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 1480, Fri Apr 23 15:37:52 2004 UTC
#	Line 13	Line 13
13	Description:	Description:
14
15	----------------------------------------------------------------------	----------------------------------------------------------------------
16		Name: Matthias Blume (blume (at) tti - c (dot) org)
17		Date: 2004/04/23 10:40:00 CDT
18		Tag: blume-20040423-ml-build
19		Description:
20
21		The ml-build script now terminates with a non-0 status when something
22		goes wrong.
23
24		----------------------------------------------------------------------
25		Name: Matthias Blume (blume (at) tti - c (dot) org)
26		Date: 2004/04/22 16:35:00 CDT
27		Tag: blume-20040422-Option
28		Description:
29
30		Made exception Option to be the same as exception Option.Option
31		(as it should be).
32
33		----------------------------------------------------------------------
34		Name: Allen Leung (leunga (at) reservoir (dot) com)
35		Date: 2004/03/19 14:40:00 EST
36		Tag: leunga-20040319-cygwin-nlffi
37		Description:
38
39		Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
40		of SML/NJ.  The problem is that
41
42		lib = dlopen(NULL, ...)
43		f   = dlsym(lib, "malloc");
44
45		does not work on Windows unless we explicitly export symbols
46		such as 'malloc' during linking.  We fixed this by explicitly
47		exporting the required symbols with the magic gcc incantation:
48
49		-Wl,--export-all cygwin.def
50
51		where cygwin.def is a file containing all the symbols that
52		we wish to export.
53
54		I suspect this is a Windows problem and we'll have to
55		do the same (somehow with windows compilers) when
56		we build the native win32 version with the system
57		calls LoadLibrary/GetProcAddress.
58
59		----------------------------------------------------------------------
60		Name: Matthias Blume (blume (at) tti - c (dot) org)
61		Date: 2004/03/04 16:35:00 CST
62		Tag: blume-20040304-intinf-fmt
63		Description:
64
65		Fixed problem with IntInf.fmt (sign would show up on the right instead
66		of on the left for BIN, OCT, and HEX).
67
68		----------------------------------------------------------------------
69		Name: Matthias Blume (blume (at) tti - c (dot) org)
70		Date: 2004/03/04 11:25:00 CST
71		Tag: blume-20040304-symlinks
72		Description:
73
74		Fixed problem with installer script (unix only) where bin/ml-yacc and
75		friends pointed (via symlinks) to absolute locations instead of just
76		.run-sml.  This was reported by Vesa A Norrman.
77
78		----------------------------------------------------------------------
79		Name: Matthias Blume (blume (at) tti - c (dot) org)
80		Date: 2004/02/13 14:50:00 CST
81		Tag: Release_110_45
82		Description:
83
84		New working version (110.45). New bootfiles.
85
86		----------------------------------------------------------------------
87		Name: Matthias Blume (blume (at) tti - c (dot) org)
88		Date: 2004/01/26 15:15:15 CST
89		Tag: blume-20040126-toplevel
90		Description:
91
92		Improved handling of exceptions at the interactive toplevel.
93
94		----------------------------------------------------------------------
95		Name: Matthias Blume (blume (at) tti - c (dot) org)
96		Date: 2004/01/26 11:25:00
97		Tag: blume-20040126-app
98		Description:
99
100		Type of top-level "app" corrected.
101		Added code for setting vp_limitPtrMask to Win32-specific runtime.
102
103		----------------------------------------------------------------------
104		Name: Matthias Blume (blume (at) tti - c (dot) org)
105		Date: 2003/11/18 17:10 CST
106		Tag: blume-20031118-basis-fiddle
107		Description:
108
109		- changed Timer interface to what might become the spec
110		- POSIX_FLAGS -> BIT_FLAGS according to spec
111		- some other minor discrepancies wrt. spec eliminated
112
113		----------------------------------------------------------------------
114		Name: Matthias Blume (blume (at) tti - c (dot) org)
115		Date: 2003/11/06 12:00:00 CST
116		Tag: Release_110_44
117		Description:
118
119		New working version (110.44). New bootfiles.
120
121		----------------------------------------------------------------------
122		Name: Matthias Blume (blume (at) tti - c (dot) org)
123		Date: 2003/11/04 11:50:00 CST
124		Tag: blume-20031104-move-libraries
125		Description:
126
127		Eliminated the "dont_move_libraries" directive in config/targets.
128		(The mechanism was broken and could not be fixed easily.  Moreover,
129		there does not seem to be any reason not to move all libraries into
130		lib during installation.  I originally implemented this directive as a
131		backward-compatibility feature when I first introduced the new CM.
132		Now that things have been stable for a long time and going back to the
133		old CM is not an option, there is no reason to keep it around.)
134
135		----------------------------------------------------------------------
136		Name: Matthias Blume (blume (at) tti - c (dot) org)
137		Date: 2003/11/03 16:00:00 CST
138		Tag: blume-20031103-installdir
139		Description:
140
141		Made installer honor INSTALLDIR variable again.  (Thanks to Chris
142		Richards for pointing out the problem and providing the solution.)
143
144		----------------------------------------------------------------------
145		Name: Matthias Blume (blume (at) tti - c (dot) org)
146		Date: 2003/10/01 17:05:00 CDT
147		Tag: blume-20031001-lal-mlrisc
148		Description:
149
150		MLRISC bug fix from Lal.
151
152		----------------------------------------------------------------------
153		Name: Matthias Blume (blume (at) tti - c (dot) org)
154		Date: 2003/09/30 16:10:00 CDT
155		Tag: blume-20030930-primio-bat
156		Description:
157
158		1. Added openVector, nullRd, and nullWr to PRIM_IO.
159		2. Improved .bat files (for Win32 port) to make things work under Win95.
