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

Parent Directory | Revision Log

---

Revision 658 - (view) (download)
Original Path: sml/trunk/HISTORY

1 :	dbm	570	This is the HISTORY file for the Yale SML/NJ CVS repository.
2 :
3 :			An entry should be made for _every_ commit to the repository.
4 :			The entries in this file will be used when creating the README
5 :			for new versions, so keep that in mind when writing the
6 :			description.
7 :
8 :			The form of an entry should be:
9 :
10 :			Name:
11 :			Date:
12 :			Tag: <post-commit CVS tag>
13 :			Description:
14 :	leunga	591	----------------------------------------------------------------------
15 :	blume	658	Name: Matthias Blume
16 :			Date: 2000/06/12 11:40:00
17 :			Tag: blume-20000612-parmakefix
18 :			Description:
19 :
20 :			More cleanup after changing the file naming scheme: This time I
21 :			repaired the parallel make mechanism for CMB.make which I broke earlier.
22 :
23 :			----------------------------------------------------------------------
24 :	leunga	657	Name: Allen Leung
25 :			Date: 2000/06/09 01:25:00
26 :			Tag: leunga-20000609-various
27 :
28 :			None of these things should affect normal SML/NJ operations
29 :
30 :			1. Peephole improvements provided by Fermin (c--)
31 :			2. New annotation DEFUSE for adding extra dependence (moby)
32 :			3. New X86 LOCK instructions (moby)
33 :			4. New machine description language for reservation tables (scheduling)
34 :			5. Fixes to various optimization/analysis modules (branch chaining, dominator
35 :			trees etc.)
36 :			6. I've changed the CM files so that they can work with versions
37 :			110.0.6, 110.25 and 110.28
38 :
39 :			----------------------------------------------------------------------
40 :	blume	652	Name: Matthias Blume
41 :	blume	656	Date: 2000/06/09 12:40:00
42 :			Tag: blume-20000609-log
43 :			Description:
44 :
45 :			- Removed all(?) remaining RCS Log entries from sources.
46 :
47 :			- Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
48 :			anchored paths).
49 :
50 :			----------------------------------------------------------------------
51 :			Name: Matthias Blume
52 :	blume	653	Date: 2000/06/07 17:00:00 JST
53 :			Tag: blume-20000607-no-implicit-anchors
54 :			Description:
55 :
56 :			1. This update changes the default setting for
57 :			CM.Control.implicit_anchors from true to false. This means that
58 :			implicit anchors are no longer permitted by default. I also tried to
59 :			make sure that nothing else still relies on implicit anchors.
60 :			(This is the next step on the schedule towards a CM that does not even
61 :			have the notion of implicit anchors anymore.)
62 :
63 :			2. More CM manual updates.
64 :
65 :			3. I managed to track down and fix the pickling bug I mentioned last
66 :			time. Because of the previously existing workaround, this entails no
67 :			immediate practical changes.
68 :
69 :			----------------------------------------------------------------------
70 :			Name: Matthias Blume
71 :	blume	652	Date: 2000/06/06 11:15:00 JST
72 :			Tag: blume-20000606-lazierpickle
73 :			Description:
74 :
75 :			!!!! NEW BOOT FILES !!!!
76 :
77 :			* The main purpose of this update is to make library pickles lazier in
78 :			order to reduce the initial space penalty for autoloading a library.
79 :			As a result, it is now possible to have $smlnj/compiler.cm
80 :			pre-registered. This should take care of the many complaints or
81 :			inquiries about missing structure Compiler. This required changes to
82 :			CM's internal data structures and small tweaks to some algorithms.
83 :
84 :			As a neat additional effect, it is no longer necessary (for the sake
85 :			of lean heap image files) to distinguish between a "minimal" CM and a
86 :			"full" CM. Now, there is only one CM (i.e., the "full" version:
87 :			$smlnj/cm.cm aka $smlnj/cm/full.cm), and it is always available at the
88 :			interactive top level. ($smlnj/cm/minimal.cm is gone.)
89 :
90 :			To make the life of compiler-hackers easier, "makeml" now also
91 :			pre-registers $smlnj/cmb.cm (aka $smlnj/cmb/current.cm). In other
92 :			words, after you bootstrap a new sml for the first time, you will not
93 :			have to autoload $smlnj/cmb.cm again afterwards. (The first time
94 :			around you will still have to do it, though.)
95 :
96 :			* A second change consists of major updates to the CM manual. There
97 :			are now several appendices with summary information and also a full
98 :			specification of the CM description file syntax.
99 :
100 :			* In directory src/system I added the script "allcross". This script
101 :			invokes sml and cross-compiles the compiler for all supported
102 :			architectures. (Useful when providing a new set of boot files.)
103 :
104 :			* There seems to be a latent bug in my "lazy pickles" mechanism. I
105 :			added a small tweak to pickle-util.sml to work around this problem,
106 :			but it is not a proper fix yet. I will investigate further. (The
107 :			effect of the bug was an inflation of library pickle size.)
108 :
109 :			* Version number increased to 110.28.1 (to avoid compatibility problems).
110 :
111 :			----------------------------------------------------------------------
112 :	leunga	646	Name: Allen Leung
113 :	leunga	648	Date: 2000/05/25 17:28 EDT
114 :			Tag: leunga-20000525-ra
115 :			Description:
116 :
117 :			Fixed a bug in freezing phase of the register allocator.
118 :
119 :			----------------------------------------------------------------------
120 :			Name: Allen Leung
121 :	leunga	647	Date: 2000/05/15 22:53 EDT
122 :	leunga	646	Tag: leunga-20000515-alpha-x86-ra
123 :			Description:
124 :
125 :			1. Alpha
126 :
127 :			Slight cleanup. Removed the instruction SGNXL
128 :
129 :			2. X86
130 :
131 :			Added the following instructions to the instruction set:
132 :
133 :			ROLx, RORx,
134 :			BTx, BTSx, BTLx, BTRx,
135 :			XCHGx, and variants with the LOCK prefix
136 :
137 :			3. Register Allocation
138 :
139 :			The module ra-rewrite-with-renaming has been improved.
140 :
141 :	leunga	647	These have no effect on SML/NJ.
142 :
143 :	leunga	646	----------------------------------------------------------------------
144 :	blume	642	Name: Matthias Blume
145 :	blume	645	Date: 2000/05/15 16:20:00 JST
146 :			Tag: blume-20000515-lightrebuild
147 :			Description:
148 :
149 :			1. I added an alternative to "-rebuild" to "makeml". The difference is
150 :			that prior to calling CMB.make' the CM-variable "LIGHT" will be
151 :			defined. In effect, the command will not build any cross-compiler
152 :			backends and therefore finish more quickly.
153 :
154 :			The "fixpt" script also takes a "-light" switch to be able to use
155 :			this new facility while compiling for a fixpoint.
