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/HISTORY
ViewVC logotype

Annotation of /sml/trunk/HISTORY

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1291 - (view) (download)

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 : george 1003 Date: yyyy/mm/dd
12 : dbm 570 Tag: <post-commit CVS tag>
13 :     Description:
14 : leunga 1142
15 : george 1136 ----------------------------------------------------------------------
16 : blume 1229 Name: Matthias Blume
17 : blume 1291 Date: 2002/12/06 22:40:00 EST
18 :     Tag: blume-20021206-cm-fileid
19 :     Description:
20 :    
21 :     Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
22 :     This fixes the problem of CM getting confused by files that suddenly
23 :     change their identity (e.g., by getting unlinked and recreated by some
24 :     text editor such as vi). There might be a better/cheaper/cleaner way
25 :     of doing this, but for now this will have to do.
26 :    
27 :     ----------------------------------------------------------------------
28 :     Name: Matthias Blume
29 : blume 1288 Date: 2002/10/28 09:50:00 EST
30 :     Tag: blume-20021028-typecheck
31 :     Description:
32 :    
33 :     Exported structure Typecheck from $smlnj/viscomp/core.cm.
34 :    
35 :     ----------------------------------------------------------------------
36 :     Name: Matthias Blume
37 : blume 1282 Date: 2002/10/17 09:10:00 EDT
38 : blume 1279 Tag: Release_110_42
39 :     Description:
40 :    
41 : blume 1282 In good old tradition, there has been a slight hiccup so that we have
42 :     to patch 110.42 after the fact. The old release tag has been replaced
43 :     (see below).
44 :    
45 :     The change solves a problem with two competing approaches the
46 :     configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
47 :     each other's way.
48 :    
49 :     This change only affects the runtime system code and the installer script.
50 :     (No new bootfiles.)
51 :    
52 :     ----------------------------------------------------------------------
53 :     Name: Matthias Blume
54 :     Date: 2002/10/16 12:00:00 EDT
55 :     Tag: Release_110_42_removed
56 :     Description:
57 :    
58 : blume 1279 New working release. New bootfiles.
59 :    
60 :     ----------------------------------------------------------------------
61 :     Name: Matthias Blume
62 : blume 1273 Date: 2002/10/10 13:10:00 EDT
63 :     Tag: blume-20021010-ppc-divs
64 :     Description:
65 :    
66 :     The mltree operator DIVS must be implemented with an overflow check on
67 :     the PPC because the hardware indicates divide-by-zero using "overflow" as
68 :     well.
69 :    
70 :     ----------------------------------------------------------------------
71 :     Name: Matthias Blume
72 : blume 1261 Date: 2002/07/23 11:45:00
73 :     Tag: blume-20020723-smlnj-home
74 :     Description:
75 :    
76 :     Sml now senses the SMLNJ_HOME environment variable. If this is set,
77 :     then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
78 :     CM_PATHCONFIG is also set), the path configuration file is assumed
79 :     to be in $SMLNJ_HOME/lib/pathconfig. This way one can easily move
80 :     the entire tree to some other place and everything will "just work".
81 :    
82 :     (Companion commands such as ml-build and ml-makedepend also sense this
83 :     variable.)
84 :    
85 :     ----------------------------------------------------------------------
86 :     Name: Matthias Blume
87 : blume 1259 Date: 2002/07/12 21:19:00 EDT
88 :     Tag: blume-20020712-liveness
89 :     Description:
90 :    
91 :     Exported two useful "step" functions from liveness module (MLRISC).
92 :    
93 :     ----------------------------------------------------------------------
94 :     Name: Matthias Blume
95 : blume 1253 Date: 2002/07/05 16:00 EDT
96 :     Tag: Release_110_41
97 :     Description:
98 :    
99 :     New working release. New bootfiles.
100 :    
101 :     ----------------------------------------------------------------------
102 :     Name: Matthias Blume
103 : blume 1250 Date: 2002/07/05 10:25:00 EDT
104 :     Tag: blume-20020705-btimp
105 :     Description:
106 :    
107 :     Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
108 :     other clients can set up backtracing support.
109 :    
110 :     ----------------------------------------------------------------------
111 :     Name: Matthias Blume
112 : blume 1249 Date: 2002/06/25 17:23:00 EDT
113 :     Tag: blume-20020625-fpmax
114 :     Description:
115 :    
116 :     Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
117 :     case. (The sense of the isNaN test was reversed -- which made min and
118 :     max always return their first argument.)
119 :    
120 :     ----------------------------------------------------------------------
121 :     Name: Matthias Blume
122 : blume 1240 Date: 2002/06/11
123 :     Tag: blume-20020611-unixpath
124 :     Description:
125 :    
126 :     Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
127 :    
128 :     ----------------------------------------------------------------------
129 :     Name: Matthias Blume
130 : blume 1238 Date: 2002/06/10 16:35:00 EDT
131 :     Tag: blume-20020610-ieeereal
132 :     Description:
133 :    
134 :     I back-ported my implementation of IEEEReal.fromString from the
135 :     idlbasis-devel branch so that we can test it.
