20041113-versiontool
28		Description:
29
30		- Stripped down the versiontool: It now only handles the version number.
31		The date string is generated at bootstrap time (during makeml).
32
33		- In a previous commit, fixed a minor issue with how polyequal is being
34		translated.  In particular, the code now "looks through" abstractions.
35		This results in slightly fewer polyEqual warnings and hopefully slightly
36		more efficient code.  Important examples for where this matters are
37		the new int64 and word64 types.
38
39		----------------------------------------------------------------------
40		Name: Matthias Blume (blume (at) tti - c (dot) org)
41		Date: 2004/11/12 00:30:00 CST
42		Tag: blume-20041112-int64
43		Description:
44
45		Structure Int64 fully hooked in.  (The implementation is not very
46		efficient, though.)
47
48		----------------------------------------------------------------------
49		Name: Matthias Blume (blume (at) tti - c (dot) org)
50		Date: 2004/11/11 17:30:00 CST
51		Tag: blume-20041111-more64
52		Description:
53
54		All the pieces of Word64 are now there, with the exception of the
55		conversions from and to LargeWord. (Eventually these need to be identities,
56		but for the time being they don't even make sense because LargeWord is
57		32-bit wide.)
58
59		Also started to add similar support for Int64, but major pieces of that
60		are still missing.
61
62		----------------------------------------------------------------------
63		Name: Matthias Blume (blume (at) tti - c (dot) org)
64		Date: 2004/11/11 00:15:00 CST
65		Tag: blume-20041111-word64
66		Description:
67
68		Structure Word64 is now (almost) complete, word literals and patterns
69		seem to work.  There are a few odd pieces missing.  In particular,
70		I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
71		at the moment.
72
73		Making Word64 official would mean that LargeWord becomes Word64.  But
74		this requires extreme care because most word-word conversions have to
75		go through LargeWord, so making a mistake means loss of efficiency or
76		worse.  Eventually there will be a solution similar to (but actually
77		simpler than) what I did with IntInf.
78
79		----------------------------------------------------------------------
80		Name: Matthias Blume (blume (at) tti - c (dot) org)
81		Date: 2004/11/10 18:12:00 CST
82		Tag: blume-20041110-64bit
83		Description:
84
85		More 64-bit hacking (but still not even half-way there yet).
86		Also, some assorted improvements to the handling of 8-bit words.
87
88		----------------------------------------------------------------------
89		Name: Matthias Blume (blume (at) tti - c (dot) org)
90		Date: 2004/11/09 17:50:00 CST
91		Tag: <none>
92		Description:
93
94		Started adding some infrastructure for supporting 64-bit int- and
95		word-types.  (Still in its very early stages.)
96
97		----------------------------------------------------------------------
98		Name: Matthias Blume (blume (at) tti - c (dot) org)
99		Date: 2004/10/28 10:45:00 CDT
100		Tag: Release_110_50
101		Description:
102
103		New working version (110.50).  NEW BOOTFILES!
104
105		=====================
106
107		Also:
108
109		- Changend config/srcarchiveurl from a file just
110		containing the URL string into a file containing
111		shell script code.  The code has access to the $VERSION variable.
112		- Made corresponding changes to config/install.sh and config/unpack.
113		- Default contents of config/srcarchiveurl uses $VERSION and
114		normally does not have to be edited to reflect a version change.
115
116		(As a result, a version change can be done by just editing
117		config/version, the rest is now automatic.)
118
119		----------------------------------------------------------------------
120		Name: Matthias Blume (blume (at) tti - c (dot) org)
121		Date: 2004/10/27 17:50:00 CDT
122		Tag: blume-20041027-btrace-msg
123		Description:
124
125		BackTrace.monitor now also reports the source of the exception that
126		triggered the trace.
127
128		----------------------------------------------------------------------
129		Name: Matthias Blume (blume (at) tti - c (dot) org)
130		Date: 2004/10/27 17:20:00 CDT
131		Tag: blume-20041027-x86-c-calls
132		Description:
133
134		This is the HISTORY entry for two earlier commits, both concerning
135		the x86 c-calls code in MLRISC:
136
137		- added a missing LOAD in the code that deals with struct arguments
138		- made sure the caller does not add the wrong number of bytes to the
139		stack pointer after a call of a function returning a struct
140		(the callee already pops the implicit argument which points to
141		the space reserved for the result)
142
143		----------------------------------------------------------------------
144		Name: Allen Leung (leunga (at) reservoir (dot) com)
145		Date: 2004/10/24 14:00:00 EST
146		Tag: leunga-20041024-x86-gas-fucomip
147		Description:
148
149		John discovered a bug in the syntax of fucomip.
150		The opcodes FU?COMIP? have been changed to
151
152		fu?comip? %st(i), %st
153
154		----------------------------------------------------------------------
155		Name: Matthias Blume (blume (at) tti - c (dot) org)
156		Date: 2004/10/20 15:06:00 CDT
157		Tag: blume-20041020-standalone-backtrace
158		Description:
159
160		Added a mechanism for getting back-trace information from standalone
161		programs.  Here is how it works:
162
163		1. The part of the program from which you want to get backtrace
164		information (usually the whole program) should be wrapped with
165		BackTrace.monitor.  This is a (unit->'a)->'a function, and your
166		main program could be modified from something like
167
168		fun main (pgm, args) = ...
