Home My Page Projects Code Snippets Project Openings SML/NJ
Summary Activity Forums Tracker Lists Tasks Docs Surveys News SCM Files

SCM Repository

[smlnj] Annotation of /sml/trunk/NOTES/HISTORY
ViewVC logotype

Annotation of /sml/trunk/NOTES/HISTORY

Parent Directory Parent Directory | Revision Log Revision Log


Revision 676 - (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 675 Name: Matthias Blume
16 : blume 676 Date: 2000/06/24 12:40:00 JST
17 :     Tag: blume-20000624-startup
18 :     Description:
19 :    
20 :     Fixes startup slowdown problem. (I was calling SrcPath.sync a _tad_
21 :     bit too often -- to put it mildly. :)
22 :    
23 :     ----------------------------------------------------------------------
24 :     Name: Matthias Blume
25 : blume 675 Date: 2000/06/23 18:20:00 JST
26 :     Tag: blume-20000623-btrace
27 :     Description:
28 :    
29 :     This updates adds a backtrace facility to aid programmers in debugging
30 :     their programs. This involves the following changes:
31 :    
32 :     1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
33 :     keeping track of the current call stack. When programs are compiled
34 :     in a special mode, the compiler will insert calls to these hooks
35 :     into the user program.
36 :     "Hook" means that it is possible for different implementations of
37 :     back-tracing to register themselves (at different times).
38 :    
39 :     2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
40 :     as an Absyn.dec->Absyn.dec rewrite. Normally this phase is turned off.
41 :     It can be turned on using this call:
42 :     SMLofNJ.Internals.BTrace.mode (SOME true);
43 :     Turning it off again:
44 :     SMLofNJ.Internals.BTrace.mode (SOME false);
45 :     Querying the current status:
46 :     SMLofNJ.Internals.BTrace.mode NONE;
47 :     Annotated programs are about twice as big as normal ones, and they
48 :     run a factor of 2 to 4 slower with a dummy back-trace plugin (one
49 :     where all hooks do nothing). The slowdown with a plugin that is
50 :     actually useful (such as the one supplied by default) is even greater,
51 :     but in the case of the default plugin it is still only an constant
52 :     factor (amortized).
53 :    
54 :     3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
55 :     with a sub-structure BTrace for controlling back-tracing. In particular,
56 :     the above-mentioned function "mode" controls whether the annotation
57 :     phase is invoked by the compiler. Another important function is
58 :     "trigger": when called it aborts the current execution and causes
59 :     the top-level loop to print a full back-trace.
60 :    
61 :     4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
62 :     for back-tracing. It keeps track of the dynamic call stack and in
63 :     addition to that it keeps a partial history at each "level" of that
64 :     stack. For example, if a tail-calls b, b tail-calls c, and c tail-calls
65 :     d and b (at separate times, dynamically), then the report will show:
66 :    
67 :     GOTO d
68 :     /c
69 :     GOTO \b
70 :     CALL a
71 :    
72 :     This shows that there was an initial non-tail call of a, then a
73 :     tail-call to b or c, looping behavior in a cluster of functions that
74 :     consist of b and c, and then a goto from that cluster (i.e., either from
75 :     b or from c) to d.
76 :    
77 :     Note that (depending on the user program) the amount of information
78 :     that the back-trace module has to keep track of at each level is bounded
79 :     by a constant. Thus, the whole implementation has the same asymptotical
80 :     complexity as the original program (both in space and in time).
81 :    
82 :     5. compiler/TopLevel/interact/evalloop.sml has been modified to
83 :     handle the special exception SMLofNJ.Internals.BTrace.BTrace
84 :     which is raised by the "trigger" function mentioned above.
85 :    
86 :     Notes on usage:
87 :    
88 :     - Annotated code works well together with unannotated code:
89 :     Unannotated calls simply do not show up at all in the backtrace.
90 :    
91 :     - It is not a good idea to let modules that were annotated during
92 :     different sessions run at the same time. This is because the compiler
93 :     chooses small integers to identify individual functions, and there
94 :     will be clashes if different modules were compiled in separate sessions.
95 :     (Nothing will crash, and you will even be told about the clashes, but
96 :     back-trace information will in general not be useful.)
97 :    
98 :     - Back-tracing can be confused by callcc and capture.
99 :    
100 :     - The only way of getting a back-trace right now is to explicitly
101 :     invoke the "trigger" function from your user program. Eventually, we
102 :     should make every exception carry back-trace information (if
103 :     available). But since this creates more overhead at "raise"-time
104 :     (similar to the current exnHistory overhead), I have not yet
105 :     implemented this. (The implementation will be rather easy.) With
106 :     exceptions carrying back-trace information, this facility will be even
107 :     more useful because users don't need to modify their programs...
108 :    
109 :     - While it is possible to compile the compiler with back-trace
110 :     annotations turned on (I did it to get some confidence in
111 :     correctness), you must make absolutely sure that core.sml and
112 :     btimp.sml are compiled WITHOUT annotation! (core.sml cannot actually
113 :     be compiled with annotation because there is no core access yet, but
114 :     if you compile btimp.sml with annotation, then the system will go into
115 :     an infinite recursion and crash.)
116 :     Since CM currently does not know about BTrace, the only way to turn
117 :     annotations on and off for different modules of the compiler is to
118 :     interrupt CMB.make, change the settings, and re-invoke it. Of course,
119 :     this is awkward and clumsy.
120 :    
121 :     Sample sessions:
122 :    
123 :     Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
124 :     - SMLofNJ.Internals.BTrace.mode (SOME true);
125 :     [autoloading]
126 :     [autoloading done]
127 :     val it = false : bool
128 :     - structure X = struct
129 :     - fun main n = let
130 :     - fun a (x, 0) = d x
131 :     - | a (x, n) = b (x, n - 1)
132 :     - and b (x, n) = c (x, n)
133 :     - and c (x, n) = a (x, n)
134 :     - and d x = e (x, 3)
135 :     - and e (x, 0) = f x
136 :     - | e (x, n) = e (x, n - 1)
137 :     - and f 0 = SMLofNJ.Internals.BTrace.trigger ()
138 :     - | f n = n * g (n - 1)
139 :     - and g n = a (n, 3)
140 :     - in
141 :     - f n
142 :     - end
143 :     - end;
144 :     structure X : sig val main : int -> int end
145 :     - X.main 3;
146 :     *** BACK-TRACE ***
147 :     GOTO stdIn:4.2-13.20: X.main[2].f
148 :     GOTO-( stdIn:4.2-13.20: X.main[2].e
149 :     GOTO stdIn:4.2-13.20: X.main[2].d
150 :     / stdIn:4.2-13.20: X.main[2].a
151 :     | stdIn:4.2-13.20: X.main[2].b
152 :     GOTO-\ stdIn:4.2-13.20: X.main[2].c
153 :     CALL stdIn:4.2-13.20: X.main[2].g
154 :     GOTO stdIn:4.2-13.20: X.main[2].f
155 :     GOTO-( stdIn:4.2-13.20: X.main[2].e
156 :     GOTO stdIn:4.2-13.20: X.main[2].d
157 :     / stdIn:4.2-13.20: X.main[2].a
158 :     | stdIn:4.2-13.20: X.main[2].b
159 :     GOTO-\ stdIn:4.2-13.20: X.main[2].c
160 :     CALL stdIn:4.2-13.20: X.main[2].g
161 :     GOTO stdIn:4.2-13.20: X.main[2].f
162 :     GOTO-( stdIn:4.2-13.20: X.main[2].e
163 :     GOTO stdIn:4.2-13.20: X.main[2].d
164 :     / stdIn:4.2-13.20: X.main[2].a
165 :     | stdIn:4.2-13.20: X.main[2].b
166 :     GOTO-\ stdIn:4.2-13.20: X.main[2].c
167 :     CALL stdIn:4.2-13.20: X.main[2].g
168 :     GOTO stdIn:4.2-13.20: X.main[2].f
169 :     CALL stdIn:2.15-17.4: X.main[2]
170 :     -
171 :    
172 :     (Note that because of a FLINt bug the above code currently does not
173 :     compile without BTrace turned on.)
174 :    
175 :     Here is another example, using my modified Tiger compiler:
176 :    
177 :     Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
178 :     - SMLofNJ.Internals.BTrace.mode (SOME true);
179 :     [autoloading]
180 :     [autoloading done]
181 :     val it = false : bool
182 :     - CM.make "sources.cm";
183 :     [autoloading]
184 :     ...
185 :     [autoloading done]
186 :     [scanning sources.cm]
187 :     [parsing (sources.cm):parse.sml]
188 :     [creating directory CM/SKEL ...]
189 :     [parsing (sources.cm):tiger.lex.sml]
190 :     ...
191 :     [wrote CM/sparc-unix/semant.sml]
192 :     [compiling (sources.cm):main.sml]
193 :     [wrote CM/sparc-unix/main.sml]
194 :     [New bindings added.]