156 :
157 :			2. I replaced all mentions of anchored paths in group owner specifications
158 :			with simple relative paths (usually starting with "..").
159 :			The rationale is that a library's internal workings should not be
160 :			compromised by the lack of some anchor. (An anchor is necessary
161 :			for someone who wants to refer to the library by an anchored path,
162 :			but it should not be necessary to build the same library in the first
163 :			place.)
164 :
165 :			3. I changed the way CM's tool mechanism determines the shell command
166 :			string used for things like ml-yacc etc. so that it does not break
167 :			when CM.Control.implicit_anchors is turned off.
168 :
169 :			----------------------------------------------------------------------
170 :			Name: Matthias Blume
171 :	blume	644	Date: 2000/05/12 18:20:00 JST
172 :			Tag: blume-20000512-ml-build
173 :			Description:
174 :
175 :			Fixed a bug in config/_ml-build that prevented ml-yacc and ml-lex from
176 :			getting installed properly (by config/install.sh).
177 :
178 :			----------------------------------------------------------------------
179 :			Name: Matthias Blume
180 :	blume	643	Date: 2000/05/12 17:30:00 JST
181 :			Tag: blume-20000512-anchors
182 :			Description:
183 :
184 :			!!! NEW BOOT FILES !!!
185 :
186 :			This change is in preparation of fading out support for "implicitly
187 :			anchored path names". I went through all sources and used the
188 :			explicit (and relatively new) $-notation. See system/README and the
189 :			CM manual for more info on this.
190 :
191 :			I also modified the anchoring scheme for some things such as "smlnj",
192 :			"MLRISC", "cm", etc. to take advantage of the fact that explicit
193 :			anchors are more expressive: anchor name and first arc do not have to
194 :			coincide. This entails the following user-visible change:
195 :
196 :			You have to write $smlnj/foo/bar instead of smlnj/foo/bar. In
197 :			particular, when you fire up sml with a command-line argument, say,
198 :			e.g.:
199 :
200 :			sml '$smlnj/cmb.cm'
201 :
202 :			At the ML toplevel prompt:
203 :
204 :			CM.autoload "$smlnj/cmb.cm";
205 :
206 :			There is also a new controller in CM.Control that can be used to turn
207 :			off all remaining support for implicit anchors by saying:
208 :
209 :			CM.autoload "$smlnj/
210 :			#set CM.Control.implicit_anchors false;
211 :
212 :			This causes CM to reject implicitly anchored paths. This is (for the
213 :			time being) less permissive than the "final" version where there will
214 :			be no more such implicit anchors and relative paths will be just that:
215 :			relative.
216 :
217 :			The next step (version after next version?) will be to make the
218 :			default for CM.Control.implicit_anchors false. After the dust has
219 :			settled, I can then produce the "final" version of this...
220 :
221 :			Note: Since bootstrapping is a bit tricky, I provided new boot files.
222 :
223 :			----------------------------------------------------------------------
224 :			Name: Matthias Blume
225 :	blume	642	Date: 2000/05/11 16:30:00 JST
226 :			Tag: blume-20000511-sources
227 :			Description:
228 :
229 :			The main change is that I added function CM.sources as a generalized
230 :			version of the earlier CM.makedepend. This entails the following
231 :			additional changes:
232 :
233 :			- CM.makedepend has been dropped.
234 :
235 :			- CM manual has been updated.
236 :
237 :			- TOOLS signature and API have been changed.
238 :
239 :			----------------------------------------------------------------------
240 :	leunga	640	Name: Allen Leung
241 :	leunga	641	Date: 2000/05/10 21:17 EDT
242 :			Tag: leunga-20000510-moby-c--ssa
243 :			Description:
244 :
245 :			Various bug fixes and new features for C--, Moby and MLRISC optimizations.
246 :			None of these affect SML/NJ.
247 :
248 :			1. Register Allocation
249 :
250 :			a. A new ra spilling module (ra/ra-spill-with-renaming) is implemented.
251 :			This module tries to remove local (i.e. basic block level) redundancies
252 :			during spilling.
253 :
254 :			b. A new framework for performing region based register allocation.
255 :			Not yet entirely functional.
256 :
257 :			2. X86
258 :
259 :			a. DefUse for POP was missing the stack pointer [found by Lal]
260 :			b. Reload for CALL was incorrect in X86Spill [found by John]
261 :			c. Various fixes in X86Spill so that it can be used correctly for
262 :			the new spilling module.
263 :
264 :			3. SSA/IR
265 :
266 :			a. New module ir/dj-dataflow.sml implements elimination based
267 :			data flow analysis.
268 :
269 :			4. MLRiscGen
270 :
271 :			a. Fix for gc type annotation
272 :
273 :			5. MDGen
274 :
275 :			Various fixes for machine description -> ml code translation. For ssa
276 :			only.
277 :
278 :			----------------------------------------------------------------------
279 :			Name: Allen Leung
280 :	leunga	640	Date: 2000/05/08 22:17 EDT
281 :			Tag: leunga-20000508-labexp
282 :			Description:
283 :
284 :			Fermin has found a few assembly problems with constant expressions
285 :			generated in LabelExp. Mostly, the problems involve extra parentheses,
286 :			which choke on dumb assemblers. This is his fix.
287 :
288 :			----------------------------------------------------------------------
289 :	dbm	635	Name: Dave MacQueen
290 :			Date: 2000/04/09 14:00 EDT
291 :			Tag: dbm-20000502-Version_110_28
292 :			Description:
293 :
294 :			1. Updated src/compiler/TopLevel/main/version.sml to version 110.28
295 :
296 :			2. Updated config/version to 110.28
297 :
298 :			3. Updated config/srcarchiveurl
299 :
300 :			3. New boot files!
301 :			ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.28/
302 :
303 :			----------------------------------------------------------------------
304 :	blume	634	Name: Matthias Blume
305 :			Date: 2000/05/01 19:05:00 JST
306 :			Tag: blume-20000501-noweb
307 :			Description:
308 :
309 :			A new noweb tool has been added. The existing system is entirely
310 :			unaffected by this, but some CM users have asked for renewed noweb
311 :			support. Everything is documented in the CM manual.
312 :
313 :			New (plugin) libraries:
314 :
315 :			noweb-tool.cm
316 :			nw-ext.cm
317 :
318 :			----------------------------------------------------------------------
319 :	dbm	633	Name: Dave MacQueen
320 :			Date: 2000/04/30 12:40PM EDT
321 :			Tag: dbm-20000430-bug_fixes
322 :			Description:
323 :
324 :			1. Fix for bug 1498
325 :			smlnj/src/system/Basis/Implementation/Unsafe/object.sig
326 :			smlnj/src/system/Basis/Implementation/Unsafe/object.sml
327 :			added toRealArray function
328 :			smlnj/src/compiler/MiscUtil/print/ppobj.sml
329 :			added check for tag Obj.RealArray to array printing case in ppObj
330 :
331 :			2. Fix for bug 1510
332 :			smlnj/src/compiler/Semant/types/typesutil.sml
333 :			fixed definition of dummyargs (used by equalTycon) so that
334 :			dummy args are distinct types
335 :
336 :			----------------------------------------------------------------------
337 :	blume	629	Name: Matthias Blume
338 :	blume	632	Date: 2000/04/30 01:00:00 JST
339 :			Tag: blume-20000430-versions
340 :			Description:
341 :
342 :			1. CM version numbering added. This is an implementation of Lal's
343 :			proposal for adding version numbers and version checking to .cm
344 :			files. Lal said that his proposal was just that -- a proposal.