169
170		to
171
172		fun main (pgm, args) = BackTrace.monitor (fn () => ...)
173
174		2. To be able to access BackTrace.monitor, you have to add
175		library $smlnj-tdp/plugins.cm to the .cm file that contains your
176		main function.
177
178		3. Remove all compiled code (i.e., all the .cm/ subdirectories that
179		CM might have created in the past for your project).
180
181		4. Build the system using this command line:
182
183		ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
184		myprog.cm MyProg.main myprog
185
186		instead of the usual
187
188		ml-build myprog.cm MyProg.main myprog
189
190		I changed a library name:
191
192		$/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
193
194		New libraries:
195
196		$smlnj-tdp/back-trace.cm
197		-- when loaded causes the back-trace plugin to be installed
198		$smlnj-tdp/coverage.cm
199		-- when loaded causes the coverage plugin to be installed
200
201		----------------------------------------------------------------------
202		Name: Matthias Blume (blume (at) tti - c (dot) org)
203		Date: 2004/10/18 16:45:00 CDT
204		Tag: blume-20041018-groupowner
205		Description:
206
207		Added an "obsolete" warning for the "group owner" syntax to CM's parser.
208
209		Eliminated group owner specs from .cm files throughout the source tree.
210
211		----------------------------------------------------------------------
212		Name: Matthias Blume (blume (at) tti - c (dot) org)
213		Date: 2004/10/15 15:45:00 CDT
214		Tag: blume-20041015-coverage
215		Description:
216
217		* Test coverage tool added!
218		* Further reorganization of tracing-, debugging-, and profiling support:
219
220		- moved original BTImp -- now called BackTrace -- into a separate
221		library called $/trace-debug-profile.cm
222		- eliminated all mentions of BTrace from SMLofNJ.Internals
223		- only the instrumentation mechanism is now left in the compiler proper
224		- BackTrace module is a plugin which is NOT plugged in by default
225		- Coverage module is another such plugin
226
227		To get the benefits of any of these plugin modules, the code in
228		question must be compiled with tdp instrumentation turned on.  This
229		can be done by setting SMLofNJ.Internals.TDP.mode to true.
230		(The ref cell is also controlled via the -Ctdp.instrument=... switch.)
231
232		Plugins are selected at link time.  (Pre-compiled instrumented code
233		can be re-loaded with different plugins in effect.)  When an
234		instrumented module is linked, whatever plugins are at that time
235		enabled will come into effect for that module.
236
237		To enable the back-trace plugin, load library $/trace-debug-profile.cm
238		and invoke BackTrace.install() (e.g., from the interactive prompt).
239		To enable the coverage plugin, load the same library and invoke
240		Coverage.install().
241
242		Back-traces are generated automatically on uncaught exceptions and
243		when the code in question explicitly invokes BackTrace.trigger().
244
245		Coverage (and execution frequency-) information must be queried
246		explicitly by calling Coverage.not_covered and Coverage.hot_spots.
247
248		----------------------------------------------------------------------
249		Name: Matthias Blume (blume (at) tti - c (dot) org)
250		Date: 2004/10/14 17:40:00 CDT
251		Tag: blume-20041014-tdp-core
252		Description:
253
254		Snapshot of a significant overhaul of how the trace/debug/profile support
255		is hooked into the system (specifically: Core and SMLofNJ.Internals).
256
257		----------------------------------------------------------------------
258		Name: Matthias Blume (blume (at) tti - c (dot) org)
259		Date: 2004/10/13 16:34:00 CDT
260		Tag: blume-20041013-tdp
261		Description:
262
263		Some rationalization of names:
264
265		structure BTrace -> structure TDPInstrument
266		etc.
267
268		This is is preparation of using the original back-trace
269		instrumentation for other purposes.  "TDP" stands for
270		Trace/Debug/Profile.
271
272		The control flag controlling whether instrumentation is on or off is now
273		registered under a different name, so instead of running sml as
274
275		sml -Cinstrument.btrace-mode=true
276
277		one has to say
278
279		sml -Ctdp.instrument=true
280
281		----------------------------------------------------------------------
282		Name: Matthias Blume (blume (at) tti - c (dot) org)
283		Date: 2004/10/11 16:37:00 CDT
284		Tag: blume-20041011-regions
285		Description:
286
287		Made some minor modifications to elabcore.sml to have source regions
288		be propagated more tightly -- resulting in better (i.e., smaller)
289		regions being reported in error- and debug messages.
290
291		----------------------------------------------------------------------
292		Name: Matthias Blume (blume (at) tti - c (dot) org)
293		Date: 2004/10/08 22:50:00 CDT
294		Tag: blume-20041008-cmkw
295		Description:
296
297		Fixed handling of keywords in .cm files: After seeing "is" the lexer
298		treats subsequent occurrences of "group", "library", "source", "is",
299		"*", and "-" as ordinary identifiers rather than keywords.
300
301		Most seriously, this fixes a problem with CM's "shell" tool.  The tool
302		is supposed to accept a tool argument called "source", but this did
303		not work because of the clash with the keyword.