195 :     val it = true : bool
196 :     - Main.compile ("../testcases/merge.tig", "foo.out");
197 :     *** BACK-TRACE ***
198 :     CALL lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
199 :     CALL lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
200 :     CALL lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
201 :     GOTO lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
202 :     CALL lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
203 :     CALL lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
204 :     CALL lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
205 :     / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
206 :     CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
207 :     CALL lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
208 :     CALL lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
209 :     CALL main.sml:1.18-118.4: Main.compile[2]
210 :     -
211 :    
212 :     ----------------------------------------------------------------------
213 : blume 673 Name: Matthias Blumen
214 :     Date: 2000/06/21 18:00:00 JST
215 :     Tag: blume-20000621-manual
216 :     Description:
217 :    
218 :     CM manual update: Path environments documented.
219 :    
220 :     ----------------------------------------------------------------------
221 : blume 666 Name: Matthias Blume
222 : blume 672 Date: 2000/06/19 13:40:00
223 :     Tag: blume-20000619-manual
224 :     Description:
225 :    
226 :     CM manual and system/README update. This only covers the fact that
227 :     there are no more implicit anchors. (Path environments and the "bind"
228 :     option to "cm" have yet to be documented.)
229 :    
230 :     ----------------------------------------------------------------------
231 :     Name: Matthias Blume
232 : blume 671 Date: 2000/06/19 11:05:00 JST
233 :     Tag: blume-20000619-chdir-bugfix
234 :     Description:
235 :    
236 :     Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
237 :    
238 :     ----------------------------------------------------------------------
239 :     Name: Matthias Blume
240 : blume 670 Date: 2000/06/18 22:00:10 JST
241 :     Tag: blume-20000618-implicit-anchors-really-gone
242 :     Description:
243 :    
244 :     I updates the previous HISTORY entry where I forgot to mention that
245 :     implicit anchors are no longer with us.
246 :    
247 :     The current update also gets rid of the (now useless) controller
248 :     CM.Control.implicit_anchors.
249 :    
250 :     ----------------------------------------------------------------------
251 :     Name: Matthias Blume
252 : blume 666 Date: 2000/06/16 17:30:00 JST
253 :     Tag: blume-20000616-anchorenv
254 :     Description:
255 :    
256 :     This patch implements the long anticipated (just kidding :) "anchor
257 :     environment" mechanism. In the course of doing this, I also
258 :     re-implemented CM's internal "SrcPath" module from scratch. The new
259 :     one should be more robust in certain boundary cases. In any case, it
260 :     is a lot cleaner than its predecessor (IMHO).
261 :    
262 :     This time, although there is yet another boot file format change, I
263 :     kept the unpickler backward-compatible. As a result, no new bootfiles
264 :     are necessary and bootstrapping is straightforward. (You cannot read
265 :     new bootfiles into an old system, but the other way around is no
266 :     problem.)
267 :    
268 :     Visible changes:
269 :    
270 : blume 670 ** 0. Implicit path anchors (without the leading $-symbol) are no
271 :     longer recognized at all. This means that such path names are not
272 :     illegal either. For example, the name basis.cm simply refers to a
273 :     local file called "basis.cm" (i.e, the name is an ordinary path
274 :     relative to .cm-files directory). Or, to put it differently, only
275 :     names that start with $ are anchored paths.
276 : blume 666
277 : blume 670 ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
278 :     vanished.
279 :    
280 :     John (Reppy) had critizised this as soon as I originally proposed and
281 : blume 666 implemented it, but at that time I did not really deeply believe
282 :     him. :) Now I came full-circle because I need the $<singlearc> syntax
283 :     in another place where it cannot be seen as an abbreviation for
284 :     $/<singlearc>. To avoid the confusion, $<singlearc> now means what it
285 :     seems to mean (i.e., it "expands" into the corresponding anchor
286 :     value).
287 :    
288 :     However, when paths are used as members in CM description files, it
289 :     continues to be true that there must be at least another arc after the
290 :     anchor. This is now enforced separately during semantic analysis
291 : blume 670 (i.e., from a lexical/syntactical point of view, the notation is ok.)
292 : blume 666
293 :     ** 2. The "cm" class now accepts an option "bind". The option's value
294 :     is a sub-option list of precisely two items -- one labeled "anchor"
295 :     and the other one labeled "value". As you might expect, "anchor" is
296 :     used to specify an anchor name to be bound, and "value" specifies what
297 :     the anchor is being bound to.
298 :    
299 :     The value must be a directory name and can be given in either standard
300 :     syntax (including the possibility that it is itself an anchored path)
301 :     or native syntax.
302 :    
303 :     Examples:
304 :    
305 :     foo.cm (bind:(anchor:bar value:$mystuff/bar))
306 :     lib.cm (bind:(anchor:a value:"H:\\x\\y\\z")) (* only works under windows *)
307 :    
308 :     and so on.
309 :    
310 :     The meaning of this is that the .cm-file will be processed with an
311 :     augmented anchor environment where the given anchor(s) is/are bound to
312 :     the given values(s).
313 :    
314 :     The rationale for having this feature is this: Suppose you are trying
315 :     to use two different (already stable) libraries a.cm and b.cm (that
316 :     you perhaps didn't write yourself). Further, suppose each of these
317 :     two libraries internally uses its own auxiliary library $aux/lib.cm.
318 :     Normally you would now have a problem because the anchor "lib" can not
319 :     be bound to more than one value globally. Therefore, the project that
320 :     uses both a.cm and b.cm must locally redirect the anchor to some other
321 :     place:
322 :    
323 :     a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
324 :     b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
325 :    
326 :     This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
327 :     /usr/lib/smlnj/b-stuff/aux.cm, respectively.
328 :    
329 :     Hard-wiring path names is a bit inflexible (and CM will verbosely warn
330 :     you when you do so at the time of CM.stabilize). Therefore, you can
331 :     also use an anchored path as the value:
332 :    
333 :     a.cm (bind:(anchor:lib value:$a-lib))
334 :     b.cm (bind:(anchor:lib value:$b-lib))
335 :    
336 :     Now you can globally configure (using the usual CM.Anchor.anchor or
337 :     pathconfig machinery) bindings for "a-lib" and "b-lib". Since "lib"
338 :     itself is always locally bound, setting it globally is no longer
339 :     meaningful or necessary (but it does not hurt either). In fact, "lib"
340 :     can still be used as a global anchor for separate purposes. As a
341 :     matter of fact, one can locally define "lib" in terms of a global
342 :     "lib":
343 :    
344 :     a.cm (bind:(anchor:lib value:$lib/a))
345 :     b.cm (bind:(anchor:lib value:$lib/b))
346 :    
347 :     ** 3: The encoding of path names has changed. This affects the way
348 :     path names are shown in CM's progress report and also the internal
349 :     protocol encoding used for parallel make.
350 :    
351 :     The encoding now uses one or more ':'-separated segments. Each
352 :     segments corresponds to a file that has been specified relative to the
353 :     file given by its preceding segment. The first segment is either
354 :     relative to the CWD, absolute, or anchored. Each segment itself is
355 :     basically a Unix pathname; all segments but the first are relative.
356 :    
357 :     Example:
358 :    
359 :     $foo/bar/baz.cm:a/b/c.sml
360 :    
361 :     This path denotes the file bar/a/b/c.sml relative to the directory
362 :     denoted by anchor "foo". Notice that the encoding also includes
363 :     baz.cm which is the .cm-file that listed a/b/c.sml. As usual, such
364 :     paths are resolved relative to the .cm-files directory, so baz.cm must
365 :     be ignored to get the "real" pathname.
366 :    
367 :     To make this fact more obvious, CM puts the names of such "virtual
368 :     arcs" into parentheses when they appear in progress reports. (No
369 :     parentheses will appear in the internal protocol encoding.) Thus,
370 :     what you really see is:
371 :    
372 :     $foo/bar/(baz.cm):a/b/c.sml
373 :    
374 :     I find this notation to be much more informative than before.
375 :    
376 :     Another new feature of the encoding is that special characters
377 :     including parentheses, colons, (back)slashes, and white space are
378 :     written as \ddd (where ddd is the decimal encoding of the character).
379 :    
380 : blume 670 *** The CM manual still needs to be updated.
381 :    
382 : blume 666 ----------------------------------------------------------------------
383 : leunga 660 Name: Allen Leung
384 :     Date: 2000/06/15 00:38:00
385 :     Tag: leunga-20000615-x86-peephole
386 :    
387 :     x86 Peephole fix by Fermin. Affects c-- and moby only.
388 :    
389 :     ----------------------------------------------------------------------
390 : blume 658 Name: Matthias Blume
391 :     Date: 2000/06/12 11:40:00
392 :     Tag: blume-20000612-parmakefix
393 :     Description:
394 :    
395 :     More cleanup after changing the file naming scheme: This time I
396 :     repaired the parallel make mechanism for CMB.make which I broke earlier.
397 :    
398 :     ----------------------------------------------------------------------
399 : leunga 657 Name: Allen Leung
400 :     Date: 2000/06/09 01:25:00
401 :     Tag: leunga-20000609-various
402 :    
403 :     None of these things should affect normal SML/NJ operations
404 :    
405 :     1. Peephole improvements provided by Fermin (c--)
406 :     2. New annotation DEFUSE for adding extra dependence (moby)
407 :     3. New X86 LOCK instructions (moby)
408 :     4. New machine description language for reservation tables (scheduling)
409 :     5. Fixes to various optimization/analysis modules (branch chaining, dominator
410 :     trees etc.)