345 :			For the time being I went ahead and implemented it so that people
346 :			can comment on it. Everything is completely backward-compatible
347 :			(except for the stable library format, i.e., new bootfiles!).
348 :
349 :			As usual, see the CM manual for details.
350 :
351 :			2. An alternative syntax for anchored paths has been implemented.
352 :			Dave has recently voiced the same concerns that I had when I did
353 :			this, so there should be some support. My take is that eventually
354 :			I will let support for the current syntax (where anchors are
355 :			"implicit") fade out in favor of the new, explicit syntax.
356 :			In order to be backward-compatible, both old and new syntax are
357 :			currently supported.
358 :
359 :			Again, see the CM manual for details.
360 :
361 :			3. Parallel make is trying to be slightly smarter: When the master
362 :			process finds a "bottleneck", i.e., when there is only one
363 :			compilation unit that can be compiled and everybody else is
364 :			waiting on it, then it will simply compile it directly instead
365 :			of clumsily telling one of the slaves to do it.
366 :
367 :			4. Support for "unsharing" added. This is necessary in order to be
368 :			able to have two different versions of the same library running
369 :			at the same time (e.g., for trying out a new MLRISC while still
370 :			having the old MLRISC linked into the current compiler, etc.)
371 :			See the CM manual.
372 :
373 :			5. Simple "makedepend" functionality added for generating Makefile
374 :			dependency information. (This is rather crude at the moment.
375 :			Expect some changes here in the future.)
376 :
377 :			6. ".fun" added as a recognized suffix for ML files. Also documented
378 :			explicitly in the manual that the fallback behavior (unknown suffix
379 :			-> ML file) is not an official feature!
380 :
381 :			7. Small changes to the pickler for stable libraries.
382 :
383 :			8. Several internal changes to CM (for cleanup/improvement).
384 :
385 :
386 :			!!!! NEW BINFILES !!!!
387 :
388 :			----------------------------------------------------------------------
389 :			Name: Matthias Blume
390 :	blume	631	Date: 2000/04/28 17:30:00 JST
391 :			Tag: blume-20000428-pathconfig
392 :			Description:
393 :
394 :			1. I changed config/install.sh to remove duplicate entries from the
395 :			lib/pathconfig file at the end. Moreover, the final version of
396 :			lib/pathconfig is sorted alphabetically. The same (sorting) is done
397 :			in src/system/installml.
398 :
399 :			2. The config/install.sh script now consistently uses relative
400 :			pathnames in lib/pathconfig whenever the anchor is in the lib
401 :			directory. (So far this was true for the libraries that come
402 :			pre-compiled and bundled as part of the bootfiles but not for
403 :			libraries that are compiled by the script itself.)
404 :
405 :			----------------------------------------------------------------------
406 :			Name: Matthias Blume
407 :	blume	629	Date: 2000/04/26 13:10:00 JST
408 :			Tag: blume-20000426-fun_suffix
409 :			Description:
410 :
411 :			Added ".fun" as a recognized file name suffix (for ML code).
412 :
413 :			----------------------------------------------------------------------
414 :	leunga	628	Name: Allen Leung
415 :			Date: 2000/04/25 17:00:00 EST
416 :			Tag: leunga-20000425-alpha-ra
417 :			Description:
418 :
419 :			1. Alpha
420 :
421 :			PSEUDOARITH was missing in AlphaRewrite. This causes an endless loop
422 :			in C--.
423 :
424 :			2. RA
425 :
426 :			Added a flag "ra-dump-size" to print out the size of the flowgraph
427 :			and the interference graph.
428 :
429 :			----------------------------------------------------------------------
430 :	dbm	627	Name: Dave MacQueen
431 :			Date: 2000/04/25/
432 :			Tag: dbm-20000425-mlyacc_doc_examples
433 :			Description:
434 :			Updated mlyacc.tex sections 5 and 7 for SML '97 and CM.
435 :			Updated all three examples in src/ml-yacc/examples to run
436 :			under 110.* using CM.make.
437 :
438 :			----------------------------------------------------------------------
439 :	leunga	624	Name: Allen Leung
440 :			Date: 2000/04/20 23:04:00 EST
441 :			Tag: leunga-20000420-ssa-c---stuff
442 :			Description:
443 :
444 :			This update synchronizes my repository with Yale's. Most of these
445 :			changes, however, do not affect SML/NJ at all (the RA is an exception).
446 :
447 :			1. Register Allocator
448 :
449 :			a. An improvement in the interference graph construction:
450 :			Given a copy
451 :
452 :			s <- t
453 :
454 :	leunga	628	no interference edge between s and t is added for this definition of s.
455 :	leunga	624
456 :			b. I've added two new spill heuristic modules that Fermin and I developed
457 :			(in the new library RA.cm). These are unused in SML/NJ but maybe
458 :			useful for others (Moby?)
459 :
460 :			2. X86
461 :
462 :			a. Various fixes in the backend provided by Fermin [C--] and Lal.
463 :
464 :			3. Alpha
465 :
466 :			a. Added the BSR instruction and code generation that goes with it [C--]
467 :			b. Other fixes too numerous to recount provided by Fermin [C--]
468 :
469 :			4. Regmaps
470 :
471 :			a. The regmaps are not initialized with the identity physical bindings
472 :			at creation time. This is unneeded.
473 :
474 :			5. MLRISC Optimizations
475 :
476 :			a. The DJ-Graph module can now compute the iterated dominance frontiers
477 :			intersects with liveness incrementally in linear time! Woohoo!
478 :			This is now used in my new SSA construction algorithm.
479 :
480 :			b. THe branch reorganization module is now smarter about linear chains of
481 :			basic blocks.
482 :
483 :
484 :			----------------------------------------------------------------------
485 :	blume	615	Name: Matthias Blume
486 :			Date: 2000/04/12 13:52:00 JST
487 :			Tag: blume_main_v110p27_1
488 :			Description:
489 :
490 :			Changed install.sh script to handle archive files without version number
491 :			and to use "boot.<arch>-<os>" instead of "sml.boot.<arch>-<os>" for the
492 :			name of the boot file archive.