160		(thanks to Aaron S. Hawley for this one)
161
162		----------------------------------------------------------------------
163		Name: Matthias Blume (blume (at) tti - c (dot) org)
164		Date: 2003/09/26 16:05:00 CDT
165		Tag: blume-20030926-wrappriv
166		Description:
167
168		Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
169
170		----------------------------------------------------------------------
171		Name: Matthias Blume (blume (at) tti - c (dot) org)
172		Date: 2003/09/26 15:00:00 CDT
173		Tag: blume-20030926-110_43_3
174		Description:
175
176		- additional cleanup
177		- version number bump, NEW BOOTFILES
178
179		----------------------------------------------------------------------
180		Name: Matthias Blume (blume (at) tti - c (dot) org)
181		Date: 2003/09/26 12:00:00 CDT
182		Tag: blume-20030926-ppautoload
183		Description:
184
185		I modified the read-eval-print loop so that the autoloader gets
186		invoked whenever the prettyprinter tries to look up a symbol that
187		is not currently defined in the toplevel environment but which
188		appears in CM's autoload registry.  As a result, we see far fewer of
189		those ?.Foo.Bar.xxx names in the prettyprinter's output.
190
191		In addition to this I tried to clean up some pieces of the Basis
192		implementation (e.g., Socket, Word8Array) in order to prevent other
193		instances of these ?.Foo.Bar.xxx names from being printed.
194
195		The mechanism that picks names for types still needs some work, though.
196		(Right now it seems that if there is a type A.t which is defined to
197		be B.u, but B is unavailable at toplevel, then A.t gets printed as
198		"?.B.u" although the perhaps more sensible solution would be to use
199		"A.t" in this case.  In other words, the prettyprinter should follow
200		a chain of DEFtycs not farther than there are corresponding toplevel
201		names in the current environment.)
202
203		----------------------------------------------------------------------
204		Name: Matthias Blume (blume (at) tti - c (dot) org)
205		Date: 2003/09/24 16:31:00 CDT
206		Tag: blume-20030924-installer
207		Description:
208
209		Another installer tweak:  All the ML code for the installer is now
210		compiled during CMB.make and put into a little library called
211		$smlnj/installer.cm.  The installation then simply invokes
212
213		sml -m $smlnj/installer.cm
214
215		and everything happens automagically.
216
217		Win32:  ML code senses value of environment variable SMLNJ_HOME.
218		Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
219		and BINDIR.
220
221		The new scheme guarantees that the ML code responsible for the installation
222		is in sync with the APIs of the main system.  Also, the installer is
223		somewhat faster because the installer script is precompiled.
224
225		----------------------------------------------------------------------
226		Name: Matthias Blume (blume (at) tti - c (dot) org)
227		Date: 2003/09/24 15:35:00 CDT
228		Tag: blume-20030924-synsock
229		Description:
230
231		Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
232		but excludes all non-blocking operations.  Defined SOCKET (in Basis)
233		and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
234		implementations of non-blocking operations from CML's Socket
235		structure.
236
237		----------------------------------------------------------------------
238		Name: Matthias Blume (blume (at) tti - c (dot) org)
239		Date: 2003/09/24 15:10:05 CDT
240		Tag: blume-20030924-sockets
241		Description:
242
243		1. Fixed SOCKET API and implementation to match Basis spec.
244		This required changing the internal representation of sockets to one
245		that remembers (for each socket file descriptor) whether it is currently
246		blocking or non-blocking.  This state is maintained lazily (i.e., a system
247		call is made only if the state actually needs to change).
248
249		2. OS-specific details of sockets were moved into separate files, thus
250		making it possible to unify the bulk of the socket implementations
251		between Unix and Win32.
252
253		3. CML's socket API changed accordingly.
254		(Note that we need to remove non-blocking functions from this API
255		since they are redundant in the case of CML!)
256
257		4. CML's socket implementation now makes use of non-blocking functions
258		provided by Basis, thus removing all OS-dependent code from this part
259		of CML.
260
261		5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
262
263		----------------------------------------------------------------------
264		Name: Matthias Blume (blume (at) tti - c (dot) org)
265		Date: 2003/09/22 12:10:00 CDT
266		Tag: blume-20030922-110_43_2
267		Description:
268
269		Made a new interim version and bootfiles for developer's bootstrapping
270		convenience.
271
272		110.43.2 -- NEW BOOTFILES
273
274		----------------------------------------------------------------------
275		Name: Matthias Blume (blume (at) tti - c (dot) org)
276		Date: 2003/09/19 15:55:00 CDT
277		Tag: blume-20030919-cmdir
278		Description:
279
280		1. new-install.sh -> install.sh
281		2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
282		3. tweaked installer so that another name instead of .cm can be chosen
283		at install time (by setting the CM_DIR_ARC environment variable
284		during installation); once installation is complete, the name is
285		fixed
286
287		----------------------------------------------------------------------
288		Name: Matthias Blume (blume (at) tti - c (dot) org)
289		Date: 2003/09/18 16:00:00 CDT
290		Tag: blume-20030918-110_43_1
291		Description:
292
293		Made a new interim version and bootfiles for developer's bootstrapping
294		convenience.
295
296		110.43.1 -- NEW BOOTFILES
297
298		----------------------------------------------------------------------
299		Name: Matthias Blume (blume (at) tti - c (dot) org)
300		Date: 2003/09/18 15:20:00 CDT
301		Tag: blume-20030918-misc
302		Description:
303
304		1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
305		this seems to be controversial at the moment) and used those in
306		Posix.ProcEnv.times.
307
308		2. Added Time.{from,to}Nanoseconds to Time.
309
310		3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
311		For example, fromLargeInt never needs to look at more than 3 "big
312		digits" to get its 53 bits of precision.
313
314		----------------------------------------------------------------------
315		Name: Matthias Blume (blume (at) tti - c (dot) org)
316		Date: 2003/09/17 16:30:00 CDT
317		Tag: blume-20030917-real32-slices
318		Description:
319
320		Added an entry to the primitive environment
321		(compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
322		added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
323
324		Removed some of the "magic" constants in real64.sml and replaced them
325		with code that generates these values from their corresponding
326		integer counterparts.