411 :     6. I've changed the CM files so that they can work with versions
412 :     110.0.6, 110.25 and 110.28
413 :    
414 :     ----------------------------------------------------------------------
415 : blume 652 Name: Matthias Blume
416 : blume 656 Date: 2000/06/09 12:40:00
417 :     Tag: blume-20000609-log
418 :     Description:
419 :    
420 :     - Removed all(?) remaining RCS Log entries from sources.
421 :    
422 :     - Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
423 :     anchored paths).
424 :    
425 :     ----------------------------------------------------------------------
426 :     Name: Matthias Blume
427 : blume 653 Date: 2000/06/07 17:00:00 JST
428 :     Tag: blume-20000607-no-implicit-anchors
429 :     Description:
430 :    
431 :     1. This update changes the default setting for
432 :     CM.Control.implicit_anchors from true to false. This means that
433 :     implicit anchors are no longer permitted by default. I also tried to
434 :     make sure that nothing else still relies on implicit anchors.
435 :     (This is the next step on the schedule towards a CM that does not even
436 :     have the notion of implicit anchors anymore.)
437 :    
438 :     2. More CM manual updates.
439 :    
440 :     3. I managed to track down and fix the pickling bug I mentioned last
441 :     time. Because of the previously existing workaround, this entails no
442 :     immediate practical changes.
443 :    
444 :     ----------------------------------------------------------------------
445 :     Name: Matthias Blume
446 : blume 652 Date: 2000/06/06 11:15:00 JST
447 :     Tag: blume-20000606-lazierpickle
448 :     Description:
449 :    
450 :     !!!! NEW BOOT FILES !!!!
451 :    
452 :     * The main purpose of this update is to make library pickles lazier in
453 :     order to reduce the initial space penalty for autoloading a library.
454 :     As a result, it is now possible to have $smlnj/compiler.cm
455 :     pre-registered. This should take care of the many complaints or
456 :     inquiries about missing structure Compiler. This required changes to
457 :     CM's internal data structures and small tweaks to some algorithms.
458 :    
459 :     As a neat additional effect, it is no longer necessary (for the sake
460 :     of lean heap image files) to distinguish between a "minimal" CM and a
461 :     "full" CM. Now, there is only one CM (i.e., the "full" version:
462 :     $smlnj/cm.cm aka $smlnj/cm/full.cm), and it is always available at the
463 :     interactive top level. ($smlnj/cm/minimal.cm is gone.)
464 :    
465 :     To make the life of compiler-hackers easier, "makeml" now also
466 :     pre-registers $smlnj/cmb.cm (aka $smlnj/cmb/current.cm). In other
467 :     words, after you bootstrap a new sml for the first time, you will not
468 :     have to autoload $smlnj/cmb.cm again afterwards. (The first time
469 :     around you will still have to do it, though.)
470 :    
471 :     * A second change consists of major updates to the CM manual. There
472 :     are now several appendices with summary information and also a full
473 :     specification of the CM description file syntax.
474 :    
475 :     * In directory src/system I added the script "allcross". This script
476 :     invokes sml and cross-compiles the compiler for all supported
477 :     architectures. (Useful when providing a new set of boot files.)
478 :    
479 :     * There seems to be a latent bug in my "lazy pickles" mechanism. I
480 :     added a small tweak to pickle-util.sml to work around this problem,
481 :     but it is not a proper fix yet. I will investigate further. (The
482 :     effect of the bug was an inflation of library pickle size.)
483 :    
484 :     * Version number increased to 110.28.1 (to avoid compatibility problems).
485 :    
486 :     ----------------------------------------------------------------------
487 : leunga 646 Name: Allen Leung
488 : leunga 648 Date: 2000/05/25 17:28 EDT
489 :     Tag: leunga-20000525-ra
490 :     Description:
491 :    
492 :     Fixed a bug in freezing phase of the register allocator.
493 :    
494 :     ----------------------------------------------------------------------
495 :     Name: Allen Leung
496 : leunga 647 Date: 2000/05/15 22:53 EDT
497 : leunga 646 Tag: leunga-20000515-alpha-x86-ra
498 :     Description:
499 :    
500 :     1. Alpha
501 :    
502 :     Slight cleanup. Removed the instruction SGNXL
503 :    
504 :     2. X86
505 :    
506 :     Added the following instructions to the instruction set:
507 :    
508 :     ROLx, RORx,
509 :     BTx, BTSx, BTLx, BTRx,
510 :     XCHGx, and variants with the LOCK prefix
511 :    
512 :     3. Register Allocation
513 :    
514 :     The module ra-rewrite-with-renaming has been improved.
515 :    
516 : leunga 647 These have no effect on SML/NJ.
517 :    
518 : leunga 646 ----------------------------------------------------------------------
519 : blume 642 Name: Matthias Blume
520 : blume 645 Date: 2000/05/15 16:20:00 JST
521 :     Tag: blume-20000515-lightrebuild
522 :     Description:
523 :    
524 :     1. I added an alternative to "-rebuild" to "makeml". The difference is
525 :     that prior to calling CMB.make' the CM-variable "LIGHT" will be
526 :     defined. In effect, the command will not build any cross-compiler
527 :     backends and therefore finish more quickly.
528 :    
529 :     The "fixpt" script also takes a "-light" switch to be able to use
530 :     this new facility while compiling for a fixpoint.
531 :    
532 :     2. I replaced all mentions of anchored paths in group owner specifications
533 :     with simple relative paths (usually starting with "..").
534 :     The rationale is that a library's internal workings should not be
535 :     compromised by the lack of some anchor. (An anchor is necessary
536 :     for someone who wants to refer to the library by an anchored path,
537 :     but it should not be necessary to build the same library in the first
538 :     place.)
539 :    
540 :     3. I changed the way CM's tool mechanism determines the shell command
541 :     string used for things like ml-yacc etc. so that it does not break
542 :     when CM.Control.implicit_anchors is turned off.
543 :    
544 :     ----------------------------------------------------------------------
545 :     Name: Matthias Blume
546 : blume 644 Date: 2000/05/12 18:20:00 JST
547 :     Tag: blume-20000512-ml-build
548 :     Description:
549 :    
550 :     Fixed a bug in config/_ml-build that prevented ml-yacc and ml-lex from
551 :     getting installed properly (by config/install.sh).
552 :    
553 :     ----------------------------------------------------------------------
554 :     Name: Matthias Blume
555 : blume 643 Date: 2000/05/12 17:30:00 JST
556 :     Tag: blume-20000512-anchors
557 :     Description:
558 :    
559 :     !!! NEW BOOT FILES !!!
560 :    
561 :     This change is in preparation of fading out support for "implicitly
562 :     anchored path names". I went through all sources and used the
563 :     explicit (and relatively new) $-notation. See system/README and the
564 :     CM manual for more info on this.
565 :    
566 :     I also modified the anchoring scheme for some things such as "smlnj",
567 :     "MLRISC", "cm", etc. to take advantage of the fact that explicit
568 :     anchors are more expressive: anchor name and first arc do not have to
569 :     coincide. This entails the following user-visible change:
570 :    
571 :     You have to write $smlnj/foo/bar instead of smlnj/foo/bar. In
572 :     particular, when you fire up sml with a command-line argument, say,
573 :     e.g.:
574 :    
575 :     sml '$smlnj/cmb.cm'
576 :    
577 :     At the ML toplevel prompt:
578 :    
579 :     CM.autoload "$smlnj/cmb.cm";
580 :    
581 :     There is also a new controller in CM.Control that can be used to turn
582 :     off all remaining support for implicit anchors by saying:
583 :    
584 :     CM.autoload "$smlnj/
585 :     #set CM.Control.implicit_anchors false;
586 :    
587 :     This causes CM to reject implicitly anchored paths. This is (for the
588 :     time being) less permissive than the "final" version where there will
589 :     be no more such implicit anchors and relative paths will be just that:
590 :     relative.
591 :    
592 :     The next step (version after next version?) will be to make the
593 :     default for CM.Control.implicit_anchors false. After the dust has
594 :     settled, I can then produce the "final" version of this...
595 :    
596 :     Note: Since bootstrapping is a bit tricky, I provided new boot files.
597 :    
598 :     ----------------------------------------------------------------------
599 :     Name: Matthias Blume
600 : blume 642 Date: 2000/05/11 16:30:00 JST
601 :     Tag: blume-20000511-sources
602 :     Description:
603 :    
604 :     The main change is that I added function CM.sources as a generalized
605 :     version of the earlier CM.makedepend. This entails the following
606 :     additional changes:
607 :    
608 :     - CM.makedepend has been dropped.
609 :    
610 :     - CM manual has been updated.
611 :    
612 :     - TOOLS signature and API have been changed.
613 :    
614 :     ----------------------------------------------------------------------
615 : leunga 640 Name: Allen Leung
616 : leunga 641 Date: 2000/05/10 21:17 EDT
617 :     Tag: leunga-20000510-moby-c--ssa
618 :     Description:
619 :    
620 :     Various bug fixes and new features for C--, Moby and MLRISC optimizations.