493 :
494 :			----------------------------------------------------------------------
495 :	dbm	607	Name: Dave MacQueen
496 :			Date: 2000/04/09 14:00 EDT
497 :			Tag: dbm-20000410-Version_110_27
498 :			Description:
499 :
500 :			1. Updated src/compiler/TopLevel/main/version.sml to version 110.27
501 :
502 :			2. Updated src/config/version to 110.27
503 :
504 :			3. New boot files!
505 :
506 :			----------------------------------------------------------------------
507 :	leunga	606	Name: Allen Leung
508 :			Date: 2000/04/09 19:09:00 EST
509 :			Tag: leunga-20000409-misc
510 :			Description:
511 :
512 :			1. Yet another fix for x86 assembly for idivl, imull, mull and friends.
513 :
514 :			2. Miscellaneous improvements to MLRISC (unused in sml/nj)
515 :
516 :			----------------------------------------------------------------------
517 :	monnier	605	Name: Stefan
518 :			Date: 2000/04/07 10:00:00 EDT
519 :			Tag: monnier-20000406-branch-handling
520 :			Description:
521 :
522 :			Improved handling of branches (mostly those generated from
523 :			polymorphic equality), removed switchoff and changed the
524 :			default optimization settings (more cpsopt and less flintopt).
525 :
526 :			----------------------------------------------------------------------
527 :	leunga	601	Name: Allen Leung
528 :	leunga	602	Date: 2000/04/06 01:30:00 EST
529 :			Tag: leunga-20000406-peephole-x86-SSA-2
530 :			Description:
531 :
532 :			Forgot a few files.
533 :
534 :			----------------------------------------------------------------------
535 :			Name: Allen Leung
536 :	leunga	601	Date: 2000/04/06 00:36:00 EST
537 :			Tag: leunga-20000406-peephole-x86-SSA
538 :			Description:
539 :
540 :			1. New Peephole code
541 :
542 :			2. Minor improvement to X86 instruction selection
543 :
544 :			3. Various fixes to SSA and machine description -> code translator
545 :
546 :			----------------------------------------------------------------------
547 :	blume	595	Name: Matthias Blume
548 :			Date: 2000/04/05 12:30:00 JST
549 :			Tag: blume_main_v110p26p2_3
550 :			Description:
551 :
552 :			This update just merges three minor cosmetic updates to CM's sources
553 :			to get ready for the 110.27 code freeze on Friday. No functionality
554 :			has changed.
555 :
556 :			----------------------------------------------------------------------
557 :	leunga	593	Name: Allen Leung
558 :	leunga	594	Date: 2000/04/04 19:39:00 EST
559 :			Tag: leunga-20000404-x86-asm
560 :			Description:
561 :
562 :			1. Fixed a problem in X86 assembly.
563 :
564 :			Things like
565 :
566 :			jmp %eax
567 :			jmp (%eax)
568 :
569 :			should be output as
570 :
571 :			jmp *%eax
572 :			jmp *(%eax)
573 :
574 :			2. Assembly output
575 :
576 :			Added a new flag
577 :
578 :			"asm-indent-copies" (default to false)
579 :
580 :			When this flag is on, parallel copies will be indented an extra level.
581 :
582 :			----------------------------------------------------------------------
583 :			Name: Allen Leung
584 :	leunga	593	Date: 2000/04/04 03:18:00 EST
585 :			Tag: leunga-20000404-C--Moby
586 :			Description:
587 :
588 :			All of these fixes are related to C--, Moby, and my own optimization
589 :			stuff; so they shouldn't affect SML/NJ.
590 :
591 :			1. X86
592 :
593 :			Various fixes related floating point, and extensions.
594 :
595 :			2. Alpha
596 :
597 :			Some extra patterns related to loads with signed/zero extension
598 :			provided by Fermin.
599 :
600 :			3. Assembly
601 :
602 :			When generating assemby, resolve the value of client defined constants,
603 :			instead of generating symbolic values. This is controlled by the
604 :			new flag "asm-resolve-constants", which is default to true.
605 :
606 :			4. Machine Descriptions
607 :
608 :			a. The precedence parser was slightly broken when parsing infixr symbols.
609 :			b. The type generalizing code had the bound variables reversed, resulting
610 :			in a problem during arity raising.
611 :			c. Various fixes in machine descriptions.
612 :
613 :			----------------------------------------------------------------------
614 :	blume	592	Name: Matthias Blume
615 :			Date: 2000/04/03 16:05:00 JST
616 :			Tag: blume_main_v110p26p2_2
617 :			Description:
618 :
619 :			I eliminated coreEnv from compInfo. Access to the "Core" structure is
620 :			now done via the ordinary static environment that is context to each
621 :			compilation unit.
622 :
623 :			To this end, I arranged that instead of "structure Core" as "structure
624 :			_Core" is bound in the pervasive environment. Core access is done via
625 :			_Core (which can never be accidentially rebound because _Core is not a
626 :			legal surface-syntax symbol).
627 :
628 :			The current solution is much cleaner because the core environment is
629 :			now simply part of the pervasive environment which is part of every
630 :			compilation unit's context anyway. In particular, this eliminates all
631 :			special-case handling that was necessary until now in order to deal
632 :			with dynamic and symbolic parts of the core environment.
633 :
634 :			Remaining hackery (to bind the "magic" symbol _Core) is localized in the
635 :			compilation mananger's bootstrap compiler (actually: in the "init group"
636 :			handling). See the comments in src/system/smlnj/init/init.cmi for
637 :			more details.
638 :
639 :			I also tried to track down all mentions of "Core" (as string argument
640 :			to Symbol.strSymbol) in the compiler and replaced them with a
641 :			reference to the new CoreSym.coreSym. Seems cleaner since the actual
642 :			name appears in one place only.
643 :
644 :			Binfile and bootfile format have not changed, but the switchover from
645 :			the old "init.cmi" to the new one is a bit tricky, so I supplied new
646 :			bootfiles anyway.
647 :
648 :			----------------------------------------------------------------------
649 :	leunga	591	Name: Allen Leung
650 :			Date: 2000/04/02 21:17:00 EST
651 :			Tag: leunga-20000402-mltree
652 :			Description:
653 :	leunga	585
654 :	leunga	591	1. Renamed the constructor CALL in MLTREE by popular demand.
655 :			2. Added a bunch of files from my repository. These are currently
656 :			used by other non-SMLNJ backends.
657 :
658 :	leunga	576	----------------------------------------------------------------------
659 :	leunga	590	Name: Allen Leung
660 :			Date: 2000/03/31 21:15:00 EST
661 :			Tag: leunga-20000331-aliasing
662 :			Description:
663 :
664 :			This update contains a rewritten (and hopefully more correct) module
665 :			for extracting aliasing information from CPS.
666 :
667 :			To turn on this feature:
668 :
669 :			Compiler.Control.CG.memDisambiguate := true
670 :
671 :			To pretty print the region information with assembly
672 :
673 :			Compiler.Control.MLRISC.getFlag "asm-show-region" := true;
674 :
675 :			To control how many levels of aliasing information are printed, use:
676 :
677 :			Compiler.Control.MLRISC.getInt "points-to-show-level" := n
678 :
679 :			The default of n is 3.