136 :    
137 :     Another small change is that ppDec tries to give more information
138 :     than just "<sig>" in the case of functors. However, this code is
139 :     broken in some mysterious way if the functor's body's signature
140 :     has not been declared by ascription but gets inferred from the
141 :     implementation. This needs fixing...
142 :    
143 :     ----------------------------------------------------------------------
144 :     Name: Matthias Blume
145 : blume 1229 Date: 2002/05/31
146 :     Tag: blume-20020531-btrace-mode
147 :     Description:
148 :    
149 :     Resurrected SMLofNJ.Internals.BTrace.mode. (It accidentally fell by
150 :     the wayside when I switched over to using Controls everywhere.)
151 :    
152 :     ----------------------------------------------------------------------
153 : george 1222 Name: Lal George
154 :     Date: 2002/05/23 12:21:40 EDT
155 :     Tag: george-20020523-visual-labels
156 :     Description:
157 :    
158 :     Labels are now displayed in the graphical output to make
159 :     the fall-through and target blocks obvious.
160 :    
161 :     ----------------------------------------------------------------------
162 : blume 1204 Name: Matthias Blume
163 : blume 1220 Date: 2002/05/22 11:03:00 EDT
164 :     Tag: blume-20020522-shrink
165 :     Description:
166 :    
167 :     John tweaked yesterday's fix for 1131 to handle an out-of-memory
168 :     situation that comes up when allocating huge arrays.
169 :    
170 :     ----------------------------------------------------------------------
171 :     Name: Matthias Blume
172 : blume 1215 Date: 2002/05/21 16:00:00 EDT
173 :     Tag: Release_110_40
174 :     Description:
175 :    
176 :     New working release (110.40). New bootfiles.
177 :    
178 :     [Also: John Reppy fixed GC bug 1131.]
179 :    
180 :     ----------------------------------------------------------------------
181 :     Name: Matthias Blume
182 : blume 1212 Date: 2002/05/21 12:35:00 EDT
183 :     Tag: blume-20020521-cmdoc
184 :     Description:
185 :    
186 :     CM documentation update.
187 :    
188 :     ----------------------------------------------------------------------
189 :     Name: Matthias Blume
190 : blume 1211 Date: 2002/05/21 10:55:00 EDT
191 :     Tag: blume-20020521-misc
192 :     Description:
193 :    
194 :     - John tweaked runtime to be silent on heap export (except when
195 :     GC messages are on).
196 :     - I added a few more things (cross-compiling versions of CMB) to
197 :     config/preloads (as suggestions).
198 :    
199 :     ----------------------------------------------------------------------
200 :     Name: Matthias Blume
201 : blume 1208 Date: 2002/05/20 22:25:00 EDT
202 :     Tag: blume-20020520-controls
203 :     Description:
204 :    
205 :     - Added ControlUtil structure to control-lib.cm.
206 :     - Use it throughout.
207 :     - Used Controls facility to define MLRISC controls (as opposed to
208 :     registering MLRISC control ref cells with Controls after the fact)
209 :     - Fixed messed-up controls priorities.
210 :    
211 :     * Removed again all the stuff from config/preloads that one wouldn't
212 :     be able to preload at the time the initial heap image is built.
213 :     (Many libraries, e.g., CML, do not exist yet at this time. The
214 :     only libraries that can be preloaded via config/preloads are those
215 :     that come bundled with the bootfiles.)
216 :    
217 :     ----------------------------------------------------------------------
218 :     Name: Matthias Blume
219 : blume 1204 Date: 2002/05/20 10:59:00 EDT
220 :     Tag: blume-20020520-preloads
221 :     Description:
222 :    
223 :     Added a lot of commented-out suggestions for things to be included
224 :     in config/preloads.
225 :    
226 :     ----------------------------------------------------------------------
227 : leunga 1203 Name: Allen Leung
228 :     Date: 2002/05/18 14:20:00 EDT
229 :     Tag: leunga-20020518-mdl
230 :     Description:
231 :    
232 :     o Made the mdl tool stuff compile and run again.
233 :     o I've disabled all the stuff that depends on RTL specifications; they
234 :     are all badly broken anyway.
235 :    
236 :     ----------------------------------------------------------------------
237 : blume 1201 Name: Matthias Blume
238 :     Date: 2002/05/17 16:49:00 EDT
239 :     Tag: blume-20020517-controls
240 :     Description:
241 :    
242 :     0. John Reppy made several modifications to the SML/NJ library.
243 :     In particular, there is a shiny new controls-lib.cm.
244 :    
245 :     1. Pushed new controls interface through compiler so that everything
246 :     compiles again.