621 :     None of these affect SML/NJ.
622 :    
623 :     1. Register Allocation
624 :    
625 :     a. A new ra spilling module (ra/ra-spill-with-renaming) is implemented.
626 :     This module tries to remove local (i.e. basic block level) redundancies
627 :     during spilling.
628 :    
629 :     b. A new framework for performing region based register allocation.
630 :     Not yet entirely functional.
631 :    
632 :     2. X86
633 :    
634 :     a. DefUse for POP was missing the stack pointer [found by Lal]
635 :     b. Reload for CALL was incorrect in X86Spill [found by John]
636 :     c. Various fixes in X86Spill so that it can be used correctly for
637 :     the new spilling module.
638 :    
639 :     3. SSA/IR
640 :    
641 :     a. New module ir/dj-dataflow.sml implements elimination based
642 :     data flow analysis.
643 :    
644 :     4. MLRiscGen
645 :    
646 :     a. Fix for gc type annotation
647 :    
648 :     5. MDGen
649 :    
650 :     Various fixes for machine description -> ml code translation. For ssa
651 :     only.
652 :    
653 :     ----------------------------------------------------------------------
654 :     Name: Allen Leung
655 : leunga 640 Date: 2000/05/08 22:17 EDT
656 :     Tag: leunga-20000508-labexp
657 :     Description:
658 :    
659 :     Fermin has found a few assembly problems with constant expressions
660 :     generated in LabelExp. Mostly, the problems involve extra parentheses,
661 :     which choke on dumb assemblers. This is his fix.
662 :    
663 :     ----------------------------------------------------------------------
664 : dbm 635 Name: Dave MacQueen
665 :     Date: 2000/04/09 14:00 EDT
666 :     Tag: dbm-20000502-Version_110_28
667 :     Description:
668 :    
669 :     1. Updated src/compiler/TopLevel/main/version.sml to version 110.28
670 :    
671 :     2. Updated config/version to 110.28
672 :    
673 :     3. Updated config/srcarchiveurl
674 :    
675 :     3. New boot files!
676 :     ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.28/
677 :    
678 :     ----------------------------------------------------------------------
679 : blume 634 Name: Matthias Blume
680 :     Date: 2000/05/01 19:05:00 JST
681 :     Tag: blume-20000501-noweb
682 :     Description:
683 :    
684 :     A new noweb tool has been added. The existing system is entirely
685 :     unaffected by this, but some CM users have asked for renewed noweb
686 :     support. Everything is documented in the CM manual.
687 :    
688 :     New (plugin) libraries:
689 :    
690 :     noweb-tool.cm
691 :     nw-ext.cm
692 :    
693 :     ----------------------------------------------------------------------
694 : dbm 633 Name: Dave MacQueen
695 :     Date: 2000/04/30 12:40PM EDT
696 :     Tag: dbm-20000430-bug_fixes
697 :     Description:
698 :    
699 :     1. Fix for bug 1498
700 :     smlnj/src/system/Basis/Implementation/Unsafe/object.sig
701 :     smlnj/src/system/Basis/Implementation/Unsafe/object.sml
702 :     added toRealArray function
703 :     smlnj/src/compiler/MiscUtil/print/ppobj.sml
704 :     added check for tag Obj.RealArray to array printing case in ppObj
705 :    
706 :     2. Fix for bug 1510
707 :     smlnj/src/compiler/Semant/types/typesutil.sml
708 :     fixed definition of dummyargs (used by equalTycon) so that
709 :     dummy args are distinct types
710 :    
711 :     ----------------------------------------------------------------------
712 : blume 629 Name: Matthias Blume
713 : blume 632 Date: 2000/04/30 01:00:00 JST
714 :     Tag: blume-20000430-versions
715 :     Description:
716 :    
717 :     1. CM version numbering added. This is an implementation of Lal's
718 :     proposal for adding version numbers and version checking to .cm
719 :     files. Lal said that his proposal was just that -- a proposal.
720 :     For the time being I went ahead and implemented it so that people
721 :     can comment on it. Everything is completely backward-compatible
722 :     (except for the stable library format, i.e., new bootfiles!).
723 :    
724 :     As usual, see the CM manual for details.
725 :    
726 :     2. An alternative syntax for anchored paths has been implemented.
727 :     Dave has recently voiced the same concerns that I had when I did
728 :     this, so there should be some support. My take is that eventually
729 :     I will let support for the current syntax (where anchors are
730 :     "implicit") fade out in favor of the new, explicit syntax.
731 :     In order to be backward-compatible, both old and new syntax are
732 :     currently supported.
733 :    
734 :     Again, see the CM manual for details.
735 :    
736 :     3. Parallel make is trying to be slightly smarter: When the master
737 :     process finds a "bottleneck", i.e., when there is only one
738 :     compilation unit that can be compiled and everybody else is
739 :     waiting on it, then it will simply compile it directly instead
740 :     of clumsily telling one of the slaves to do it.
741 :    
742 :     4. Support for "unsharing" added. This is necessary in order to be
743 :     able to have two different versions of the same library running
744 :     at the same time (e.g., for trying out a new MLRISC while still
745 :     having the old MLRISC linked into the current compiler, etc.)
746 :     See the CM manual.
747 :    
748 :     5. Simple "makedepend" functionality added for generating Makefile
749 :     dependency information. (This is rather crude at the moment.
750 :     Expect some changes here in the future.)
751 :    
752 :     6. ".fun" added as a recognized suffix for ML files. Also documented
753 :     explicitly in the manual that the fallback behavior (unknown suffix
754 :     -> ML file) is not an official feature!
755 :    
756 :     7. Small changes to the pickler for stable libraries.
757 :    
758 :     8. Several internal changes to CM (for cleanup/improvement).
759 :    
760 :    
761 :     !!!! NEW BINFILES !!!!
762 :    
763 :     ----------------------------------------------------------------------
764 :     Name: Matthias Blume
765 : blume 631 Date: 2000/04/28 17:30:00 JST
766 :     Tag: blume-20000428-pathconfig
767 :     Description:
768 :    
769 :     1. I changed config/install.sh to remove duplicate entries from the
770 :     lib/pathconfig file at the end. Moreover, the final version of
771 :     lib/pathconfig is sorted alphabetically. The same (sorting) is done
772 :     in src/system/installml.
773 :    
774 :     2. The config/install.sh script now consistently uses relative
775 :     pathnames in lib/pathconfig whenever the anchor is in the lib
776 :     directory. (So far this was true for the libraries that come
777 :     pre-compiled and bundled as part of the bootfiles but not for
778 :     libraries that are compiled by the script itself.)
779 :    
780 :     ----------------------------------------------------------------------
781 :     Name: Matthias Blume
782 : blume 629 Date: 2000/04/26 13:10:00 JST
783 :     Tag: blume-20000426-fun_suffix
784 :     Description:
785 :    
786 :     Added ".fun" as a recognized file name suffix (for ML code).
787 :    
788 :     ----------------------------------------------------------------------
789 : leunga 628 Name: Allen Leung
790 :     Date: 2000/04/25 17:00:00 EST
791 :     Tag: leunga-20000425-alpha-ra
792 :     Description:
793 :    
794 :     1. Alpha
795 :    
796 :     PSEUDOARITH was missing in AlphaRewrite. This causes an endless loop
797 :     in C--.
798 :    
799 :     2. RA
800 :    
801 :     Added a flag "ra-dump-size" to print out the size of the flowgraph
802 :     and the interference graph.
803 :    
804 :     ----------------------------------------------------------------------
805 : dbm 627 Name: Dave MacQueen
806 :     Date: 2000/04/25/
807 :     Tag: dbm-20000425-mlyacc_doc_examples
808 :     Description:
809 :     Updated mlyacc.tex sections 5 and 7 for SML '97 and CM.
810 :     Updated all three examples in src/ml-yacc/examples to run
811 :     under 110.* using CM.make.
812 :    
813 :     ----------------------------------------------------------------------
814 : leunga 624 Name: Allen Leung
815 :     Date: 2000/04/20 23:04:00 EST
816 :     Tag: leunga-20000420-ssa-c---stuff
817 :     Description:
818 :    
819 :     This update synchronizes my repository with Yale's. Most of these
820 :     changes, however, do not affect SML/NJ at all (the RA is an exception).
821 :    
822 :     1. Register Allocator
823 :    
824 :     a. An improvement in the interference graph construction:
825 :     Given a copy
826 :    
827 :     s <- t
828 :    
829 : leunga 628 no interference edge between s and t is added for this definition of s.
830 : leunga 624
831 :     b. I've added two new spill heuristic modules that Fermin and I developed
832 :     (in the new library RA.cm). These are unused in SML/NJ but maybe
833 :     useful for others (Moby?)
834 :    
835 :     2. X86
836 :    
837 :     a. Various fixes in the backend provided by Fermin [C--] and Lal.
838 :    
839 :     3. Alpha
840 :    
841 :     a. Added the BSR instruction and code generation that goes with it [C--]
842 :     b. Other fixes too numerous to recount provided by Fermin [C--]
843 :    
844 :     4. Regmaps
845 :    
846 :     a. The regmaps are not initialized with the identity physical bindings
847 :     at creation time. This is unneeded.