680 :
681 :			----------------------------------------------------------------------
682 :	dbm	589	Name: David MacQueen
683 :			Date: 2000/03/31 11:15:00 EST
684 :			Tag: dbm-20000331-runtime_fix
685 :			Description:
686 :
687 :			This update contains:
688 :
689 :			1. runtime/c-lib/c-libraries.c
690 :			includes added in revision 1.2 caused compilation errors on hppa-hpux
691 :
692 :			2. fix for bug 1556
693 :			system/Basis/Implementation/NJ/internal-signals.sml
694 :
695 :			----------------------------------------------------------------------
696 :	blume	587	Name: Matthias Blume
697 :	blume	588	Date: 2000/03/31 18:00:00 JST
698 :			Tag: blume_main_v110p26p2_1
699 :			Description:
700 :
701 :			This update contains:
702 :
703 :			1. A small change to CM's handling of stable libraries:
704 :			CM now maintains one "global" modmap that is used for all stable
705 :			libraries. The use of such a global modmap maximizes sharing and
706 :			minimizes the need for re-traversing parts of environments during
707 :			modmap construction. (However, this has minor impact since modmap
708 :			construction seems to account for just one percent or less of total
709 :			compile time.)
710 :
711 :			2. I added a "genmap" phase to the statistics. This is where I got the
712 :			"one percent" number (see above).
713 :
714 :			3. CM's new tool parameter mechanism just became _even_ better. :)
715 :			- The parser understands named parameters and recursive options.
716 :			- The "make" and "shell" tools use these new features.
717 :			(This makes it a lot easier to cascade these tools.)
718 :			- There is a small syntax change: named parameters use a
719 :
720 :			<name> : ( <option> ... ) or
721 :			<name> : <string>
722 :
723 :			syntax. Previously, named parameters were implemented in an
724 :			ad-hoc fashion by each tool individually (by parsing strings)
725 :			and had the form
726 :
727 :			<name>=<string>
728 :
729 :			See the CM manual for a full description of these issues.
730 :
731 :			----------------------------------------------------------------------
732 :			Name: Matthias Blume
733 :	blume	587	Date: 2000/03/30 18:00:00 JST
734 :			Tag: blume_main_v110p26p2_0
735 :			Description:
736 :
737 :			!!!!! WARNING !!!!!!
738 :			!! New binfiles !!
739 :			!!!!!!!!!!!!!!!!!!!!
740 :
741 :			This update contains:
742 :
743 :			1. Moderate changes to CM:
744 :
745 :			- Changes to CM's tools mechanism. In particular, it is now possible
746 :			to have tools that accept additional "command line" parameters
747 :			(specified in the .cm file at each instance where the tool's class is
748 :			used).
749 :
750 :			This was done to accomodate the new "make" and "shell" tools which
751 :			facilitate fairly seemless hookup to portions of code managed using
752 :			Makefiles or Shell scripts.
753 :
754 :			There are no classes "shared" or "private" anymore. Instead, the
755 :			sharing annotation is now a parameter to the "sml" class.
756 :
757 :			There is a bit of generic machinery for implementing one's own
758 :			tools that accept command-line parameters. However, I am not yet fully
759 :			satisfied with that part, so expect changes here in the future.
760 :
761 :			All existing tools are described in the CM manual.
762 :
763 :			- Slightly better error handling. (CM now surpresses many followup
764 :			error messages that tended to be more annoying than helpful.)
765 :
766 :			2. Major changes to the compiler's static environment data structures.
767 :
768 :			- no CMStaticEnv anymore.
769 :			- no CMEnv, no "BareEnvironment" (actually, _only_ BareEnvironment,
770 :			but it is called Environment), no conversions between different
771 :			kinds of static environments
772 :
773 :			- There is still a notion of a "modmap", but such modmaps are generated
774 :			on demand at the time when they are needed. This sounds slow, but I
775 :			sped up the code that generates modmaps enough for this not to lead to
776 :			a slowdown of the compiler (at least I didn't detect any).
777 :
778 :			- To facilitate rapid modmap generation, static environments now
779 :			contain an (optional) "modtree" structure. Modtree annotations are
780 :			constructed by the unpickler during unpickling. (This means that
781 :			the elaborator does not have to worry about modtrees at all.)
782 :			Modtrees have the advantage that they are compositional in the same
783 :			way as the environment data structure itself is compositional.
784 :			As a result, modtrees never hang on to parts of an environment that
785 :			has already been rendered "stale" by filtering or rebinding.
786 :
787 :			- I went through many, many trials and errors before arriving at the
788 :			current solution. (The initial idea of "linkpaths" did not work.)
789 :			But the result of all this is that I have touched a lot of files that
790 :			depend on the "modules" and "types" data structures (most of the
791 :			elaborator). There were a lot of changes during my "linkpath" trials
792 :			that could have been reverted to their original state but weren't.
793 :			Please, don't be too harsh on me for messing with this code a bit more
794 :			than what was strictly necessary... (I _did_ resist the tempation
795 :			of doing any "global reformatting" to avoid an untimely death at
796 :			Dave's hands. :)
797 :
798 :			- One positive aspect of the previous point: At least I made sure that
799 :			all files that I touched now compile without warnings (other than
800 :			"polyEqual").
801 :
802 :			- compiler now tends to run "leaner" (i.e., ties up less memory in
803 :			redundant modmaps)
804 :
805 :			----------------------------------------------------------------------
806 :	leunga	580	Name: Allen Leung
807 :	leunga	585	Date: 2000/03/29 18:00:00
808 :			Tag: leunga-20000327-mlriscGen_hppa_alpha_x86
809 :	leunga	586	Boot files (optional): ftp://react-ilp.cs.nyu.edu/leunga/110.26.1-sml.boot.x86-unix-20000330.tar.gz
810 :	leunga	585	Description:
811 :
812 :			This update contains *MAJOR* changes to the way code is generated from CPS
813 :			in the module mlriscGen, and in various backend modules.
814 :
815 :			CHANGES
816 :			=======
817 :
818 :			1. MLRiscGen: forward propagation fix.
819 :
820 :			There was a bug in forward propagation introduced at about the same time
821 :			as the MLRISC x86 backend, which prohibits coalescing to be
822 :			performed effectively in loops.
823 :
824 :			Effect: speed up of loops in RISC architectures.
825 :			By itself, this actually slowed down certain benchmarks on the x86.
826 :
827 :			2. MLRiscGen: forward propagating addresses from consing.
828 :
829 :			I've changed the way consing code is generated. Basically I separated
830 :			out the initialization part:
831 :
832 :			store tag, offset(allocptr)
833 :			store elem1, offset+4(allocptr)
834 :			store elem2, offset+8(allocptr)
835 :			...