327
328		Made all(?) the slice-related changes to the Basis and made everything
329		compile again...
330
331		----------------------------------------------------------------------
332		Name: Matthias Blume (blume (at) tti - c (dot) org)
333		Date: 2003/09/15 17:45:00 CDT
334		Tag: blume-20030915-rbase
335		Description:
336
337		Fixed bug in Real.fromLargeInt.
338
339		----------------------------------------------------------------------
340		Name: Matthias Blume (blume (at) tti - c (dot) org)
341		Date: 2003/09/13 18:11:00 CDT
342		Tag: blume-20030913-libinstall
343		Description:
344
345		Minor bugfix in config/libinstall (set anchor with path to
346		standalone tool after installing it, otherwise libraries that
347		need ml-lex or ml-yacc won't compile the first time the installer
348		runs).
349
350		----------------------------------------------------------------------
351		Name: Matthias Blume (blume (at) tti - c (dot) org)
352		Date: 2003/09/12 11:45:00 CDT
353		Tag: blume-20030912-various
354		Description:
355
356		- fixed bug in Real.toLargeInt
357		- fixed bug in Posix.ProcEnv.times
358		- changed inputLine functions to return an option
359		- minor installer improvements / bugfixes
360		- changed default @SMLalloc parameter for x86/celeron to 64k
361
362		----------------------------------------------------------------------
363		Name: Matthias Blume (blume (at) tti - c (dot) org)
364		Date: 2003/09/09 22:00:00 CDT
365		Tag: Release_110_43
366		Description:
367
368		New working release 110.43.  New bootfiles.
369
370		----------------------------------------------------------------------
371		Name: Matthias Blume (blume (at) tti - c (dot) org)
372		Date: 2003/09/09 19:20:00 CDT
373		Tag: blume-20030909-installer
374		Description:
375
376		Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
377		Modified config/install.bat to take advantage of it.  Also modified
378		config/install.sh (and called it config/new-install.sh) to take advantage
379		of it on Unix systems.  (The SML code is (supposed to be) platform-
380		independent.)
381
382		The installer can now install everything under Win32
383		as well as under *nix as long as it compiles.
384
385		Other changes:
386
387		- made CML compile again under Win32
388		- made eXene compile under Win32 (by providing a fake structure UnixSock
389		and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
390		- fixed a bug in nowhere: it assumed that type OS.Process.status is the
391		same as type int;  under Win32 it isn't
392		- fixed some slice-related problems in the win32-specific parts of CML
393		- added a functor argument "sameVol" to os-path-fn.sml in the Basis
394		(under Win32, the volume name is case-insensitive, and the
395		OS.Path code compares volume names for equality)
396
397		----------------------------------------------------------------------
398		Name: Matthias Blume (blume (at) tti - c (dot) org)
399		Date: 2003/09/08 11:55:00 CDT
400		Tag: blume-20030908-fullpath
401		Description:
402
403		Made Win32 version of OS.FileSys.fullPath return current directory
404		when given an empty string.  This is what the spec says, and incidentally,
405		CM depends on it.  (CM otherwise goes into an infinite loop in certain
406		cases when presented with the name of a non-existing .cm file.)
407
408		----------------------------------------------------------------------
409		Name: Matthias Blume (blume (at) tti - c (dot) org)
410		Date: 2003/09/04 16:30:00 CDT
411		Tag: blume-20030905-slices-etc
412		Description:
413
414		1. Changed interface to vectors and arrays in Basis to match
415		(draft) Basis spec.
416		2. Added signatures and implementations of slices according to
417		Basis spec.
418		3. Edited source code throughout the system to make it compile again
419		under 1. and 2.  (In some cases code had to be added to have it
420		match the new signatures.)
421		4. MLRISC should be backward-compatible: the copies of the originals
422		of files that needed to change under 3. were retained, the .cm files
423		check the compiler version number and use old versions when
424		appropriate.
425		5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
426		dirstream -> string option (in accordance with Basis spec).
427		6. When generating code that counts lines, ml-lex used function
428		CharVector.foldli, taking advantage of its old interface.
429		This has been replaced with the corresponding code from
430		CharVectorSlice.  (html-lex must be re-lexed!)
431		7. BitArray in smlnj-lib/Util has been extended/modified to match the
432		new MONO_ARRAY signature.  (Do we need BitArraySlice?)
433		8. Removed temporary additions (fromInternal, toInternal) from the
434		(now obsolete) IntInf in smlnj-lib/Util.
435		9. Cleaned up structure Byte.
436		10. Added localOffset, scan, and fromString to Date (according to spec).
437		Cleaned/corrected implementation of Date.
438		(Still need to check for correctness; implement better canonicalizeDate.)
439		11. Added "scan" to signature IEEE_REAL.
440		12. Some improvements to IntInf [in particular: efficiency-hack for
441		mod and rem when second operand is 2 (for parity checks).]
442		13. Changed representation of type Time.time, using a single IntInf.int
443		value counting microseconds.  This considerably simplified the
444		implementation of structure Time.  We now support negative time
445		values; scan and fromString handle signs.
446		14. Functor PrimIO now takes two additional arguments (VectorSlice and
447		ArraySlice).
448
449		----------------------------------------------------------------------
450		Name: Matthias Blume (blume (at) tti - c (dot) org)
451		Date: 2003/08/28 17:00:00 CDT
452		Tag: blume-20030828-intinf
453		Description:
454
455		This is a major update which comes with a version number bump
456		(110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
457		and an implementation of IntInf in the Basis.
458
459		There are a fairly large number of related changes and updates throughout
460		the system:
461
462		Basis:
463		- Implemented IntInf.
464		- Made LargeInt a projection of IntInf (by filtering through INTEGER).
465		- Added some missing Real64 operations, most notably Real.toLargeInt.
466		- Added FixedInt as a synonym for Int32.
467
468		compiler:
469		* Added support for a built-in intinf type.