304
305		----------------------------------------------------------------------
306		Name: Matthias Blume (blume (at) tti - c (dot) org)
307		Date: 2004/10/07 16:00:00 CDT
308		Tag: blume-20041007-cleanup
309		Description:
310
311		Assorted cleanup work:
312
313		- got rid of intstrmap in favor of using the library's
314		hash table implementation
315		- threw out most of the pathnames stuff, as it was not used anyway
316		- simplified tokentable implementation
317		- fixed some minor spelling errors
318
319		----------------------------------------------------------------------
320		Name: Matthias Blume (blume (at) tti - c (dot) org)
321		Date: 2004/10/06 15:15:15 CDT
322		Tag: blume-20041006-handler
323		Description:
324
325		Cleaned up the absyn to reflect the invariant that HANDLE always
326		carries a FNexp as part of the type definition.  This eliminates some
327		superfluous sanity checks at runtime down the road.
328
329		Some minor cleanup of the btrace code.
330
331		----------------------------------------------------------------------
332		Name: Matthias Blume (blume (at) tti - c (dot) org)
333		Date: 2004/10/01 10:20:30 CDT
334		Tag: blume-20041001-slave
335		Description:
336
337		Added hack to make slave mode work in the presence of the version
338		tool.  (Still, since the master does two passes over the code for
339		CMB.make, the release number gets bumped twice when slaves are
340		attached.  I don't know if this is worth fixing...)
341
342		----------------------------------------------------------------------
343		Name: Matthias Blume (blume (at) tti - c (dot) org)
344		Date: 2004/09/30 10:55:00 CDT
345		Tag: blume-20040930-version
346		Description:
347
348		* Moved the "version" magic into its own little library under
349		src/system/smlnj/internal.  This avoids expensive reconstruction of
350		a stable src/compiler/core.cm.
351
352		* At the same time, structure CompilerVersion is now known as
353		structure SMLNJVersion.
354
355		* Arranged for the version tool to NOT kick in when rebuilding the system
356		(makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
357		Also, loading the versiontool does not work when rebuilding the system
358		because CM is not properly initialized at that time.
359
360		----------------------------------------------------------------------
361		Name: Matthias Blume (blume (at) tti - c (dot) org)
362		Date: 2004/09/29 14:00:00 CDT
363		Tag: blume-20040929-autoversion
364		Description:
365
366		Implemented some CM magic to have
367		file src/compiler/TopLevel/main/version.sml
368		generated automagically.
369		The version is taken from two files: config/version and config/release.
370		The first is expected to contain a two-part version number such as 110.49.
371		The second should contain a single number, but it may be missing.
372
373		If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
374		time the version tool is loaded (which is the first time you say CMB.make),
375		then the tool will increment the value stored in config/release every
376		time CMB.make is invoked.
377
378		The binfile format is now insensitive to anything beyond the first
379		two components of a version number, so bumping the release does not render
380		binfiles incompatible.  Auto-bumping can be used to keep track of versions
381		during development without invalidating existing binfiles.
382
383		In any case, every CMB.make updates the date information in version.sml.
384		(This is the date that is printed in the banner.)
385
386		----------------------------------------------------------------------
387		Name: Matthias Blume (blume (at) tti - c (dot) org)
388		Date: 2004/09/28 10:53:00 CDT
389		Tag: blume-20040928-controls
390		Description:
391
392		Some cleanup of the controls code.
393
394		----------------------------------------------------------------------
395		Name: Matthias Blume (blume (at) tti - c (dot) org)
396		Date: 2004/09/27 22:08:00 CDT
397		Tag: blume-20040927-controls
398		Description:
399
400		Added two pieces of functionality to the Controls interface:
401
402		1.   val save'restore: 'a control -> unit -> unit
403
404		grabs the current value of the control in stage 1 and restores it
405		in stage 2.
406
407		2.   val set' : 'a control * 'a -> unit -> unit
408
409		stores the given value into the control in stage 2 (i.e., delayed)
410		but does all error checking in stage 1.
411		(This is for string controls that need to do parse their argument --
412		something that might fail.  In some cases, notably in CM, one
413		already knows the intended argument but wants to delay the actual
414		assignment until a time when error recovery would be more difficult.)
415
416		Changed the handling of controls in tool arguments to classes "sml" and
417		"lazysml":
418		- use Controls.save'restore as a more robust way of restoring the
419		old value (in particular: without having to re-parse the string)
420		- use controls to handle the "overload" keyword in the init group
421		(I believe this change actually fixes a long-standing obscure bug.)
422
423		----------------------------------------------------------------------
424		Name: Matthias Blume (blume (at) tti - c (dot) org)
425		Date: 2004/09/27 17:00:00 CDT
426		Tag: blume-20040927-lazysml
427		Description:
428
429		Added a new tool class called "lazysml" to CM's tool chest.  The only
430		difference to "sml" is that compilation is done with Control.lazysml
431		set to true.  A source of class "lazysml" is automatically recognized
432		by a file name suffix of ".lml".
433
434		In addition to the above feature, the original class "sml" now also
435		supports a tool argument "lazy" which has the same effect.  As a
436		result, the following three lines are equivalent:
437
438		foo.sml : lazysml
439		foo.sml : sml (lazy)
440		foo.sml (lazy)
441
442		The setting goes into effect both during parsing and during
443		compilation.  The original setting is restored right after parsing and
444		after compilation, respectively.