836 :			store elemn, offset+4n(allocptr)
837 :
838 :			and the address computation part:
839 :
840 :			celladdr <- offset+4+alloctpr
841 :
842 :			and move the address computation part
843 :
844 :			Effect: register pressure is generally lower as a result. This
845 :			makes compilation of certain expressions much faster, such as
846 :			long lists with non-trivial elements.
847 :
848 :			[(0,0), (0,0), .... (0,0)]
849 :
850 :			3. MLRiscGen: base pointer elimination.
851 :
852 :			As part of the linkage mechanism, we generate the sequence:
853 :
854 :			L: ... <- start of the code fragment
855 :
856 :			L1:
857 :			base pointer <- linkreg - L1 + L
858 :
859 :			The base pointer was then used for computing relocatable addresses
860 :			in the code fragment. Frequently (such as in lots of continuations)
861 :			this is not needed. We now eliminate this sequence whenever possible.
862 :
863 :			For compile time efficiency, I'm using a very stupid local heuristic.
864 :			But in general, this should be done as a control flow analysis.
865 :
866 :			Effect: Smaller code size. Speed up of most programs.
867 :
868 :			4. Hppa back end
869 :
870 :			Long jumps in span dependence resolution used to depend on the existence
871 :			of the base pointer.
872 :
873 :			A jump to a long label L was expanded into the following sequence:
874 :
875 :			LDIL %hi(L-8192), %r29
876 :			LDO %lo(L-8192)(%r29), %r29
877 :			ADD %r29, baseptr, %r29
878 :			BV,n %r0(%r29)
879 :
880 :			In the presence of change (3) above, this will not work. I've changed
881 :			it so that the following sequence of instructions are generated, which
882 :			doesn't mention the base pointer at all:
883 :
884 :			BL,n L', %r29 /* branch and link, L' + 4 -> %r29 */
885 :			L': ADDIL L-(L'+4), %r29 /* Compute address of L */
886 :			BV,n %r0(%r29) /* Jump */
887 :
888 :			5. Alpha back end
889 :
890 :			New alpha instructions LDB/LDW have been added, as per Fermin's
891 :			suggestions. This is unrelated to all other changes.
892 :
893 :			6. X86 back end
894 :
895 :			I've changed andl to testl in the floating point test sequence
896 :			whenever appropriate. The Intel optimization guide states that
897 :			testl is perferable to andl.
898 :
899 :			7. RA (x86 only)
900 :
901 :			I've improved the spill propagation algorithm, using an approximation
902 :			of maximal weighted independent sets. This seems to be necessary to
903 :			alleviate the negative effect in light of the slow down in (1).
904 :
905 :			I'll write down the algorithm one of these days.
906 :
907 :			8. MLRiscGen: frequencies
908 :
909 :			I've added an annotation that states that all call gc blocks have zero
910 :			execution frequencies. This improves register allocation on the x86.
911 :
912 :			BENCHMARKS
913 :			==========
914 :
915 :			I've only perform the comparison on 110.25.
916 :
917 :			The platforms are:
918 :
919 :			HPPA A four processor HP machine (E9000) with 5G of memory.
920 :			X86 A 300Hhz Pentium II with 128M of memory, and
921 :			SPARC An Ultra sparc 2 with 512M of memory.
922 :
923 :			I used the following parameters for the SML benchmarks:
924 :
925 :			@SMLalloc
926 :			HPPA 256k
927 :			SPARC 512k
928 :			X86 256k
929 :
930 :			COMPILATION TIME
931 :			----------------
932 :			Here are the numbers comparing the compilation times of the compilers.
933 :			I've only compared 110.25 compiling the new sources versus
934 :			a fixpoint version of the new compiler compiling the same.
935 :
936 :			110.25 New
937 :			Total Time in RA Spill+Reload Total Time In RA Spill+Reload
938 :			HPPA 627s 116s 2684+3584 599s 95s 1003+1879
939 :			SPARC 892s 173s 2891+3870 708s 116s 1004+1880
940 :			X86 999s 315s 94006+130691 987s 296s 108877+141957
941 :
942 :			110.25 New
943 :			Code Size Code Size
944 :			HPPA 8596736 8561421
945 :			SPARC 8974299 8785143
946 :			X86 9029180 8716783
947 :
948 :			So in summary, things are at least as good as before. Dramatic
949 :			reduction in compilation is obtained on the Sparc; I can't explain it,
950 :			but it is reproducible. Perhaps someone should try to reproduce this
951 :			on their own machines.
952 :
953 :			SML BENCHMARKS
954 :			--------------
955 :
956 :			On the average, all benchmarks perform at least as well as before.
957 :
958 :			HPPA Compilation Time Spill+Reload Run Time
959 :			110.25 New 110.25 New 110.25 New
960 :
961 :			barnesHut 3.158 3.015 4.75% 1+1 0+0 2.980 2.922 2.00%
962 :			boyer 6.152 5.708 7.77% 0+0 0+0 0.218 0.213 2.34%
963 :			count-graphs 1.168 1.120 4.32% 0+0 0+0 22.705 23.073 -1.60%
964 :			fft 0.877 0.792 10.74% 1+3 1+3 0.602 0.587 2.56%
965 :			knuthBendix 3.180 2.857 11.32% 0+0 0+0 0.675 0.662 2.02%
966 :			lexgen 6.190 5.290 17.01% 0+0 0+0 0.913 0.788 15.86%
967 :			life 0.803 0.703 14.22% 25+25 0+0 0.153 0.140 9.52%
968 :			logic 2.048 2.007 2.08% 6+6 1+1 4.133 4.008 3.12%
969 :			mandelbrot 0.077 0.080 -4.17% 0+0 0+0 0.765 0.712 7.49%
970 :			mlyacc 22.932 20.937 9.53% 154+181 32+57 0.468 0.430 8.91%
971 :			nucleic 5.183 5.060 2.44% 2+2 0+0 0.125 0.120 4.17%
972 :			ratio-regions 3.357 3.142 6.84% 0+0 0+0 116.225 113.173 2.70%
973 :			ray 1.283 1.290 -0.52% 0+0 0+0 2.887 2.855 1.11%
974 :			simple 6.307 6.032 4.56% 28+30 5+7 3.705 3.658 1.28%
975 :			tsp 0.888 0.862 3.09% 0+0 0+0 7.040 6.893 2.13%
976 :			vliw 24.378 23.455 3.94% 106+127 25+45 2.758 2.707 1.91%
977 :			--------------------------------------------------------------------------
978 :			Average 6.12% 4.09%
979 :
980 :			SPARC Compilation Time Spill+Reload Run Time
981 :			110.25 New 110.25 New 110.25 New
982 :
983 :			barnesHut 3.778 3.592 5.20% 2+2 0+0 3.648 3.453 5.65%