848 :    
849 :     5. MLRISC Optimizations
850 :    
851 :     a. The DJ-Graph module can now compute the iterated dominance frontiers
852 :     intersects with liveness incrementally in linear time! Woohoo!
853 :     This is now used in my new SSA construction algorithm.
854 :    
855 :     b. THe branch reorganization module is now smarter about linear chains of
856 :     basic blocks.
857 :    
858 :    
859 :     ----------------------------------------------------------------------
860 : blume 615 Name: Matthias Blume
861 :     Date: 2000/04/12 13:52:00 JST
862 :     Tag: blume_main_v110p27_1
863 :     Description:
864 :    
865 :     Changed install.sh script to handle archive files without version number
866 :     and to use "boot.<arch>-<os>" instead of "sml.boot.<arch>-<os>" for the
867 :     name of the boot file archive.
868 :    
869 :     ----------------------------------------------------------------------
870 : dbm 607 Name: Dave MacQueen
871 :     Date: 2000/04/09 14:00 EDT
872 :     Tag: dbm-20000410-Version_110_27
873 :     Description:
874 :    
875 :     1. Updated src/compiler/TopLevel/main/version.sml to version 110.27
876 :    
877 :     2. Updated src/config/version to 110.27
878 :    
879 :     3. New boot files!
880 :    
881 :     ----------------------------------------------------------------------
882 : leunga 606 Name: Allen Leung
883 :     Date: 2000/04/09 19:09:00 EST
884 :     Tag: leunga-20000409-misc
885 :     Description:
886 :    
887 :     1. Yet another fix for x86 assembly for idivl, imull, mull and friends.
888 :    
889 :     2. Miscellaneous improvements to MLRISC (unused in sml/nj)
890 :    
891 :     ----------------------------------------------------------------------
892 : monnier 605 Name: Stefan
893 :     Date: 2000/04/07 10:00:00 EDT
894 :     Tag: monnier-20000406-branch-handling
895 :     Description:
896 :    
897 :     Improved handling of branches (mostly those generated from
898 :     polymorphic equality), removed switchoff and changed the
899 :     default optimization settings (more cpsopt and less flintopt).
900 :    
901 :     ----------------------------------------------------------------------
902 : leunga 601 Name: Allen Leung
903 : leunga 602 Date: 2000/04/06 01:30:00 EST
904 :     Tag: leunga-20000406-peephole-x86-SSA-2
905 :     Description:
906 :    
907 :     Forgot a few files.
908 :    
909 :     ----------------------------------------------------------------------
910 :     Name: Allen Leung
911 : leunga 601 Date: 2000/04/06 00:36:00 EST
912 :     Tag: leunga-20000406-peephole-x86-SSA
913 :     Description:
914 :    
915 :     1. New Peephole code
916 :    
917 :     2. Minor improvement to X86 instruction selection
918 :    
919 :     3. Various fixes to SSA and machine description -> code translator
920 :    
921 :     ----------------------------------------------------------------------
922 : blume 595 Name: Matthias Blume
923 :     Date: 2000/04/05 12:30:00 JST
924 :     Tag: blume_main_v110p26p2_3
925 :     Description:
926 :    
927 :     This update just merges three minor cosmetic updates to CM's sources
928 :     to get ready for the 110.27 code freeze on Friday. No functionality
929 :     has changed.
930 :    
931 :     ----------------------------------------------------------------------
932 : leunga 593 Name: Allen Leung
933 : leunga 594 Date: 2000/04/04 19:39:00 EST
934 :     Tag: leunga-20000404-x86-asm
935 :     Description:
936 :    
937 :     1. Fixed a problem in X86 assembly.
938 :    
939 :     Things like
940 :    
941 :     jmp %eax
942 :     jmp (%eax)
943 :    
944 :     should be output as
945 :    
946 :     jmp *%eax
947 :     jmp *(%eax)
948 :    
949 :     2. Assembly output
950 :    
951 :     Added a new flag
952 :    
953 :     "asm-indent-copies" (default to false)
954 :    
955 :     When this flag is on, parallel copies will be indented an extra level.
956 :    
957 :     ----------------------------------------------------------------------
958 :     Name: Allen Leung
959 : leunga 593 Date: 2000/04/04 03:18:00 EST
960 :     Tag: leunga-20000404-C--Moby
961 :     Description:
962 :    
963 :     All of these fixes are related to C--, Moby, and my own optimization
964 :     stuff; so they shouldn't affect SML/NJ.
965 :    
966 :     1. X86
967 :    
968 :     Various fixes related floating point, and extensions.
969 :    
970 :     2. Alpha
971 :    
972 :     Some extra patterns related to loads with signed/zero extension
973 :     provided by Fermin.
974 :    
975 :     3. Assembly
976 :    
977 : monnier 659 When generating assembly, resolve the value of client defined constants,
978 : leunga 593 instead of generating symbolic values. This is controlled by the
979 :     new flag "asm-resolve-constants", which is default to true.
980 :    
981 :     4. Machine Descriptions
982 :    
983 :     a. The precedence parser was slightly broken when parsing infixr symbols.
984 :     b. The type generalizing code had the bound variables reversed, resulting
985 :     in a problem during arity raising.
986 :     c. Various fixes in machine descriptions.
987 :    
988 :     ----------------------------------------------------------------------
989 : blume 592 Name: Matthias Blume
990 :     Date: 2000/04/03 16:05:00 JST
991 :     Tag: blume_main_v110p26p2_2
992 :     Description:
993 :    
994 :     I eliminated coreEnv from compInfo. Access to the "Core" structure is
995 :     now done via the ordinary static environment that is context to each
996 :     compilation unit.
997 :    
998 :     To this end, I arranged that instead of "structure Core" as "structure
999 :     _Core" is bound in the pervasive environment. Core access is done via
1000 : monnier 659 _Core (which can never be accidentally rebound because _Core is not a
1001 : blume 592 legal surface-syntax symbol).
1002 :    
1003 :     The current solution is much cleaner because the core environment is
1004 :     now simply part of the pervasive environment which is part of every
1005 :     compilation unit's context anyway. In particular, this eliminates all
1006 :     special-case handling that was necessary until now in order to deal
1007 :     with dynamic and symbolic parts of the core environment.
1008 :    
1009 :     Remaining hackery (to bind the "magic" symbol _Core) is localized in the
1010 : monnier 659 compilation manager's bootstrap compiler (actually: in the "init group"
1011 : blume 592 handling). See the comments in src/system/smlnj/init/init.cmi for
1012 :     more details.
1013 :    
1014 :     I also tried to track down all mentions of "Core" (as string argument
1015 :     to Symbol.strSymbol) in the compiler and replaced them with a
1016 :     reference to the new CoreSym.coreSym. Seems cleaner since the actual
1017 :     name appears in one place only.
1018 :    
1019 :     Binfile and bootfile format have not changed, but the switchover from
1020 :     the old "init.cmi" to the new one is a bit tricky, so I supplied new
1021 :     bootfiles anyway.
1022 :    
1023 :     ----------------------------------------------------------------------
1024 : leunga 591 Name: Allen Leung
1025 :     Date: 2000/04/02 21:17:00 EST
1026 :     Tag: leunga-20000402-mltree
1027 :     Description:
1028 : leunga 585
1029 : leunga 591 1. Renamed the constructor CALL in MLTREE by popular demand.
1030 :     2. Added a bunch of files from my repository. These are currently
1031 :     used by other non-SMLNJ backends.
1032 :    
1033 : leunga 576 ----------------------------------------------------------------------
1034 : leunga 590 Name: Allen Leung
1035 :     Date: 2000/03/31 21:15:00 EST
1036 :     Tag: leunga-20000331-aliasing
1037 :     Description:
1038 :    
1039 :     This update contains a rewritten (and hopefully more correct) module
1040 :     for extracting aliasing information from CPS.
1041 :    
1042 :     To turn on this feature:
1043 :    
1044 :     Compiler.Control.CG.memDisambiguate := true
1045 :    
1046 :     To pretty print the region information with assembly
1047 :    
1048 :     Compiler.Control.MLRISC.getFlag "asm-show-region" := true;
1049 :    
1050 :     To control how many levels of aliasing information are printed, use:
1051 :    
1052 :     Compiler.Control.MLRISC.getInt "points-to-show-level" := n
1053 :    
1054 :     The default of n is 3.
1055 :    
1056 :     ----------------------------------------------------------------------
1057 : dbm 589 Name: David MacQueen
1058 :     Date: 2000/03/31 11:15:00 EST
1059 :     Tag: dbm-20000331-runtime_fix
1060 :     Description:
1061 :    
1062 :     This update contains:
1063 :    
1064 :     1. runtime/c-lib/c-libraries.c
1065 :     includes added in revision 1.2 caused compilation errors on hppa-hpux
1066 :    
1067 :     2. fix for bug 1556
1068 :     system/Basis/Implementation/NJ/internal-signals.sml
1069 :    
1070 :     ----------------------------------------------------------------------
1071 : blume 587 Name: Matthias Blume
1072 : blume 588 Date: 2000/03/31 18:00:00 JST
1073 :     Tag: blume_main_v110p26p2_1
1074 :     Description:
1075 :    
1076 :     This update contains:
1077 :    
1078 :     1. A small change to CM's handling of stable libraries:
1079 :     CM now maintains one "global" modmap that is used for all stable
1080 :     libraries. The use of such a global modmap maximizes sharing and
1081 :     minimizes the need for re-traversing parts of environments during
1082 :     modmap construction. (However, this has minor impact since modmap
1083 :     construction seems to account for just one percent or less of total
1084 :     compile time.)