470		- literals
471		- pattern matching
472		- conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
473		- overloading on literals and operations
474
475		This required adding a primitive type intinf, some additional
476		primops, and implementations for several non-trivial intinf
477		operations in Core.  (The intinf type is completely abstract
478		to the compiler; all operations get delegated back to the Core.)
479
480		* Intinf equality is handled by polyequal.  However, the compiler
481		does not print its usual warning in this case (since polyequal
482		is the right thing to do there).
483
484		* Improved the organization of structure InlineT.
485
486		* A word about conversion primops:
487		If conversions involving intinf do not cancel out during
488		CPS contract, then the compiler must insert calls to Core functions.
489		Since all core access must be resolved already during the FLINT
490		translate phase, it would be too late a the time of CPS contract
491		to add new Core calls.  For this reason, conversion primops
492		for intinf carry two arguments:  1. the numeric argument that
493		they are supposed to convert, and 2. the Core function that
494		can help with this conversion if necessary.  If CPS contract
495		eliminates a primop, then the associated Core function becomes
496		dead and goes away.  Intinf conversion primops that do not get
497		eliminated by CPS contract get rewritten into calls of their
498		core functions by a separate, new phase.
499
500		interactive system:
501		- Control.Print.intinfDepth controls max length of intinf constants
502		being printed. (Analogous to Control.Print.stringDepth.)
503		- Cleanup in printutil and pputil: got rid of unused stuff and
504		duplicates; replaced some of the code with code that makes better
505		use of library functionality.
506
507		CM:
508		Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
509		are no longer silent.
510
511		CKIT:
512		Fixed mismatched uses of Int32 and LargeInt.  I always decided
513		in favor of LargeInt -- which is now the same as IntInf.
514		CKIT-knowledgable people should check whether this is what's
515		intended and otherwise change things back to using Int32 or
516		FixedInt.
517
518		Throughout the code:
519		Started using IntInf.int literals and built-in operations
520		(e.g., comparison with 0) where this seems appropriate.
521
522
523		----------------------------------------------------------------------
524		Name: Dave MacQueen (dbm@cs.uchicago.edu)
525		Date: 2003/08/13 11:36:00 CDT
526		Tag: dbm-20030813-mcz-merge1
527		Description:
528
529		Merging changes from the mcz-branch development branch into trunk.
530		These changes involve replacement of the emulated old prettyprinter
531		interface with direct use of the SML/NJ Lib PP library, and fixing
532		of a couple of bugs (895, 1186) relating to error messages.  A new
533		prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
534		has been added.
535
536		----------------------------------------------------------------------
537		Name: Matthias Blume (blume (at) tti - c (dot) org)
538		Date: 2003/08/11 15:45:00 CDT
539		Tag: blume-20030811-windows
540		Description:
541
542		Version number bumped to 110.42.9.  NEW BOOTFILES!!!
543
544		http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
545
546		This patch restores SML/NJ's ability to run under win32.  There are a
547		number of changes, including fixes for several bugs that had gone
548		unnoticed until now:
549
550		- uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
551		trying to run the win32 version from within cygwin.)
552
553		- There are a number of simple .bat scripts that substitute for their
554		corresponding Unix shell-scripts.  (See below.)
555
556		- The internals of ml-build have been modified slightly.  The main
557		difference is that instead of calling ".link-sml" (or link-sml.bat)
558		using OS.Process.system, the ML process delegates this task back
559		to the script.  Otherwise problems arise in mixed environments such
560		as Cygwin where scripts look and work like Unix scripts, but
561		where OS.Process.system cannot run them.
562
563		- In CM, the srcpath pickler used native pathname syntax -- which
564		is incorrect in the case of cross-compilation.  The new pickle format
565		is independent of platform-specific naming conventions.
566
567		- Path configuration files (such as lib/pathconfig) can now choose
568		between native and standard syntax.  Placing a line of the form
569
570		standard!
571
572		into the file causes all subsequent paths to be interpreted using
573		CM standard pathname syntax (= Unix conventions); a line
574
575		native!
576
577		switches back to native style.  This was needed so that
578		path config files can be written portably, see src/system/pathconfig.
579
580		- Runtime system:
581
582		- win32-filesys.c:  get_file_time and set_file_time now
583		access modification time, not creation time.
584
585		- I/O code made aware of new array representation.
586
587		- Bug fixes in X86.prim.masm.
588
589		- src/system/makeml made aware of win32. (For use under cygwin
590		and other Unix-environments for windows.)
591
592		- In Basis, fixed off-by-one error in win32-io.sml (function vecF)
593		which caused BinIO.inputAll to fail consistently.
594
595		.bat scripts:
596
597		Windows .bat scripts assume that SMLNJ_HOME is defined.
598
599		- sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
600		applications (sml, ml-yacc, ml-lex).
601		- ml-build.bat: analogous to ml-build.
602		- config\install.bat:  Analogous to config/install.sh.  This requires
603		that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
604		(nmake etc. must be on the path, and vcvars32 must have been run.)
605		Moreover, sources for ml-lex and ml-yacc need to exist under src,
606		and the bootfile hierarchy must have been unpacked under
607		sml.boot.x86-win32.
608		The script is very primitive and does a poor job at error checking.
609		It only installs the base system, ml-lex, and ml-yacc.  No other
610		libraries are being installed (i.e., you get only those that
611		are part of the compiler.)
612		- link-sml.bat:  analogous to .link-sml, but not currently used
613
614		Unrelated bug fixes:
615
616		- ml-nlffigen now exports structures ST_* corresponding to incomplete
617		types.
618		- Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
619		otherwise.)