445
446		In addition to all the above, there is also a general mechanism to set
447		ANY of the "controls" that are available at the command line via
448		"-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
449		apply as well. (In fact, "lazy" is implemented as a special case of
450		the general mechanism.)
451
452		The .cm file syntax uses a new keyword tool argument called "with".
453		There are several ways of indicating the desired settings:
454
455		foo.sml (with:parser.quotations=true)
456		foo.sml (with:(name:parser.quotations value:true))
457		foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
458		foo.sml (with:(name1=value1 name2=value2 ...))
459		foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
460
461		etc.
462
463		Another possible abbreviation is to leave out the =v or value:v part
464		if the name refers to a boolean control (in which case the value is
465		taken to be true).  Thus, one could get lazy sml also by saying:
466
467		foo.sml (with:parser.lazy-keyword=true)
468		foo.sml (with:parser.lazy-keyword)
469		foo.sml (with:(name:parser.lazy-keyword value:true))
470		foo.sml (with:(name:parser.lazy-keyword))
471
472		----------------------------------------------------------------------
473		Name: Matthias Blume (blume (at) tti - c (dot) org)
474		Date: 2004/09/24 16:22:00 CDT
475		Tag: blume-20040924-ppc-long-branch
476		Description:
477
478		Turned message about "emiting long form of branch" off by default.
479		Added a control flag to turn it back on when desired.
480
481		----------------------------------------------------------------------
482		Name: Matthias Blume (blume (at) tti - c (dot) org)
483		Date: 2004/09/24 16:05:00 CDT
484		Tag: blume-20040924-rounding
485		Description:
486
487		Applied patch for setting rounding modes under Mac OS X.  Thanks to
488		Melissa O'Neill for providing the code!
489
490		----------------------------------------------------------------------
491		Name: Matthias Blume (blume (at) tti - c (dot) org)
492		Date: 2004/09/23 17:30:00 CDT
493		Tag: blume-20040923-envvars
494		Description:
495
496		1. Changed definition of type ControlRegistry.registry_tree to
497		include control_info (i.e., the name of the controlling environment
498		variable).
499
500		2. Added command-line flags -e and -E to print the names of environment
501		variables that can be used to control internal settings.
502		(This uses the new API mentioned in 1.)
503
504		----------------------------------------------------------------------
505		Name: Matthias Blume (blume (at) tti - c (dot) org)
506		Date: 2004/09/13 16:50:00 CDT
507		Tag: Release_110_49
508		Description:
509
510		New working version (110.49).  NEW BOOTFILES!
511
512		----------------------------------------------------------------------
513		Name: Matthias Blume
514		Date: 2004/09/13 16:20:00 CDT
515		Tag: blume-20040913-config-mlrisc
516		Description:
517
518		Put target "mlrisc" back into the default list.
519		(There is no harm in having it, and some users have expressed their
520		wish to have "mlrisc" included by default.)
521
522		----------------------------------------------------------------------
523		Name: John Reppy
524		Date: 2004/09/13
525		Tag: jhr-20040913-signals
526		Description:
527
528		Fixed the signal masking code to properly nest mask/unmask operations
529		on a per-signal basis.
530
531		----------------------------------------------------------------------
532		Name: Matthias Blume (blume (at) tti - c (dot) org)
533		Date: 2004/09/08 13:20:00 CDT
534		Tag: blume-20040908-heap-magic
535		Description:
536
537		Bumped the heap macig number to 0x09082004 to account for the changed
538		layout of the ML frame under MacOS X.
539
540		----------------------------------------------------------------------
541		Name: Allen Leung (leunga (at) reservoir (dot) com)
542		Date: 2004/09/03 11:26:00 EST
543		Tag: leunga-20040903-cygwin-install
544		Description:
545
546		Added a patch to _arch-n-opsys to enable the Cygwin runtime.
547		The Cygwin runtime is turned on by setting the environment
548		variable SMLNJ_CYGWIN_RUNTIME to 1.
549
550		----------------------------------------------------------------------
551		Name: Matthias Blume (blume (at) tti - c (dot) org)
552		Date: 2004/08/31 17:15:00 CDT
553		Tag: blume-20040831-core
554		Description:
555
556		Added some exports to src/compiler/core.cm upon request by J. Joemann.
557
558		----------------------------------------------------------------------
559		Name: Matthias Blume (blume (at) tti - c (dot) org)
560		Date: 2004/08/30 17:55:00 CDT
561		Tag: blume-20040830-installer
562		Description:
563
564		Upon request by Johannes Joemann:
565
566		- improved ML code of installer to fall back to coping when renaming
567		fails (i.e., when source and target are on different file systems);
568		the code compiles but has yet to be tested in anger
569		- removed mlrisc from list of default targets (config/targets)
570
571		----------------------------------------------------------------------
572		Name: Matthias Blume (blume (at) tti - c (dot) org)
573		Date: 2004/08/27 17:20:00 CDT
574		Tag: blume-20040827-ptreql
575		Description:
576
577		Added ptreql primop to structure InlineT (upon request from Larry
578		Paulson).
579
580		----------------------------------------------------------------------
581		Name: Allen Leung (leunga (at) reservoir (dot) com)
582		Date: 2004/08/15 21:21:00 EST
583		Tag: leunga-110_48-udgraph
584		Description:
585
586		Another bug fix from Carl Hauser:
587
588		diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
589		> 48c48
590		> <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
591		> e::L)
592		> ---
593		> >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
594		> Without this, any deletion of an edge in an undirected graph does severe
595		> violence to the graph.