1085 :    
1086 :     2. I added a "genmap" phase to the statistics. This is where I got the
1087 :     "one percent" number (see above).
1088 :    
1089 :     3. CM's new tool parameter mechanism just became _even_ better. :)
1090 :     - The parser understands named parameters and recursive options.
1091 :     - The "make" and "shell" tools use these new features.
1092 :     (This makes it a lot easier to cascade these tools.)
1093 :     - There is a small syntax change: named parameters use a
1094 :    
1095 :     <name> : ( <option> ... ) or
1096 :     <name> : <string>
1097 :    
1098 :     syntax. Previously, named parameters were implemented in an
1099 :     ad-hoc fashion by each tool individually (by parsing strings)
1100 :     and had the form
1101 :    
1102 :     <name>=<string>
1103 :    
1104 :     See the CM manual for a full description of these issues.
1105 :    
1106 :     ----------------------------------------------------------------------
1107 :     Name: Matthias Blume
1108 : blume 587 Date: 2000/03/30 18:00:00 JST
1109 :     Tag: blume_main_v110p26p2_0
1110 :     Description:
1111 :    
1112 :     !!!!! WARNING !!!!!!
1113 :     !! New binfiles !!
1114 :     !!!!!!!!!!!!!!!!!!!!
1115 :    
1116 :     This update contains:
1117 :    
1118 :     1. Moderate changes to CM:
1119 :    
1120 :     - Changes to CM's tools mechanism. In particular, it is now possible
1121 :     to have tools that accept additional "command line" parameters
1122 :     (specified in the .cm file at each instance where the tool's class is
1123 :     used).
1124 :    
1125 : monnier 659 This was done to accommodate the new "make" and "shell" tools which
1126 :     facilitate fairly seamless hookup to portions of code managed using
1127 : blume 587 Makefiles or Shell scripts.
1128 :    
1129 :     There are no classes "shared" or "private" anymore. Instead, the
1130 :     sharing annotation is now a parameter to the "sml" class.
1131 :    
1132 :     There is a bit of generic machinery for implementing one's own
1133 :     tools that accept command-line parameters. However, I am not yet fully
1134 :     satisfied with that part, so expect changes here in the future.
1135 :    
1136 :     All existing tools are described in the CM manual.
1137 :    
1138 : monnier 659 - Slightly better error handling. (CM now suppresses many followup
1139 : blume 587 error messages that tended to be more annoying than helpful.)
1140 :    
1141 :     2. Major changes to the compiler's static environment data structures.
1142 :    
1143 :     - no CMStaticEnv anymore.
1144 :     - no CMEnv, no "BareEnvironment" (actually, _only_ BareEnvironment,
1145 :     but it is called Environment), no conversions between different
1146 :     kinds of static environments
1147 :    
1148 :     - There is still a notion of a "modmap", but such modmaps are generated
1149 :     on demand at the time when they are needed. This sounds slow, but I
1150 :     sped up the code that generates modmaps enough for this not to lead to
1151 :     a slowdown of the compiler (at least I didn't detect any).
1152 :    
1153 :     - To facilitate rapid modmap generation, static environments now
1154 :     contain an (optional) "modtree" structure. Modtree annotations are
1155 :     constructed by the unpickler during unpickling. (This means that
1156 :     the elaborator does not have to worry about modtrees at all.)
1157 :     Modtrees have the advantage that they are compositional in the same
1158 :     way as the environment data structure itself is compositional.
1159 :     As a result, modtrees never hang on to parts of an environment that
1160 :     has already been rendered "stale" by filtering or rebinding.
1161 :    
1162 :     - I went through many, many trials and errors before arriving at the
1163 :     current solution. (The initial idea of "linkpaths" did not work.)
1164 :     But the result of all this is that I have touched a lot of files that
1165 :     depend on the "modules" and "types" data structures (most of the
1166 :     elaborator). There were a lot of changes during my "linkpath" trials
1167 :     that could have been reverted to their original state but weren't.
1168 :     Please, don't be too harsh on me for messing with this code a bit more
1169 :     than what was strictly necessary... (I _did_ resist the tempation
1170 :     of doing any "global reformatting" to avoid an untimely death at
1171 :     Dave's hands. :)
1172 :    
1173 :     - One positive aspect of the previous point: At least I made sure that
1174 :     all files that I touched now compile without warnings (other than
1175 :     "polyEqual").
1176 :    
1177 :     - compiler now tends to run "leaner" (i.e., ties up less memory in
1178 :     redundant modmaps)
1179 :    
1180 :     ----------------------------------------------------------------------
1181 : leunga 580 Name: Allen Leung
1182 : leunga 585 Date: 2000/03/29 18:00:00
1183 :     Tag: leunga-20000327-mlriscGen_hppa_alpha_x86
1184 : leunga 586 Boot files (optional): ftp://react-ilp.cs.nyu.edu/leunga/110.26.1-sml.boot.x86-unix-20000330.tar.gz
1185 : leunga 585 Description:
1186 :    
1187 :     This update contains *MAJOR* changes to the way code is generated from CPS
1188 :     in the module mlriscGen, and in various backend modules.
1189 :    
1190 :     CHANGES
1191 :     =======
1192 :    
1193 :     1. MLRiscGen: forward propagation fix.
1194 :    
1195 :     There was a bug in forward propagation introduced at about the same time
1196 :     as the MLRISC x86 backend, which prohibits coalescing to be
1197 :     performed effectively in loops.
1198 :    
1199 :     Effect: speed up of loops in RISC architectures.
1200 :     By itself, this actually slowed down certain benchmarks on the x86.
1201 :    
1202 :     2. MLRiscGen: forward propagating addresses from consing.
1203 :    
1204 :     I've changed the way consing code is generated. Basically I separated
1205 :     out the initialization part:
1206 :    
1207 :     store tag, offset(allocptr)
1208 :     store elem1, offset+4(allocptr)
1209 :     store elem2, offset+8(allocptr)
1210 :     ...
1211 :     store elemn, offset+4n(allocptr)
1212 :    
1213 :     and the address computation part:
1214 :    
1215 :     celladdr <- offset+4+alloctpr
1216 :    
1217 :     and move the address computation part
1218 :    
1219 :     Effect: register pressure is generally lower as a result. This
1220 :     makes compilation of certain expressions much faster, such as
1221 :     long lists with non-trivial elements.
1222 :    
1223 :     [(0,0), (0,0), .... (0,0)]
1224 :    
1225 :     3. MLRiscGen: base pointer elimination.
1226 :    
1227 :     As part of the linkage mechanism, we generate the sequence:
1228 :    
1229 :     L: ... <- start of the code fragment
1230 :    
1231 :     L1:
1232 :     base pointer <- linkreg - L1 + L
1233 :    
1234 :     The base pointer was then used for computing relocatable addresses
1235 :     in the code fragment. Frequently (such as in lots of continuations)
1236 :     this is not needed. We now eliminate this sequence whenever possible.
1237 :    
1238 :     For compile time efficiency, I'm using a very stupid local heuristic.
1239 :     But in general, this should be done as a control flow analysis.
1240 :    
1241 :     Effect: Smaller code size. Speed up of most programs.
1242 :    
1243 :     4. Hppa back end
1244 :    
1245 :     Long jumps in span dependence resolution used to depend on the existence
1246 :     of the base pointer.
1247 :    
1248 :     A jump to a long label L was expanded into the following sequence:
1249 :    
1250 :     LDIL %hi(L-8192), %r29
1251 :     LDO %lo(L-8192)(%r29), %r29
1252 :     ADD %r29, baseptr, %r29
1253 :     BV,n %r0(%r29)
1254 :    
1255 :     In the presence of change (3) above, this will not work. I've changed
1256 :     it so that the following sequence of instructions are generated, which
1257 :     doesn't mention the base pointer at all:
1258 :    
1259 :     BL,n L', %r29 /* branch and link, L' + 4 -> %r29 */
1260 :     L': ADDIL L-(L'+4), %r29 /* Compute address of L */
1261 :     BV,n %r0(%r29) /* Jump */
1262 :    
1263 :     5. Alpha back end
1264 :    
1265 :     New alpha instructions LDB/LDW have been added, as per Fermin's
1266 :     suggestions. This is unrelated to all other changes.
1267 :    
1268 :     6. X86 back end
1269 :    
1270 :     I've changed andl to testl in the floating point test sequence
1271 :     whenever appropriate. The Intel optimization guide states that
1272 : monnier 659 testl is preferable to andl.
1273 : leunga 585
1274 :     7. RA (x86 only)
1275 :    
1276 :     I've improved the spill propagation algorithm, using an approximation
1277 :     of maximal weighted independent sets. This seems to be necessary to
1278 :     alleviate the negative effect in light of the slow down in (1).