620
621		----------------------------------------------------------------------
622		Name: Dave MacQueen (macqueen@cs.uchicago.edu)
623		Date: 2003/06/17
624		Tag: macqueen-20030617-bug895
625		Description:
626
627		Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
628		Tag will be used for new development branch (mcz-branch) for
629		use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
630
631		----------------------------------------------------------------------
632		Name: Matthias Blume (blume (at) tti - c (dot) org)
633		Date: 2003/05/27 16:55:00 CDT
634		Tag: blume-20030527-polyeq
635		Description:
636
637		Tried to eliminated most cases of polymorphic equality.
638
639		----------------------------------------------------------------------
640		Name: Matthias Blume (blume (at) tti - c (dot) org)
641		Date: 2003/05/21 17:45:00 CDT
642		Tag: blume-20030517-complete
643		Description:
644
645		Two changes:
646
647		1. Added a flag for controlling whether non-exhaustive bindings will
648		be treated as errors (default is false).
649		2. Cleaned up the *entire* source tree so that CMB.make goes through
650		without a single non-exhaustive match- or bind warning.
651
652		----------------------------------------------------------------------
653		Name: Matthias Blume (blume (at) tti - c (dot) org)
654		Date: 2003/05/17 10:20:00 CDT
655		Tag: blume-20030517-absyn
656		Description:
657
658		1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
659
660		This mainly affects the quality of error messages.  However, some
661		of the code is now more straightforward than before. (Treatment of
662		the above four constructs in translate.sml is much simpler than
663		the "macro-expansion" that was going on before.  Plus, the mach-
664		compiler no longer gets invoked just to be able to compile an
665		if-expression.)
666
667		2. The ErrorMsg.Error exception is now caught and absorbed by the
668		interactive loop.
669
670		----------------------------------------------------------------------
671		Name: Allen Leung
672		Date: 2003/05/16 13:05:00 CDT
673		Tag: leunga-20030516-cygwin-runtime
674		Description:
675
676		Ported the runtime system to cygwin, which uses the unix
677		x86-unix bin files.  Missing/buggy features:
678
679		o getnetbyname, getnetbyaddr: these functions seem to be missing in
680		the Cygwin library.
681		o Ctrl-C handling may be flaky.
682		o Windows system calls and Windows I/O are not supported.
683
684		A new set of binfiles is located at:
685
686		http://www.dorsai.org/~leunga/boot.x86-unix.tgz
687
688		This is only needed for bootstrapping the cygwin version of smlnj.
689		Other x86 versions can use the existing binfiles.
690
691		----------------------------------------------------------------------
692		Name: Matthias Blume
693		Date: 2003/04/08 15:42:00 CDT
694		Tag: blume-20030408-listpair
695		Description:
696
697		1. Added a target 'mlrisc' to installer.
698
699		2. Added missing elements to structure ListPair.
700
701		----------------------------------------------------------------------
702		Name: Allen Leung
703		Date: 2003/01/07 10:40:00 EST
704		Tag: leunga-20030107-int-rem
705		Description:
706
707		Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
708		The arguments to the SUBL instruction were swapped.
709
710		----------------------------------------------------------------------
711		Name: Matthias Blume
712		Date: 2002/12/12 16:25:00 EST
713		Tag: blume-20021212-risc-ra
714		Description:
715
716		Fixed a serious bug in the rewrite code for FP spilling/reloading that
717		sent the RA into an infinite loop when floating point registers get
718		spilled.  (Because of this bug, e.g., nucleic stopped compiling between
719		110.37 and 110.38.)
720		There was another set of potential problems related to the handling of
721		MLRISC annotations (but those did not yet cause real problems, apparently).
722
723		----------------------------------------------------------------------
724		Name: Matthias Blume
725		Date: 2002/12/06 22:40:00 EST
726		Tag: blume-20021206-cm-fileid
727		Description:
728
729		Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
730		This fixes the problem of CM getting confused by files that suddenly
731		change their identity (e.g., by getting unlinked and recreated by some
732		text editor such as vi).  There might be a better/cheaper/cleaner way
733		of doing this, but for now this will have to do.
734
735		----------------------------------------------------------------------
736		Name: Matthias Blume
737		Date: 2002/10/28 09:50:00 EST
738		Tag: blume-20021028-typecheck
739		Description:
740
741		Exported structure Typecheck from $smlnj/viscomp/core.cm.
742
743		----------------------------------------------------------------------
744		Name: Matthias Blume
745		Date: 2002/10/17 09:10:00 EDT
746		Tag: Release_110_42
747		Description:
748
749		In good old tradition, there has been a slight hiccup so that we have
750		to patch 110.42 after the fact.   The old release tag has been replaced
751		(see below).
752
753		The change solves a problem with two competing approaches the
754		configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
755		each other's way.
756
757		This change only affects the runtime system code and the installer script.
758		(No new bootfiles.)
759
760		----------------------------------------------------------------------
761		Name: Matthias Blume
762		Date: 2002/10/16 12:00:00 EDT
763		Tag: Release_110_42_removed
764		Description:
765
766		New working release.  New bootfiles.
767
768		----------------------------------------------------------------------
769		Name: Matthias Blume
770		Date: 2002/10/10 13:10:00 EDT
771		Tag: blume-20021010-ppc-divs
772		Description:
773
774		The mltree operator DIVS must be implemented with an overflow check on
775		the PPC because the hardware indicates divide-by-zero using "overflow" as
776		well.
777
778		----------------------------------------------------------------------
779		Name: Matthias Blume
780		Date: 2002/07/23 11:45:00
781		Tag: blume-20020723-smlnj-home
782		Description:
783
784		Sml now senses the SMLNJ_HOME environment variable.  If this is set,
785		then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
786		CM_PATHCONFIG is also set), the path configuration file is assumed
787		to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
788		the entire tree to some other place and everything will "just work".
789
790		(Companion commands such as ml-build and ml-makedepend also sense this
791		variable.)