596
597		----------------------------------------------------------------------
598		Name: Allen Leung (leunga (at) reservoir (dot) com)
599		Date: 2004/08/10 23:35:00 EST
600		Tag: leunga-110_48-ppc
601		Description:
602
603		The IBM/MacOS syntax switch on PPC was incorrectly swapped.
604
605		----------------------------------------------------------------------
606		Name: Matthias Blume (blume (at) tti - c (dot) org)
607		Date: 2004/08/10 12:00:00 CDT
608		Tag: Release_110_48
609		Description:
610
611		New working version (110.48).  NEW BOOTFILES!
612
613		----------------------------------------------------------------------
614		Name: Allen Leung (leunga (at) reservoir (dot) com)
615		Date: 2004/08/09 12:21:00 EST
616		Tag: leunga-110_47-dijsktra
617		Description:
618
619		Bug fix from Carl Hauser:
620
621		single_source_shortest_paths in dijkstra.sml was observed to get wrong
622		answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
623
624		The problem is that following the expression A.update(dist,s,Num.zero)
625		it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
626
627		----------------------------------------------------------------------
628		Name: Matthias Blume (blume (at) tti - c (dot) org)
629		Date: 2004/08/06 18:10:00 CDT
630		Tag: blume-20040806-cmdline
631		Description:
632
633		Fiddled with handling of command-line options:
634
635		* sml now quits after processing the command line
636		if -H, -S, -h<n>, or -s<n> appears as the last
637		command-line argument
638		* a new option -q terminates the session when encountered on
639		the command line; subsequent arguments will be ignored
640		* bug fixes: short (erroneous) arguments are no longer ignored
641		completely
642
643		----------------------------------------------------------------------
644		Name: Allen Leung (leunga (at) reservoir (dot) com)
645		Date: 2004/08/04 18:17:00 EST
646		Tag: leunga-110_47-ppc-ibm-asm
647		Description:
648
649		- Added minimal IBM assembly syntax support for PowerPC.
650
651		- Cygwin: manually changed the file cygwin.def.  Some exported symbols have
652		been altered in the runtime.  We need an automatic way to keep the file
653		in sync.
654
655		----------------------------------------------------------------------
656		Name: Matthias Blume (blume (at) tti - c (dot) org)
657		Date: 2004/08/04 14:00:00 CDT
658		Tag: Release_110_47
659		Description:
660
661		New working version (110.47).  NEW BOOTFILES!
662
663		----------------------------------------------------------------------
664		Name: Matthias Blume (blume (at) tti - c (dot) org)
665		Date: 2004/08/03 14:25:00 CDT
666		Tag: blume-20040803-callingconv
667		Description:
668
669		Added low-level support for choosing C calling conventions by
670		twiddling the type of rawccall.  (See
671		src/compiler/Semant/types/cproto.sml for details.)
672
673		----------------------------------------------------------------------
674		Name: Matthias Blume (blume (at) tti - c (dot) org)
675		Date: 2004/08/02 15:55:00 CDT
676		Tag: blume-20040802-backout
677		Description:
678
679		Backed out of change to win32-filesys.c.  The earlier patch to
680		get_file_time caused CM to produce files with the wrong time stamp.
681
682		----------------------------------------------------------------------
683		Name: Matthias Blume (blume (at) tti - c (dot) org)
684		Date: 2004/08/02 14:45:00 CDT
685		Tag: blume-20040802-nlffi-win32
686		Description:
687
688		Added NLFFI support for Win32, adapted from a patch provided by David
689		Hansel.  This is currently completely untested.  Also, the issue
690		concerning stdcall vs. ccall is still unresolved.
691
692		----------------------------------------------------------------------
693		Name: Matthias Blume (blume (at) tti - c (dot) org)
694		Date: 2004/07/30 17:55:00 CDT
695		Tag: blume-20040730-various
696		Description:
697
698		Gearing up towards 110.47...
699
700		- various minor bugfixes to ml-nlffigen
701		- a beginning of a manual for nlffi
702
703		- eliminated 'export name=value' in config/install.sh as this does
704		not work with certain versions of /bin/sh
705		(Thanks to David King at Motorola for catching this.)
706
707		- several bugfixes provided or suggested by David Hansel at Reactive Systems:
708		- added a test for tm==NULL to gmtime.c and localtime.c
709		- applied patch for incorrect GetFileTime under win32
710		- toSeconds -> toMilliseconds in Win32/win32-process.sml
711
712		----------------------------------------------------------------------
713		Name: Matthias Blume (blume (at) tti - c (dot) org)
714		Date: 2004/07/21 18:20:00 CDT
715		Tag: blume-20040721-nlffigen
716		Description:
717
718		- Fixed minor issue in ml-nlffigen: Now generate structure T_foo
719		for a typedef to an incomplete type, but leave out the "typ" member.
720		(This is just for consistency.)