984 :			boyer 6.632 6.110 8.54% 0+0 0+0 0.258 0.242 6.90%
985 :			count-graphs 1.435 1.325 8.30% 0+0 0+0 33.672 34.737 -3.07%
986 :			fft 0.980 0.940 4.26% 3+9 2+6 0.838 0.827 1.41%
987 :			knuthBendix 3.590 3.138 14.39% 0+0 0+0 0.962 0.967 -0.52%
988 :			lexgen 6.593 6.072 8.59% 1+1 0+0 1.077 1.078 -0.15%
989 :			life 0.972 0.868 11.90% 26+26 0+0 0.143 0.140 2.38%
990 :			logic 2.525 2.387 5.80% 7+7 1+1 5.625 5.158 9.05%
991 :			mandelbrot 0.090 0.093 -3.57% 0+0 0+0 0.855 0.728 17.39%
992 :			mlyacc 26.732 23.827 12.19% 162+189 32+57 0.550 0.560 -1.79%
993 :			nucleic 6.233 6.197 0.59% 3+3 0+0 0.163 0.173 -5.77%
994 :			ratio-regions 3.780 3.507 7.79% 0+0 0+0 133.993 131.035 2.26%
995 :			ray 1.595 1.550 2.90% 1+1 0+0 3.440 3.418 0.63%
996 :			simple 6.972 6.487 7.48% 29+32 5+7 3.523 3.525 -0.05%
997 :			tsp 1.115 1.063 4.86% 0+0 0+0 7.393 7.265 1.77%
998 :			vliw 27.765 24.818 11.87% 110+135 25+45 2.265 2.135 6.09%
999 :			----------------------------------------------------------------------------
1000 :			Average 6.94% 2.64%
1001 :
1002 :			X86 Compilation Time Spill+Reload Run Time
1003 :			110.25 New 110.25 New 110.25 New
1004 :
1005 :			barnesHut 5.530 5.420 2.03% 593+893 597+915 3.532 3.440 2.66%
1006 :			boyer 8.768 7.747 13.19% 493+199 301+289 0.327 0.297 10.11%
1007 :			count-graphs 2.040 2.010 1.49% 298+394 315+457 26.578 28.660 -7.26%
1008 :			fft 1.327 1.302 1.92% 112+209 115+210 1.055 0.962 9.71%
1009 :			knuthBendix 5.218 5.475 -4.69% 451+598 510+650 0.928 0.932 -0.36%
1010 :			lexgen 9.970 9.623 3.60% 1014+841 1157+885 0.947 0.928 1.97%
1011 :			life 1.183 1.183 0.00% 162+182 145+148 0.127 0.103 22.58%
1012 :			logic 3.285 3.512 -6.45% 514+684 591+836 5.682 5.577 1.88%
1013 :			mandelbrot 0.147 0.143 2.33% 38+41 33+54 0.703 0.690 1.93%
1014 :			mlyacc 35.457 32.763 8.22% 3496+4564 3611+4860 0.552 0.550 0.30%
1015 :			nucleic 7.100 6.888 3.07% 239+168 201+158 0.175 0.173 0.96%
1016 :			ratio-regions 6.388 6.843 -6.65% 1182+257 981+300 120.142 120.345 -0.17%
1017 :			ray 2.332 2.338 -0.29% 346+398 402+494 3.593 3.540 1.51%
1018 :			simple 9.912 9.903 0.08% 1475+941 1579+1168 3.057 3.178 -3.83%
1019 :			tsp 1.623 1.532 5.98% 266+200 250+211 8.045 7.878 2.12%
1020 :			vliw 33.947 35.470 -4.29% 2629+2774 2877+3171 2.072 1.890 9.61%
1021 :			----------------------------------------------------------------------------
1022 :			Average 1.22% 3.36%
1023 :
1024 :			----------------------------------------------------------------------
1025 :			Name: Allen Leung
1026 :	leunga	583	Date: 2000/03/23 16:25:00
1027 :			Tag: leunga-20000323-fix_x86_alpha
1028 :			Description:
1029 :
1030 :			1. X86 fixes/changes
1031 :
1032 :			a. The old code generated for SETcc was completely wrong.
1033 :			The Intel optimization guide is VERY misleading.
1034 :
1035 :			2. ALPHA fixes/changes
1036 :
1037 :			a. Added the instructions LDBU, LDWU, STB, STW as per Fermin's suggestion.
1038 :			b. Added a new mode byteWordLoadStores to the functor parameter to Alpha()
1039 :			c. Added reassociation code for address computation.
1040 :
1041 :			----------------------------------------------------------------------
1042 :			Name: Allen Leung
1043 :	leunga	580	Date: 2000/03/22 01:23:00
1044 :			Tag: leunga-20000322-fix_x86_hppa_ra
1045 :			Description:
1046 :
1047 :			1. X86 fixes/changes
1048 :
1049 :			a. x86Rewrite bug with MUL3 (found by Lal)
1050 :			b. Added the instructions FSTS, FSTL
1051 :
1052 :			2. PA-RISC fixes/changes
1053 :
1054 :			a. B label should not be a delay slot candidate! Why did this work?
1055 :			b. ADDT(32, REG(32, r), LI n) now generates one instruction instead of two,
1056 :			as it should be.
1057 :			c. The assembly syntax for fstds and fstdd was wrong.
1058 :			d. Added the composite instruction COMICLR/LDO, which is the immediate
1059 :			operand variant of COMCLR/LDO.
1060 :
1061 :			3. Generic MLRISC
1062 :
1063 :			a. shuffle.sml rewritten to be slightly more efficient
1064 :			b. DIV bug in mltree-simplify fixed (found by Fermin)
1065 :
1066 :			4. Register Allocator
1067 :
1068 :			a. I now release the interference graph earlier during spilling.
1069 :			May improve memory usage.
1070 :
1071 :			----------------------------------------------------------------------
1072 :	blume	577	Name: Matthias Blume
1073 :	blume	578	Date: 2000/03/14 14:15:32
1074 :			Tag: blume_main_v110p26p1_2
1075 :			Description:
1076 :
1077 :			1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
1078 :			additional argument called "template" which is an optional string that
1079 :			specifiel the layout of the tool command line. See the CM manual for
1080 :			explanation.
1081 :
1082 :			2. A special-purpose tool can be "regisitered" by simply dropping the
1083 :			corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
1084 :			directory where the .cm file lives that uses this tool. (The
1085 :			behavior/misfeature until now was to look for the tool description
1086 :			files in the current working directory.) As before, tool description
1087 :			files could also be anchored -- in which case they can live anywhere
1088 :			they like. Following the recent e-mail discussion, this change should
1089 :			make it easier to have special-purpose tools that are shipped together
1090 :			with the sources of the program that uses them.
1091 :
1092 :			----------------------------------------------------------------------
1093 :			Name: Matthias Blume
1094 :	blume	577	Date: 2000/03/10 07:48:34
1095 :			Tag: blume_main_v110p26p1_1
1096 :			Description:
1097 :
1098 :			I added a re-written version of Dave's fixpt script to src/system.