792
793		----------------------------------------------------------------------
794		Name: Matthias Blume
795		Date: 2002/07/12 21:19:00 EDT
796		Tag: blume-20020712-liveness
797		Description:
798
799		Exported two useful "step" functions from liveness module (MLRISC).
800
801		----------------------------------------------------------------------
802		Name: Matthias Blume
803		Date: 2002/07/05 16:00 EDT
804		Tag: Release_110_41
805		Description:
806
807		New working release.  New bootfiles.
808
809		----------------------------------------------------------------------
810		Name: Matthias Blume
811		Date: 2002/07/05 10:25:00 EDT
812		Tag: blume-20020705-btimp
813		Description:
814
815		Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
816		other clients can set up backtracing support.
817
818		----------------------------------------------------------------------
819		Name: Matthias Blume
820		Date: 2002/06/25 17:23:00 EDT
821		Tag: blume-20020625-fpmax
822		Description:
823
824		Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
825		case.  (The sense of the isNaN test was reversed -- which made min and
826		max always return their first argument.)
827
828		----------------------------------------------------------------------
829		Name: Matthias Blume
830		Date: 2002/06/11
831		Tag: blume-20020611-unixpath
832		Description:
833
834		Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
835
836		----------------------------------------------------------------------
837		Name: Matthias Blume
838		Date: 2002/06/10 16:35:00 EDT
839		Tag: blume-20020610-ieeereal
840		Description:
841
842		I back-ported my implementation of IEEEReal.fromString from the
843		idlbasis-devel branch so that we can test it.
844
845		Another small change is that ppDec tries to give more information
846		than just "<sig>" in the case of functors.  However, this code is
847		broken in some mysterious way if the functor's body's signature
848		has not been declared by ascription but gets inferred from the
849		implementation.  This needs fixing...
850
851		----------------------------------------------------------------------
852		Name: Matthias Blume
853		Date: 2002/05/31
854		Tag: blume-20020531-btrace-mode
855		Description:
856
857		Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
858		the wayside when I switched over to using Controls everywhere.)
859
860		----------------------------------------------------------------------
861		Name: Lal George
862		Date: 2002/05/23 12:21:40 EDT
863		Tag: george-20020523-visual-labels
864		Description:
865
866		Labels are now displayed in the graphical output to make
867		the fall-through and target blocks obvious.
868
869		----------------------------------------------------------------------
870		Name: Matthias Blume
871		Date: 2002/05/22 11:03:00 EDT
872		Tag: blume-20020522-shrink
873		Description:
874
875		John tweaked yesterday's fix for 1131 to handle an out-of-memory
876		situation that comes up when allocating huge arrays.
877
878		----------------------------------------------------------------------
879		Name: Matthias Blume
880		Date: 2002/05/21 16:00:00 EDT
881		Tag: Release_110_40
882		Description:
883
884		New working release (110.40).  New bootfiles.
885
886		[Also: John Reppy fixed GC bug 1131.]
887
888		----------------------------------------------------------------------
889		Name: Matthias Blume
890		Date: 2002/05/21 12:35:00 EDT
891		Tag: blume-20020521-cmdoc
892		Description:
893
894		CM documentation update.
895
896		----------------------------------------------------------------------
897		Name: Matthias Blume
898		Date: 2002/05/21 10:55:00 EDT
899		Tag: blume-20020521-misc
900		Description:
901
902		- John tweaked runtime to be silent on heap export (except when
903		GC messages are on).
904		- I added a few more things (cross-compiling versions of CMB) to
905		config/preloads (as suggestions).
906
907		----------------------------------------------------------------------
908		Name: Matthias Blume
909		Date: 2002/05/20 22:25:00 EDT
910		Tag: blume-20020520-controls
911		Description:
912
913		- Added ControlUtil structure to control-lib.cm.
914		- Use it throughout.
915		- Used Controls facility to define MLRISC controls (as opposed to
916		registering MLRISC control ref cells with Controls after the fact)
917		- Fixed messed-up controls priorities.
918
919		* Removed again all the stuff from config/preloads that one wouldn't
920		be able to preload at the time the initial heap image is built.
921		(Many libraries, e.g., CML, do not exist yet at this time.  The
922		only libraries that can be preloaded via config/preloads are those
923		that come bundled with the bootfiles.)
924
925		----------------------------------------------------------------------
926		Name: Matthias Blume
927		Date: 2002/05/20 10:59:00 EDT
928		Tag: blume-20020520-preloads
929		Description:
930
931		Added a lot of commented-out suggestions for things to be included
932		in config/preloads.
933
934		----------------------------------------------------------------------
935		Name: Allen Leung
936		Date: 2002/05/18 14:20:00 EDT
937		Tag: leunga-20020518-mdl
938		Description:
939
940		o Made the mdl tool stuff compile and run again.
941		o I've disabled all the stuff that depends on RTL specifications; they
942		are all badly broken anyway.
943
944		----------------------------------------------------------------------
945		Name: Matthias Blume
946		Date: 2002/05/17 16:49:00 EDT
947		Tag: blume-20020517-controls
948		Description:
949
950		0. John Reppy made several modifications to the SML/NJ library.
951		In particular, there is a shiny new controls-lib.cm.
952
953		1. Pushed new controls interface through compiler so that everything
954		compiles again.
955
956		2. Added FormatComb and FORMAT_COMB to the CML version of the
957		SML/NJ library (so that CML compiles again).
958
959		3. Modified init scripts because XXX_DEFAULT environment variables
960		are no longer with us.  (Boot-time initialization is now done
961		using the same environment variables that are also used for
962		startup-time initialization of controls.)
963
964		----------------------------------------------------------------------
965		Name: Lal George
966		Date: 2002/05/15 09:20:10 EDT
967		Tag: george-20020515-pseudo-op-decls
968		Description:
969
970		All pseudo-ops emitted before the first segment declaration
971		such as TEXT, DATA, and BSS directives are assumed to be global
972		declarations and are emitted first in the assembly file. This is
973		useful in a number of situations where one has pseudo-ops that are not
974		specific to any segment, and also works around the constraint that one
975		cannot have client pseudo-ops in the TEXT segment.
976
977		Because no segment is associated with these declarations it is
978		an error to allocate any space or objects before the first segment
979		directive and an exception will be raised. However, we cannot make
980		this check for client pseudo-ops.