1279 :    
1280 :     I'll write down the algorithm one of these days.
1281 :    
1282 :     8. MLRiscGen: frequencies
1283 :    
1284 :     I've added an annotation that states that all call gc blocks have zero
1285 :     execution frequencies. This improves register allocation on the x86.
1286 :    
1287 :     BENCHMARKS
1288 :     ==========
1289 :    
1290 :     I've only perform the comparison on 110.25.
1291 :    
1292 :     The platforms are:
1293 :    
1294 :     HPPA A four processor HP machine (E9000) with 5G of memory.
1295 :     X86 A 300Hhz Pentium II with 128M of memory, and
1296 :     SPARC An Ultra sparc 2 with 512M of memory.
1297 :    
1298 :     I used the following parameters for the SML benchmarks:
1299 :    
1300 :     @SMLalloc
1301 :     HPPA 256k
1302 :     SPARC 512k
1303 :     X86 256k
1304 :    
1305 :     COMPILATION TIME
1306 :     ----------------
1307 :     Here are the numbers comparing the compilation times of the compilers.
1308 :     I've only compared 110.25 compiling the new sources versus
1309 :     a fixpoint version of the new compiler compiling the same.
1310 :    
1311 :     110.25 New
1312 :     Total Time in RA Spill+Reload Total Time In RA Spill+Reload
1313 :     HPPA 627s 116s 2684+3584 599s 95s 1003+1879
1314 :     SPARC 892s 173s 2891+3870 708s 116s 1004+1880
1315 :     X86 999s 315s 94006+130691 987s 296s 108877+141957
1316 :    
1317 :     110.25 New
1318 :     Code Size Code Size
1319 :     HPPA 8596736 8561421
1320 :     SPARC 8974299 8785143
1321 :     X86 9029180 8716783
1322 :    
1323 :     So in summary, things are at least as good as before. Dramatic
1324 :     reduction in compilation is obtained on the Sparc; I can't explain it,
1325 :     but it is reproducible. Perhaps someone should try to reproduce this
1326 :     on their own machines.
1327 :    
1328 :     SML BENCHMARKS
1329 :     --------------
1330 :    
1331 :     On the average, all benchmarks perform at least as well as before.
1332 :    
1333 :     HPPA Compilation Time Spill+Reload Run Time
1334 :     110.25 New 110.25 New 110.25 New
1335 :    
1336 :     barnesHut 3.158 3.015 4.75% 1+1 0+0 2.980 2.922 2.00%
1337 :     boyer 6.152 5.708 7.77% 0+0 0+0 0.218 0.213 2.34%
1338 :     count-graphs 1.168 1.120 4.32% 0+0 0+0 22.705 23.073 -1.60%
1339 :     fft 0.877 0.792 10.74% 1+3 1+3 0.602 0.587 2.56%
1340 :     knuthBendix 3.180 2.857 11.32% 0+0 0+0 0.675 0.662 2.02%
1341 :     lexgen 6.190 5.290 17.01% 0+0 0+0 0.913 0.788 15.86%
1342 :     life 0.803 0.703 14.22% 25+25 0+0 0.153 0.140 9.52%
1343 :     logic 2.048 2.007 2.08% 6+6 1+1 4.133 4.008 3.12%
1344 :     mandelbrot 0.077 0.080 -4.17% 0+0 0+0 0.765 0.712 7.49%
1345 :     mlyacc 22.932 20.937 9.53% 154+181 32+57 0.468 0.430 8.91%
1346 :     nucleic 5.183 5.060 2.44% 2+2 0+0 0.125 0.120 4.17%
1347 :     ratio-regions 3.357 3.142 6.84% 0+0 0+0 116.225 113.173 2.70%
1348 :     ray 1.283 1.290 -0.52% 0+0 0+0 2.887 2.855 1.11%
1349 :     simple 6.307 6.032 4.56% 28+30 5+7 3.705 3.658 1.28%
1350 :     tsp 0.888 0.862 3.09% 0+0 0+0 7.040 6.893 2.13%
1351 :     vliw 24.378 23.455 3.94% 106+127 25+45 2.758 2.707 1.91%
1352 :     --------------------------------------------------------------------------
1353 :     Average 6.12% 4.09%
1354 :    
1355 :     SPARC Compilation Time Spill+Reload Run Time
1356 :     110.25 New 110.25 New 110.25 New
1357 :    
1358 :     barnesHut 3.778 3.592 5.20% 2+2 0+0 3.648 3.453 5.65%
1359 :     boyer 6.632 6.110 8.54% 0+0 0+0 0.258 0.242 6.90%
1360 :     count-graphs 1.435 1.325 8.30% 0+0 0+0 33.672 34.737 -3.07%
1361 :     fft 0.980 0.940 4.26% 3+9 2+6 0.838 0.827 1.41%
1362 :     knuthBendix 3.590 3.138 14.39% 0+0 0+0 0.962 0.967 -0.52%
1363 :     lexgen 6.593 6.072 8.59% 1+1 0+0 1.077 1.078 -0.15%
1364 :     life 0.972 0.868 11.90% 26+26 0+0 0.143 0.140 2.38%
1365 :     logic 2.525 2.387 5.80% 7+7 1+1 5.625 5.158 9.05%
1366 :     mandelbrot 0.090 0.093 -3.57% 0+0 0+0 0.855 0.728 17.39%
1367 :     mlyacc 26.732 23.827 12.19% 162+189 32+57 0.550 0.560 -1.79%
1368 :     nucleic 6.233 6.197 0.59% 3+3 0+0 0.163 0.173 -5.77%
1369 :     ratio-regions 3.780 3.507 7.79% 0+0 0+0 133.993 131.035 2.26%
1370 :     ray 1.595 1.550 2.90% 1+1 0+0 3.440 3.418 0.63%
1371 :     simple 6.972 6.487 7.48% 29+32 5+7 3.523 3.525 -0.05%
1372 :     tsp 1.115 1.063 4.86% 0+0 0+0 7.393 7.265 1.77%
1373 :     vliw 27.765 24.818 11.87% 110+135 25+45 2.265 2.135 6.09%
1374 :     ----------------------------------------------------------------------------
1375 :     Average 6.94% 2.64%
1376 :    
1377 :     X86 Compilation Time Spill+Reload Run Time
1378 :     110.25 New 110.25 New 110.25 New
1379 :    
1380 :     barnesHut 5.530 5.420 2.03% 593+893 597+915 3.532 3.440 2.66%
1381 :     boyer 8.768 7.747 13.19% 493+199 301+289 0.327 0.297 10.11%
1382 :     count-graphs 2.040 2.010 1.49% 298+394 315+457 26.578 28.660 -7.26%
1383 :     fft 1.327 1.302 1.92% 112+209 115+210 1.055 0.962 9.71%
1384 :     knuthBendix 5.218 5.475 -4.69% 451+598 510+650 0.928 0.932 -0.36%
1385 :     lexgen 9.970 9.623 3.60% 1014+841 1157+885 0.947 0.928 1.97%
1386 :     life 1.183 1.183 0.00% 162+182 145+148 0.127 0.103 22.58%
1387 :     logic 3.285 3.512 -6.45% 514+684 591+836 5.682 5.577 1.88%
1388 :     mandelbrot 0.147 0.143 2.33% 38+41 33+54 0.703 0.690 1.93%
1389 :     mlyacc 35.457 32.763 8.22% 3496+4564 3611+4860 0.552 0.550 0.30%
1390 :     nucleic 7.100 6.888 3.07% 239+168 201+158 0.175 0.173 0.96%
1391 :     ratio-regions 6.388 6.843 -6.65% 1182+257 981+300 120.142 120.345 -0.17%
1392 :     ray 2.332 2.338 -0.29% 346+398 402+494 3.593 3.540 1.51%
1393 :     simple 9.912 9.903 0.08% 1475+941 1579+1168 3.057 3.178 -3.83%
1394 :     tsp 1.623 1.532 5.98% 266+200 250+211 8.045 7.878 2.12%
1395 :     vliw 33.947 35.470 -4.29% 2629+2774 2877+3171 2.072 1.890 9.61%
1396 :     ----------------------------------------------------------------------------
1397 :     Average 1.22% 3.36%
1398 :    
1399 :     ----------------------------------------------------------------------
1400 :     Name: Allen Leung
1401 : leunga 583 Date: 2000/03/23 16:25:00
1402 :     Tag: leunga-20000323-fix_x86_alpha
1403 :     Description:
1404 :    
1405 :     1. X86 fixes/changes
1406 :    
1407 :     a. The old code generated for SETcc was completely wrong.
1408 :     The Intel optimization guide is VERY misleading.
1409 :    
1410 :     2. ALPHA fixes/changes
1411 :    
1412 :     a. Added the instructions LDBU, LDWU, STB, STW as per Fermin's suggestion.
1413 :     b. Added a new mode byteWordLoadStores to the functor parameter to Alpha()
1414 :     c. Added reassociation code for address computation.