721		- Started to produce what is supposed to become better (i.e., comprehensive)
722		documentation of what ml-nlffigen does and produces.
723
724		----------------------------------------------------------------------
725		Name: Matthias Blume (blume (at) tti - c (dot) org)
726		Date: 2004/07/14 16:25:00 CDT
727		Tag: blume-20040714-union
728		Description:
729
730		Added C_UNION to c-calls/c-types.sml and updated the machinery
731		(ml-nlffigen, cproto.sml) that conveys C function interface
732		information to the code generator.
733
734		However, the actual architecture-specific implementation of function
735		arguments and results that are C unions is still not implemented.
736
737		----------------------------------------------------------------------
738		Name: Allen Leung (leunga (at) reservoir (dot) com)
739		Date: 2004/07/14 14:38:00 EST
740		Tag: leunga-110_46_1-ppc-lwzu
741		Description:
742
743		Added these instructions to the PowerPC architecture:
744		LBZU(X), LHZU(X), LWZU(X),
745		STWU(X), STFDU, STFSU
746
747		etc...
748
749		Note: I haven't added their instruction encoding into the description.
750
751		----------------------------------------------------------------------
752		Name: Allen Leung (leunga (at) reservoir (dot) com)
753		Date: 2004/07/13 15:04:00 EST
754		Tag: leunga-110_46_1-ppc-lwarx
755		Description:
756
757		Added the two instructions LWARX and STWCX to the PowerPC
758		instruction set.
759
760		A (untested) rewrite of loop-structure.sml.  The old version
761		is completely broken.
762
763		----------------------------------------------------------------------
764		Name: Matthias Blume (blume (at) tti - c (dot) org)
765		Date: 2004/07/13 13:50:00 CDT
766		Tag: blume-20040713-nlffi
767		Description:
768
769		- use paramAlloc to report c-calls with too many arguments
770		(for PPC version where parameter area is pre-allocated)
771		- added ccall_maxargspace to machspec (to implement the above)
772		- made "make" commend in CM's "make" tool configurable
773		- added option (default: on) for passing the name of the SML/NJ's "bin"
774		directory to "make"; the call looks like this:
775
776		make <options> SMLNJ_BINDIR=<dir> <target>
777
778		This can be used by the Makefile to, e.g., pick the "right" version
779		of ml-nlffigen.
780		- minor code tweaks
781
782		----------------------------------------------------------------------
783		Name: Matthias Blume (blume (at) tti - c (dot) org)
784		Date: 2004/07/12 22:50:00 CDT
785		Tag: blume-110_46_1-macosx-nlffi
786		Description:
787
788		NLFFI under Mac OS X now working (sort of).  This is largely untested,
789		though.
790
791		Note:  1. You have to make a new, clean build of the runtime system.
792		2. There are new BOOTFILES, you have to use them!
793		(Doing the bootstrap process yourself would be *very* painful!
794		If you absolutely have to do it, build the system under
795		a different architecture and then cross-compile.)
796
797		Version bumped to 110.46.1 to account for runtime data format changes.
798
799		----------------------------------------------------------------------
800		Name: Matthias Blume (blume (at) tti - c (dot) org)
801		Date: 2004/06/18 14:30:00 CDT
802		Tag: blume-20040618-unix
803		Description:
804
805		Changed the implementation of structure Unix so that the same stream
806		is returned every time one of the {text,bin}{In,Out}streamOf functions
807		is invoked on the same proc.  This is not what the spec currently
808		says -- although IMO it arguably should.  (See discussion below.)
809
810		----------------------------------------------------------------------
811		Name: Matthias Blume (blume (at) tti - c (dot) org)
812		Date: 2004/06/17 18:15:00 CDT
813		Tag: Release_110_46
814		Description:
815
816		New working version (110.46).  NEW BOOTFILES!
817
818		----------------------------------------------------------------------
819		Name: Matthias Blume (blume (at) tti - c (dot) org)
820		Date: 2004/06/17 17:20:00 CDT
821		Tag: blume-20040617-timer-unix
822		Description:
823
824		Changed the interface of structures Timer and Unix to match the most
825		recent Basis spec.
826
827		In the case of Unix there still seems to be an open/weird issue:
828
829		The {text,bin}{In,Out}streamOf functions are supposed to create
830		fresh streams whenever they are called -- as opposed to have them
831		return the same stream every time.  This design is supposed to
832		prevent space leaks caused by proc values hanging on to streams.
833
834		The reap function, on the other hand, is supposed to close the
835		streams.  This cannot be done without having a handle on the
836		stream in proc after all...
837
838		I took the liberty to implement the following stopgap solution:
839
840		The proc value hangs on to the most recently created stream(s).
841		Reap closes those.  If either or both of the two streams hadn't
842		been created at all yet, then reap will close the corresponding
843		file descriptors directly.
844
845		PS: I don't understand the original space leak argument anymore.  If
846		a proc hangs on to the imperative stream, then I/O operations on those
847		will advance the state of the cached stream and avoid the space leak.