981
982		These top level declarations are a field in the CFG graph_info.
983		In theory you can continue to add to this field after the CFG has been
984		built -- provided you know what you are doing;-)
985
986		----------------------------------------------------------------------
987		Name: Matthias Blume
988		Date: 2002/05/13 16:40:00 EDT
989		Tag: blume-20020513-pp-etc
990		Description:
991
992		A few minor bugfixes:
993
994		- Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
995		(Bogus printouts for redefined bindings still occur.  Compiler
996		bug should no longer occur now.  We need to redo the prettyprinter
997		from scratch.)
998
999		- CM pathname printer now also adds escape sequences for ( and )
1000
1001		- commend and docu fixes for ml-nlffi
1002
1003		----------------------------------------------------------------------
1004		Name: Matthias Blume
1005		Date: 2002/05/10 16:40:00 EDT
1006		Tag: blume-20020510-erg-textio
1007		Description:
1008
1009		Applied the following bugfix provided by Emden Gansner:
1010
1011		Output is corrupted when outputSubstr is used rather than output.
1012
1013		The problem occurs when a substring
1014
1015		ss = (s, dataStart, dataLen)
1016
1017		where dataStart > 0, fills a stream buffer with avail bytes left.
1018		avail bytes of s, starting at index dataStart, are copied into the
1019		buffer, the buffer is flushed, and then the remaining dataLen-avail
1020		bytes of ss are copied into the beginning of the buffer. Instead of
1021		starting this copy at index dataStart+avail in s, the current code
1022		starts the copy at index avail.
1023
1024		Fix:
1025		In text-io-fn.sml, change line 695 from
1026		val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
1027		to
1028		val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
1029
1030		----------------------------------------------------------------------
1031		Name: Matthias Blume
1032		Date: 2002/04/12 13:55:00 EDT
1033		Tag: blume-20020412-assyntax
1034		Description:
1035
1036		1. Grabbed newer assyntax.h from the XFree86 project.
1037		2. Fiddled with how to compile X86.prim.asm without warnings.
1038		3. (Very) Minor cleanup in CM.
1039
1040		----------------------------------------------------------------------
1041		Name: Matthias Blume
1042		Date: 2002/04/01 (no joke!) 17:07:00 EST
1043		Tag: blume-20020401-x86div
1044		Description:
1045
1046		Added full support for div/mod/rem/quot on the x86, using the machine
1047		instruction's two results (without clumsily recomputing the remainder)
1048		directly where appropriate.
1049
1050		Some more extensive power-of-two support was added to the x86 instruction
1051		selector (avoiding expensive divs, mods, and muls where they can be
1052		replaced with cheaper shifts and masks).  However, this sort of thing
1053		ought to be done earlier, e.g., within the CPS optimizer so that
1054		all architectures benefit from it.
1055
1056		The compiler compiles to a fixed point, but changes might be somewhat
1057		fragile nevertheless.  Please, report any strange things that you might
1058		see wrt. div/mod/quot/rem...
1059
1060		----------------------------------------------------------------------
1061		Name: Matthias Blume
1062		Date: 2002/03/29 17:22:00
1063		Tag: blume-20020329-div
1064		Description:
1065
1066		Fixed my broken div/mod logic.  Unfortunately, this means that the
1067		inline code for div/mod now has one more comparison than before.
1068		Fast paths (quotient > 0 or remainder = 0) are not affected, though.
1069		The problem was with quotient = 0, because that alone does not tell
1070		us which way the rounding went.  One then has to look at whether
1071		remainder and divisor have the same sign...  :(
1072
1073		Anyway, I replaced the bootfiles with fresh ones...
1074
1075		----------------------------------------------------------------------
1076		Name: Matthias Blume
1077		Date: 2002/03/29 14:10:00 EST
1078		Tag: blume-20020329-inlprims
1079		Description:
1080
1081		NEW BOOTFILES!!!    Version number bumped to 110.39.3.
1082
1083		Primops have changed. This means that the bin/boot-file formats have
1084		changed as well.
1085
1086		To make sure that there is no confusion, I made a new version.
1087
1088
1089		CHANGES:
1090
1091		* removed REMT from mltree (remainder should never overflow).
1092
1093		* added primops to deal with divisions of all flavors to the frontend
1094
1095		* handled these primops all the way through so they map to their respective
1096		MLRISC support
1097
1098		* used these primops in the implementation of Int, Int32, Word, Word32
1099
1100		* removed INLDIV, INLMOD, and INLREM as they are no longer necessary
1101
1102		* parameterized INLMIN, INLMAX, and INLABS by a numkind
1103
1104		* translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
1105		floating point
1106
1107		* used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
1108		and Real (but Real.abs maps to a separate floating-point-only primop)
1109
1110
1111		TODO items:
1112
1113		* Hacked Alpha32 instruction selection, disabling the selection of REMx
1114		instructions because the machine instruction encoder cannot handle
1115		them.  (Hppa, PPC, and Sparc instruction selection did not handle
1116		REM in the first place, and REM is supported by the x86 machine coder.)
1117
1118		* Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
1119		selection phase.  (The two can be streamlined because the hardware
1120		delivers both quotient and remainder at the same time anyway.)
1121
1122		* Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
1123		(Currently the behavior is inconsistent both across architectures and
1124		wrt. the draft Basis spec.)
1125
1126		* Word8 should eventually be handled natively, too.
1127
1128		* There seems to be one serious bug in mltree-gen.sml.  It appears, though,
1129		as if there currently is no execution path that could trigger it in
1130		SML/NJ.  (The assumptions underlying functions arith and promotable do not
1131		hold for things like multiplication and division.)