1415 :    
1416 :     ----------------------------------------------------------------------
1417 :     Name: Allen Leung
1418 : leunga 580 Date: 2000/03/22 01:23:00
1419 :     Tag: leunga-20000322-fix_x86_hppa_ra
1420 :     Description:
1421 :    
1422 :     1. X86 fixes/changes
1423 :    
1424 :     a. x86Rewrite bug with MUL3 (found by Lal)
1425 :     b. Added the instructions FSTS, FSTL
1426 :    
1427 :     2. PA-RISC fixes/changes
1428 :    
1429 :     a. B label should not be a delay slot candidate! Why did this work?
1430 :     b. ADDT(32, REG(32, r), LI n) now generates one instruction instead of two,
1431 :     as it should be.
1432 :     c. The assembly syntax for fstds and fstdd was wrong.
1433 :     d. Added the composite instruction COMICLR/LDO, which is the immediate
1434 :     operand variant of COMCLR/LDO.
1435 :    
1436 :     3. Generic MLRISC
1437 :    
1438 :     a. shuffle.sml rewritten to be slightly more efficient
1439 :     b. DIV bug in mltree-simplify fixed (found by Fermin)
1440 :    
1441 :     4. Register Allocator
1442 :    
1443 :     a. I now release the interference graph earlier during spilling.
1444 :     May improve memory usage.
1445 :    
1446 :     ----------------------------------------------------------------------
1447 : blume 577 Name: Matthias Blume
1448 : blume 578 Date: 2000/03/14 14:15:32
1449 :     Tag: blume_main_v110p26p1_2
1450 :     Description:
1451 :    
1452 :     1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
1453 :     additional argument called "template" which is an optional string that
1454 : monnier 659 specifies the layout of the tool command line. See the CM manual for
1455 : blume 578 explanation.
1456 :    
1457 : monnier 659 2. A special-purpose tool can be "registered" by simply dropping the
1458 : blume 578 corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
1459 :     directory where the .cm file lives that uses this tool. (The
1460 :     behavior/misfeature until now was to look for the tool description
1461 :     files in the current working directory.) As before, tool description
1462 :     files could also be anchored -- in which case they can live anywhere
1463 :     they like. Following the recent e-mail discussion, this change should
1464 :     make it easier to have special-purpose tools that are shipped together
1465 :     with the sources of the program that uses them.
1466 :    
1467 :     ----------------------------------------------------------------------
1468 :     Name: Matthias Blume
1469 : blume 577 Date: 2000/03/10 07:48:34
1470 :     Tag: blume_main_v110p26p1_1
1471 :     Description:
1472 :    
1473 :     I added a re-written version of Dave's fixpt script to src/system.
1474 :     Changes relative to the original version:
1475 :     - sh-ified (not everybody has ksh)
1476 :     - automatically figures out which architecture it runs on
1477 :     - uses ./makeml a bit more cleverly
1478 :     - never invokes ./installml (and, thus, does not clobber your
1479 :     good and working installation of sml in case something goes wrong)
1480 :     - accepts max iteration count using option "-iter <n>"
1481 :     - accepts a "base" name using option "-base <base>"
1482 :    
1483 :     It does not build any extraneous heap images but directly rebuilds
1484 :     bin- and boot-hierarchies using makeml's "-rebuild" switch. Finally,
1485 :     it can incorporate existing bin- and boot- hierarchies. For example,
1486 :     suppose the base is set to "sml" (which is the default). Then it
1487 :     successively builds
1488 :    
1489 :     sml.bin.<arch>-unix and sml.boot.<arch>-unix
1490 :     then sml1.bin.<arch>-unix and sml1.boot.<arch>-unix
1491 :     then sml2.bin.<arch>-unix and sml2.boot.<arch>-unix
1492 :     ...
1493 :     then sml<n>.bin.<arch>-unix and sml<n>.boot.<arch>-unix
1494 :    
1495 :     and so on. If any of these already exist, it will just use what's
1496 :     there. In particular, many people will have the initial set of bin
1497 :     and boot files around, so this saves time for at least one full
1498 :     rebuild. Having sets of the form <base><k>.{bin,boot}.<arch>-unix for
1499 :     <k>=1,2,... is normally not a good idea when invoking fixpt. However,
1500 :     they might be the result of an earlier partial run of fixpt (which
1501 : monnier 659 perhaps got accidentally killed). In this case, fixpt will quickly
1502 : blume 577 move through what exists before continuing where it left off earlier,
1503 :     and, thus, saves a lot of time.
1504 :    
1505 :     ----------------------------------------------------------------------
1506 : leunga 576 Name: Allen Leung
1507 :     Date: 00/03/10 02:20:00
1508 :     Tag: leunga-20000310-fix_x86_asm_ra
1509 :     Description:
1510 : dbm 570
1511 : leunga 576 More assembly output problems involving the indexed addressing mode
1512 :     on the x86 have been found and corrected. Thanks to Fermin Reig for the
1513 :     fix.
1514 :    
1515 :     The interface and implementation of the register allocator have been changed
1516 :     slightly to accommodate the possibility to skip the register allocation
1517 :     phases completely and go directly to memory allocation. This is needed
1518 :     for C-- use.
1519 :    
1520 : dbm 570 ----------------------------------------------------------------------
1521 : blume 572 Name: Matthias Blume
1522 : blume 575 Date: 00/03/09 10:23:53
1523 :     Tag: blume_main_v110p26p1_0
1524 :     Description:
1525 :    
1526 :     * Complete re-organization of library names. Many libraries have been
1527 :     consolidated so that they share the same path anchor. For example,
1528 :     all MLRISC-related libraries are anchored at MLRISC, most libraries that
1529 :     are SML/NJ-specific are under "smlnj". Notice that names like
1530 :     host-cmb.cm or host-compiler.cm no longer exist. See system/README
1531 :     for a complete description of the new naming scheme. Quick reference:
1532 :    
1533 :     host-cmb.cm -> smlnj/cmb.cm
1534 :     host-compiler.cm -> smlnj/compiler.cm
1535 :     full-cm.cm -> smlnj/cm.cm
1536 :     <arch>-<os>.cm -> smlnj/cmb/<arch>-<os>.cm
1537 :     <arch>-compiler.cm -> smlnj/compiler/<arch>.cm
1538 :    
1539 :     * Bug fixes in CM.
1540 :     - exceptions in user code are being passed through (i.e., reach top level)
1541 :     - more bugs in paranoia mode fixed
1542 :     - bug related to checking group owners fixed
1543 :    
1544 :     * New install.sh script that automagically fetches archive files:
1545 :     The new file config/srcarchiveurl must contain the URL of the
1546 :     (remote) directory that contains bin files (or other source archives).
1547 :     If install.sh does not find the archive locally, it tries to get
1548 :     it from that remote directory.
1549 :     This should simplify installation further: For machines that have
1550 :     access to the internet, just fetch <version>-config.tgz, unpack it,
1551 : monnier 659 edit config/targets, and go (run config/install.sh). The script will
1552 : blume 575 fetch everything else that it might need all by itself.
1553 :    
1554 :     For CVS users, this mechanism is not relevant for source archives, but
1555 :     it is convenient for getting new sets of binfiles.
1556 :    
1557 :     Archives should be tar files compressed with either gzip, compress, or
1558 :     bzip2. The script recognizes .tgz, .tar, tar.gz, tz, .tar.Z, and .tar.bz2.
1559 :    
1560 :     ----------------------------------------------------------------------
1561 :     Name: Matthias Blume
1562 : blume 572 Date: 2000/03/07 04:01:04
1563 :     Tag: blume_main_v110_26_2
1564 : dbm 570 Description:
1565 : blume 572 - size info in BOOTLIST
1566 :     * no fixed upper limits for number of bootfiles or length of
1567 :     bootfile names in runtime
1568 :     * falling back to old behavior if no BOOTLIST size info found
1569 :     - allocation size heuristics in .run-sml
1570 :     * tries to read cache size from /proc/cpuinfo (this is important for
1571 :     small-cache Celeron systems!)
1572 :     - install.sh robustified
1573 :     - CM manual updates
1574 :     - paranoid mode
1575 :     * no more CMB.deliver() (i.e., all done by CMB.make())
1576 :     * can re-use existing sml.boot.* files
1577 :     * init.cmi now treated as library
1578 :     * library stamps for consistency checks
1579 :     - sml.boot.<arch>-<os>/PIDMAP file
1580 :     * This file is read by the CM startup code. This is used to minimize
1581 :     the amount of dynamic state that needs to be stowed away for the
1582 :     purpose of sharing between interactive system and user code.
1583 :     - CM.Anchor.anchor instead of CM.Anchor.{set,cancel}
1584 :     * Upon request by Elsa. Anchors now controlled by get-set-pair
1585 :     like most other CM state variables.
1586 :     - Compiler.CMSA eliminated
1587 :     * No longer supported by CM anyway.
1588 :     - fixed bugs in pickler that kept biting Stefan
1589 :     * past refs to past refs (was caused by the possibility that
1590 :     ad-hoc sharing is more discriminating than hash-cons sharing)
1591 :     * integer overflow on LargeInt.minInt
1592 :     - ml-{lex,yacc} build scripts now use new mechanism
1593 :     for building standalone programs
1594 :     - fixed several gcc -Wall warnings that were caused by missing header
1595 :     files, missing initializations, etc., in runtime (not all warnings
1596 :     eliminated, though)

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