848
849		----------------------------------------------------------------------
850		Name: Matthias Blume (blume (at) tti - c (dot) org)
851		Date: 2004/05/28 16:45:00 CDT
852		Tag: blume-20040528-basis
853		Description:
854
855		Added signature PACK_REAL and exported functor PrimIO.
856
857		----------------------------------------------------------------------
858		Name: Matthias Blume (blume (at) tti - c (dot) org)
859		Date: 2004/05/25 16:00:00 CDT
860		Tag: blume-20040525-group-owner
861		Description:
862
863		CM now ignores (but still accepts) the "owner" information in group
864		descriptions.  The owner of a group is its next enclosing
865		library. Each group must have a unique owner.  (There is a virtual
866		"toplevel" library that owns groups which are not nested within a real
867		library.)  Previously, each group had to explicitly declare its owner,
868		and CM would check that such a declaration is correct.  The new scheme
869		is to have CM check that for each group there is precisely one owning
870		library.
871
872		The advantage of the new scheme is that the programmer no longer needs
873		to maintain the somewhat annoying owner information.  The downside is
874		that CM cannot enforce the ownership rule across multiple runs of
875		CM.make.  Fortunately, enclosing the same group in two different
876		libraries A and B which are not part of the same program does not
877		cause real problems.
878
879		----------------------------------------------------------------------
880		Name: Matthias Blume (blume (at) tti - c (dot) org)
881		Date: 2004/05/20 16:00:00 CDT
882		Tag: blume-20040520-win32
883		Description:
884
885		Made the win32 version work again.  (Strangely, a misplaced comma had
886		slipped into win32-process.c which prevented the runtime from being
887		compiled correctly.)
888
889		Also, included a minor addition to ml-build.bat analogous to what was
890		done in blume-20040519-ml-build.
891
892		----------------------------------------------------------------------
893		Name: Matthias Blume (blume (at) tti - c (dot) org)
894		Date: 2004/05/19 22:10:00 CDT
895		Tag: blume-20040519-ml-build
896		Description:
897
898		Arranged for ml-build to clean up after itself a little bit better.
899		The script generates a temporary SML source file and compiles it using
900		CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
901		gets rid of those at the end, so they don't accumulate under .cm.
902
903		This required a minor change to install.sh because the name of the
904		metadata directory (default: .cm) is actually configurable at
905		installation time.
906
907		----------------------------------------------------------------------
908		Name: Matthias Blume (blume (at) tti - c (dot) org)
909		Date: 2004/05/18 15:50:00 CDT
910		Tag: blume-20040518-mkreader
911		Description:
912
913		Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
914		implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
915
916		----------------------------------------------------------------------
917		Name: Matthias Blume (blume (at) tti - c (dot) org)
918		Date: 2004/05/11 14:35:00 CDT
919		Tag: blume-20040511-win32sock
920		Description:
921
922		Added previously missing support for many socket-related functions
923		under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
924		for the voluminous patch!
925
926		(I have not tested this patch under win32 yet.)
927
928		Here is David's e-mail:
929
930		Hi,
931
932		Attached to this email you find a diff against sml/nj 110.45
933		that will enable socket support under Windows.
934
935		To apply the patch (using unix or cygwin)
936		1) gunzip runtime.diff.gz
937		2) "cd" into "src/runtime" in the source tree of a fresh
938		110.45 installation.
939		3) patch -p 1 < [your/path/to]runtime.diff
940
941		The code compiles fine but has NOT yet been extensively tested.
942		I only ran a few tests for basic socket client functionality
943		(which worked fine).  Especially the functions that use ioctl
944		are not tested at all and might not work (see below).
945
946		I implemented this since we want to move to a newer version of sml/nj
947		but need socket support in order to use it.  This is the first time I
948		even had a look at the sml/nj source,  so please review my changes
949		before making this part of the distribution!  Here are a few issues
950		that I think might be better for someone to solve who is more
951		familiar with the sml/nj source (and socket programming):
952
953		- getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
954		exception since I could not figure out what the windows equivalent
955		of these functions is
956
957		- In sockets-osdep.h there are a some #include statements that are
958		only used in a few files that include sockets-osdep.h
959
960		- In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
961		does not process its return value since I don't know how to
962		report an error upwards.
963
964		- It would probably be good to have a call to WSACleanup() when
965		the library is unloaded (if there is such a possibility).
966		Otherwise I think Windows will take care of this automatically
967		when the process finishes.
968
969		- I used ioctlsocket() as a replacement for ioctl() but I have
970		no idea if that is actually the proper replacement on Windows.
971
972		- All these issues are marked in the code by "FIXME" comments.
973
974		We use sml/nj extensively in our products and are quite happy
975		with it.  I hope this contribution will help you.
976
977		Keep up the good work!
978
979		David
980
981		----------------------------------------------------------------------
982		Name: Matthias Blume (blume (at) tti - c (dot) org)
983		Date: 2004/05/11 14:20:00 CDT
984		Tag: blume-20040511-installml
985		Description:
986
987		Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
988		the patch.)
989
990		----------------------------------------------------------------------
991		Name: Matthias Blume (blume (at) tti - c (dot) org)
992		Date: 2004/05/11 14:05:00 CDT
993		Tag: blume-20040511-nlffi-netbsd
994		Description:
995
996		Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
997		the patch.)
998
999		----------------------------------------------------------------------
1000		Name: Matthias Blume (blume (at) tti - c (dot) org)
1001		Date: 2004/05/11 12:05:00 CDT
1002		Tag: blume-20040511-exports
1003		Description:
1004
1005		As per request by Adam Chlipala <adam@hcoop.net>, extended various
1006		export lists in compiler-related .cm-files.