1132
1133		----------------------------------------------------------------------
1134		Name: Matthias Blume
1135		Date: 2002/03/27 16:27:00 EST
1136		Tag: blume-20020327-mlrisc-divisions
1137		Description:
1138
1139		Added support for all four division operations (ML's div, mod, quot,
1140		and rem) to MLRISC.  In the course of doing so, I also rationalized
1141		the naming (no more annoying switch-around of DIV and QUOT), by
1142		parameterizing the operation by div_rounding_mode (which can be either
1143		DIV_TO_ZERO or DIV_TO_NEGINF).
1144
1145		The generic MLTreeGen functor takes care of compiling all four
1146		operations down to only round-to-zero div.
1147
1148		Missing pieces:
1149
1150		* Doing something smarter than relying on MLTreeGen on architectures
1151		like, e.g., the x86 where hardware division delivers both quotient and
1152		remainder at the same time.  With this, the implementation of the
1153		round-to-neginf operations could be further streamlined.
1154
1155		* Remove inlining support for div/mod/rem from the frontend and replace it
1156		with primops that get carried through to the backend.  Do this for all
1157		int and word types.
1158
1159		----------------------------------------------------------------------
1160		Name: Matthias Blume
1161		Date: 2002/03/25 17:25:00 EST
1162		Tag: blume-20020325-divmod
1163		Description:
1164
1165		I improved (hopefully without breaking them) the implementation of Int.div,
1166		Int.mod, and Int.rem.   For this, the code in translate.sml now takes
1167		advantage of the following observations:
1168
1169		Let  q = x quot y      r = x rem y
1170		d = x div  y      m = x mod y
1171
1172		where "quot" is the round-to-zero version of integer division that
1173		hardware usually provides.  Then we have:
1174
1175		r = x - q * y        where neither the * nor the - will overflow
1176		d = if q >= 0 orelse x = q * y then q else q - 1
1177		where neither the * nor the - will overflow
1178		m = if q >= 0 orelse r = 0 then r else r + y
1179		where the + will not overflow
1180
1181		This results in substantial simplification of the generated code.
1182		The following table shows the number of CFG nodes and edges generated
1183		for
1184		fun f (x, y) = x OPER y
1185		(* with OPER \in div, mod, quot, rem *)
1186
1187
1188		OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
1189		--------------------------------------------------------
1190		div |         24 |         39 |         12 |         16
1191		mod |         41 |         71 |         12 |         16
1192		quot |          8 |         10 |          8 |         10
1193		rem |         10 |         14 |          8 |         10
1194
1195
1196		----------------------------------------------------------------------
1197		Name: Matthias Blume
1198		Date: 2002/03/25 22:06:00 EST
1199		Tag: blume-20020325-cprotobug
1200		Description:
1201
1202		Fixed a bug in cproto (c prototype decoder).
1203
1204		----------------------------------------------------------------------
1205		Name: Matthias Blume
1206		Date: 2002/03/25 16:00:00 EST
1207		Tag: blume-20020325-raw-primops
1208		Description:
1209
1210		I did some cleanup to Allen's new primop code and
1211		replaced yesterday's bootfiles with new ones.
1212		(But they are stored in the same place.)
1213
1214		----------------------------------------------------------------------
1215		Name: Matthias Blume
1216		Date: 2002/03/24 22:40:00 EST
1217		Tag: blume-20020324-bootfiles
1218		Description:
1219
1220		Made the bootfiles that Allen asked for.
1221
1222		----------------------------------------------------------------------
1223		Name: Allen Leung
1224		Date: 2002/03/23 15:50:00 EST
1225		Tag: leunga-20020323-flint-cps-rcc-primops
1226		Description:
1227
1228		1. Changes to FLINT primops:
1229
1230		(* make a call to a C-function;
1231		* The primop carries C function prototype information and specifies
1232		* which of its (ML-) arguments are floating point. C prototype
1233		* information is for use by the backend, ML information is for
1234		* use by the CPS converter. *)
1235		| RAW_CCALL of { c_proto: CTypes.c_proto,
1236		ml_args: ccall_type list,
1237		ml_res_opt: ccall_type option,
1238		reentrant : bool
1239		} option
1240		(* Allocate uninitialized storage on the heap.
1241		* The record is meant to hold short-lived C objects, i.e., they
1242		* are not ML pointers.  With the tag, the representation is
1243		* the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
1244		*)
1245		| RAW_RECORD of {tag:bool,sz:int}
1246		and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
1247
1248		2.  These CPS primops are now overloaded:
1249
1250		rawload of {kind:numkind}
1251		rawstore of {kind:numkind}
1252
1253		The one argument form is:
1254
1255		rawload {kind} address
1256
1257		The two argument form is:
1258
1259		rawload {kind} [ml object, byte-offset]
1260
1261		3. RAW_CCALL/RCC now takes two extra arguments:
1262
1263		a. The first is whether the C call is reentrant, i.e., whether
1264		ML state should be saved and restored.
1265		b. The second argument is a string argument specifying the name of
1266		library and the C function.
1267
1268		These things are currently not handled in the code generator, yet.
1269
1270		4. In CProto,
1271
1272		An encoding type of "bool" means "ml object" and is mapped into
1273		C prototype of PTR.  Note that "bool" is different than "string",
1274		even though "string" is also mapped into PTR, because "bool"
1275		is assigned an CPS type of BOGt, while "string" is assigned INT32t.
1276
1277		5. Pickler/unpicker
1278
1279		Changed to handle RAW_RECORD and newest RAW_CCALL
1280
1281		6. MLRiscGen,
1282
1283		1. Changed to handle the new rawload/rawstore/rawrecord operators.
1284		2. Code for handling C Calls has been moved to a new module CPSCCalls,
1285		in the file CodeGen/cpscompile/cps-c-calls.sml
1286
1287		7. Added the conditional move operator
1288
1289		condmove of branch
1290
1291		to cps.  Generation of this is still buggy so it is currently
1292		disabled.
1293
1294		----------------------------------------------------------------------
1295	Name: Lal George	Name: Lal George
1296	Date: 2002/03/22 14:18:25 EST	Date: 2002/03/22 14:18:25 EST
1297	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.1480

---

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