1099 :			Changes relative to the original version:
1100 :			- sh-ified (not everybody has ksh)
1101 :			- automatically figures out which architecture it runs on
1102 :			- uses ./makeml a bit more cleverly
1103 :			- never invokes ./installml (and, thus, does not clobber your
1104 :			good and working installation of sml in case something goes wrong)
1105 :			- accepts max iteration count using option "-iter <n>"
1106 :			- accepts a "base" name using option "-base <base>"
1107 :
1108 :			It does not build any extraneous heap images but directly rebuilds
1109 :			bin- and boot-hierarchies using makeml's "-rebuild" switch. Finally,
1110 :			it can incorporate existing bin- and boot- hierarchies. For example,
1111 :			suppose the base is set to "sml" (which is the default). Then it
1112 :			successively builds
1113 :
1114 :			sml.bin.<arch>-unix and sml.boot.<arch>-unix
1115 :			then sml1.bin.<arch>-unix and sml1.boot.<arch>-unix
1116 :			then sml2.bin.<arch>-unix and sml2.boot.<arch>-unix
1117 :			...
1118 :			then sml<n>.bin.<arch>-unix and sml<n>.boot.<arch>-unix
1119 :
1120 :			and so on. If any of these already exist, it will just use what's
1121 :			there. In particular, many people will have the initial set of bin
1122 :			and boot files around, so this saves time for at least one full
1123 :			rebuild. Having sets of the form <base><k>.{bin,boot}.<arch>-unix for
1124 :			<k>=1,2,... is normally not a good idea when invoking fixpt. However,
1125 :			they might be the result of an earlier partial run of fixpt (which
1126 :			perhaps got accidentially killed). In this case, fixpt will quickly
1127 :			move through what exists before continuing where it left off earlier,
1128 :			and, thus, saves a lot of time.
1129 :
1130 :			----------------------------------------------------------------------
1131 :	leunga	576	Name: Allen Leung
1132 :			Date: 00/03/10 02:20:00
1133 :			Tag: leunga-20000310-fix_x86_asm_ra
1134 :			Description:
1135 :	dbm	570
1136 :	leunga	576	More assembly output problems involving the indexed addressing mode
1137 :			on the x86 have been found and corrected. Thanks to Fermin Reig for the
1138 :			fix.
1139 :
1140 :			The interface and implementation of the register allocator have been changed
1141 :			slightly to accommodate the possibility to skip the register allocation
1142 :			phases completely and go directly to memory allocation. This is needed
1143 :			for C-- use.
1144 :
1145 :	dbm	570	----------------------------------------------------------------------
1146 :	blume	572	Name: Matthias Blume
1147 :	blume	575	Date: 00/03/09 10:23:53
1148 :			Tag: blume_main_v110p26p1_0
1149 :			Description:
1150 :
1151 :			* Complete re-organization of library names. Many libraries have been
1152 :			consolidated so that they share the same path anchor. For example,
1153 :			all MLRISC-related libraries are anchored at MLRISC, most libraries that
1154 :			are SML/NJ-specific are under "smlnj". Notice that names like
1155 :			host-cmb.cm or host-compiler.cm no longer exist. See system/README
1156 :			for a complete description of the new naming scheme. Quick reference:
1157 :
1158 :			host-cmb.cm -> smlnj/cmb.cm
1159 :			host-compiler.cm -> smlnj/compiler.cm
1160 :			full-cm.cm -> smlnj/cm.cm
1161 :			<arch>-<os>.cm -> smlnj/cmb/<arch>-<os>.cm
1162 :			<arch>-compiler.cm -> smlnj/compiler/<arch>.cm
1163 :
1164 :			* Bug fixes in CM.
1165 :			- exceptions in user code are being passed through (i.e., reach top level)
1166 :			- more bugs in paranoia mode fixed
1167 :			- bug related to checking group owners fixed
1168 :
1169 :			* New install.sh script that automagically fetches archive files:
1170 :			The new file config/srcarchiveurl must contain the URL of the
1171 :			(remote) directory that contains bin files (or other source archives).
1172 :			If install.sh does not find the archive locally, it tries to get
1173 :			it from that remote directory.
1174 :			This should simplify installation further: For machines that have
1175 :			access to the internet, just fetch <version>-config.tgz, unpack it,
1176 :			edit config/targets, and go (run config/install.sh). The scipt will
1177 :			fetch everything else that it might need all by itself.
1178 :
1179 :			For CVS users, this mechanism is not relevant for source archives, but
1180 :			it is convenient for getting new sets of binfiles.
1181 :
1182 :			Archives should be tar files compressed with either gzip, compress, or
1183 :			bzip2. The script recognizes .tgz, .tar, tar.gz, tz, .tar.Z, and .tar.bz2.
1184 :
1185 :			----------------------------------------------------------------------
1186 :			Name: Matthias Blume
1187 :	blume	572	Date: 2000/03/07 04:01:04
1188 :			Tag: blume_main_v110_26_2
1189 :	dbm	570	Description:
1190 :	blume	572	- size info in BOOTLIST
1191 :			* no fixed upper limits for number of bootfiles or length of
1192 :			bootfile names in runtime
1193 :			* falling back to old behavior if no BOOTLIST size info found
1194 :			- allocation size heuristics in .run-sml
1195 :			* tries to read cache size from /proc/cpuinfo (this is important for
1196 :			small-cache Celeron systems!)
1197 :			- install.sh robustified
1198 :			- CM manual updates
1199 :			- paranoid mode
1200 :			* no more CMB.deliver() (i.e., all done by CMB.make())
1201 :			* can re-use existing sml.boot.* files
1202 :			* init.cmi now treated as library
1203 :			* library stamps for consistency checks
1204 :			- sml.boot.<arch>-<os>/PIDMAP file
1205 :			* This file is read by the CM startup code. This is used to minimize
1206 :			the amount of dynamic state that needs to be stowed away for the
1207 :			purpose of sharing between interactive system and user code.
1208 :			- CM.Anchor.anchor instead of CM.Anchor.{set,cancel}
1209 :			* Upon request by Elsa. Anchors now controlled by get-set-pair
1210 :			like most other CM state variables.
1211 :			- Compiler.CMSA eliminated
1212 :			* No longer supported by CM anyway.
1213 :			- fixed bugs in pickler that kept biting Stefan
1214 :			* past refs to past refs (was caused by the possibility that
1215 :			ad-hoc sharing is more discriminating than hash-cons sharing)
1216 :			* integer overflow on LargeInt.minInt
1217 :			- ml-{lex,yacc} build scripts now use new mechanism
1218 :			for building standalone programs
1219 :			- fixed several gcc -Wall warnings that were caused by missing header
1220 :			files, missing initializations, etc., in runtime (not all warnings
1221 :			eliminated, though)

---

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