1007
1008		----------------------------------------------------------------------
1009		Name: Matthias Blume (blume (at) tti - c (dot) org)
1010		Date: 2004/05/11 11:35:00 CDT
1011		Tag: blume-20040511-allsource
1012		Description:
1013
1014		The installer now honors the "src-smlnj" target again, although its meaning
1015		has changed from "all sources required for the compiler" to "all sources
1016		the installer knows about".  In other words, if you enable "src-smlnj"
1017		in the "targets" file, then the installer will pull in sources for
1018		everything.  (Notice that this refers to source code only.  Compiled
1019		code is still only installed for modules that were requested explicitly
1020		or which are required for other modules that were requested explicitly.)
1021
1022		----------------------------------------------------------------------
1023		Name: Matthias Blume (blume (at) tti - c (dot) org)
1024		Date: 2004/04/23 17:40:00 CDT
1025		Tag: blume-20040423-ieee-scan
1026		Description:
1027
1028		Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1029		in the exponent calculation we get INF or ZERO (depending on the mantissa
1030		and the sign of the exponent).
1031
1032		----------------------------------------------------------------------
1033		Name: Matthias Blume (blume (at) tti - c (dot) org)
1034		Date: 2004/04/23 10:40:00 CDT
1035		Tag: blume-20040423-ml-build
1036		Description:
1037
1038		The ml-build script now terminates with a non-0 status when something
1039		goes wrong.
1040
1041		----------------------------------------------------------------------
1042		Name: Matthias Blume (blume (at) tti - c (dot) org)
1043		Date: 2004/04/22 16:35:00 CDT
1044		Tag: blume-20040422-Option
1045		Description:
1046
1047		Made exception Option to be the same as exception Option.Option
1048		(as it should be).
1049
1050		----------------------------------------------------------------------
1051		Name: Allen Leung (leunga (at) reservoir (dot) com)
1052		Date: 2004/03/19 14:40:00 EST
1053		Tag: leunga-20040319-cygwin-nlffi
1054		Description:
1055
1056		Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1057		of SML/NJ.  The problem is that
1058
1059		lib = dlopen(NULL, ...)
1060		f   = dlsym(lib, "malloc");
1061
1062		does not work on Windows unless we explicitly export symbols
1063		such as 'malloc' during linking.  We fixed this by explicitly
1064		exporting the required symbols with the magic gcc incantation:
1065
1066		-Wl,--export-all cygwin.def
1067
1068		where cygwin.def is a file containing all the symbols that
1069		we wish to export.
1070
1071		I suspect this is a Windows problem and we'll have to
1072		do the same (somehow with windows compilers) when
1073		we build the native win32 version with the system
1074		calls LoadLibrary/GetProcAddress.
1075
1076		----------------------------------------------------------------------
1077		Name: Matthias Blume (blume (at) tti - c (dot) org)
1078		Date: 2004/03/04 16:35:00 CST
1079		Tag: blume-20040304-intinf-fmt
1080		Description:
1081
1082		Fixed problem with IntInf.fmt (sign would show up on the right instead
1083		of on the left for BIN, OCT, and HEX).
1084
1085		----------------------------------------------------------------------
1086		Name: Matthias Blume (blume (at) tti - c (dot) org)
1087		Date: 2004/03/04 11:25:00 CST
1088		Tag: blume-20040304-symlinks
1089		Description:
1090
1091		Fixed problem with installer script (unix only) where bin/ml-yacc and
1092		friends pointed (via symlinks) to absolute locations instead of just
1093		.run-sml.  This was reported by Vesa A Norrman.
1094
1095		----------------------------------------------------------------------
1096		Name: Matthias Blume (blume (at) tti - c (dot) org)
1097		Date: 2004/02/13 14:50:00 CST
1098		Tag: Release_110_45
1099		Description:
1100
1101		New working version (110.45). New bootfiles.
1102
1103		----------------------------------------------------------------------
1104		Name: Matthias Blume (blume (at) tti - c (dot) org)
1105		Date: 2004/01/26 15:15:15 CST
1106		Tag: blume-20040126-toplevel
1107		Description:
1108
1109		Improved handling of exceptions at the interactive toplevel.
1110
1111		----------------------------------------------------------------------
1112	Name: Matthias Blume (blume (at) tti - c (dot) org)	Name: Matthias Blume (blume (at) tti - c (dot) org)
1113	Date: 2004/01/26	Date: 2004/01/26 11:25:00
1114	Tag: blume-20040126-app	Tag: blume-20040126-app
1115	Description:	Description:
1116
1117	Type of top-level "app" corrected.	Type of top-level "app" corrected.
1118		Added code for setting vp_limitPtrMask to Win32-specific runtime.
1119
1120	----------------------------------------------------------------------	----------------------------------------------------------------------
1121	Name: Matthias Blume (blume (at) tti - c (dot) org)	Name: Matthias Blume (blume (at) tti - c (dot) org)

---

Colored Diff Long Colored Diff Unidiff Context Diff Side by Side

Legend: Removed from v.1446   changed lines   Added in v.1701

---

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