247 :    
248 :     2. Added FormatComb and FORMAT_COMB to the CML version of the
249 :     SML/NJ library (so that CML compiles again).
250 :    
251 :     3. Modified init scripts because XXX_DEFAULT environment variables
252 :     are no longer with us. (Boot-time initialization is now done
253 :     using the same environment variables that are also used for
254 :     startup-time initialization of controls.)
255 :    
256 :     ----------------------------------------------------------------------
257 : george 1192 Name: Lal George
258 :     Date: 2002/05/15 09:20:10 EDT
259 :     Tag: george-20020515-pseudo-op-decls
260 :     Description:
261 :    
262 :     All pseudo-ops emitted before the first segment declaration
263 :     such as TEXT, DATA, and BSS directives are assumed to be global
264 :     declarations and are emitted first in the assembly file. This is
265 :     useful in a number of situations where one has pseudo-ops that are not
266 :     specific to any segment, and also works around the constraint that one
267 :     cannot have client pseudo-ops in the TEXT segment.
268 :    
269 :     Because no segment is associated with these declarations it is
270 :     an error to allocate any space or objects before the first segment
271 :     directive and an exception will be raised. However, we cannot make
272 :     this check for client pseudo-ops.
273 :    
274 :     These top level declarations are a field in the CFG graph_info.
275 :     In theory you can continue to add to this field after the CFG has been
276 :     built -- provided you know what you are doing;-)
277 :    
278 :     ----------------------------------------------------------------------
279 : blume 1175 Name: Matthias Blume
280 : blume 1190 Date: 2002/05/13 16:40:00 EDT
281 :     Tag: blume-20020513-pp-etc
282 :     Description:
283 :    
284 :     A few minor bugfixes:
285 :    
286 :     - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
287 :     (Bogus printouts for redefined bindings still occur. Compiler
288 :     bug should no longer occur now. We need to redo the prettyprinter
289 :     from scratch.)
290 :    
291 :     - CM pathname printer now also adds escape sequences for ( and )
292 :    
293 :     - commend and docu fixes for ml-nlffi
294 :    
295 :     ----------------------------------------------------------------------
296 :     Name: Matthias Blume
297 : blume 1189 Date: 2002/05/10 16:40:00 EDT
298 :     Tag: blume-20020510-erg-textio
299 :     Description:
300 :    
301 :     Applied the following bugfix provided by Emden Gansner:
302 :    
303 :     Output is corrupted when outputSubstr is used rather than output.
304 :    
305 :     The problem occurs when a substring
306 :    
307 :     ss = (s, dataStart, dataLen)
308 :    
309 :     where dataStart > 0, fills a stream buffer with avail bytes left.
310 :     avail bytes of s, starting at index dataStart, are copied into the
311 :     buffer, the buffer is flushed, and then the remaining dataLen-avail
312 :     bytes of ss are copied into the beginning of the buffer. Instead of
313 :     starting this copy at index dataStart+avail in s, the current code
314 :     starts the copy at index avail.
315 :    
316 :     Fix:
317 :     In text-io-fn.sml, change line 695 from
318 :     val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
319 :     to
320 :     val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
321 :    
322 :     ----------------------------------------------------------------------
323 :     Name: Matthias Blume
324 : blume 1186 Date: 2002/04/12 13:55:00 EDT
325 :     Tag: blume-20020412-assyntax
326 :     Description:
327 :    
328 :     1. Grabbed newer assyntax.h from the XFree86 project.
329 :     2. Fiddled with how to compile X86.prim.asm without warnings.
330 :     3. (Very) Minor cleanup in CM.
331 :    
332 :     ----------------------------------------------------------------------
333 :     Name: Matthias Blume
334 : blume 1185 Date: 2002/04/01 (no joke!) 17:07:00 EST
335 :     Tag: blume-20020401-x86div
336 :     Description:
337 :    
338 :     Added full support for div/mod/rem/quot on the x86, using the machine
339 :     instruction's two results (without clumsily recomputing the remainder)
340 :     directly where appropriate.
341 :    
342 :     Some more extensive power-of-two support was added to the x86 instruction
343 :     selector (avoiding expensive divs, mods, and muls where they can be
344 :     replaced with cheaper shifts and masks). However, this sort of thing
345 :     ought to be done earlier, e.g., within the CPS optimizer so that
346 :     all architectures benefit from it.
347 :    
348 :     The compiler compiles to a fixed point, but changes might be somewhat
349 :     fragile nevertheless. Please, report any strange things that you might
350 :     see wrt. div/mod/quot/rem...
351 :    
352 :     ----------------------------------------------------------------------
353 :     Name: Matthias Blume
354 : blume 1184 Date: 2002/03/29 17:22:00
355 :     Tag: blume-20020329-div
356 :     Description:
357 :    
358 :     Fixed my broken div/mod logic. Unfortunately, this means that the
359 :     inline code for div/mod now has one more comparison than before.
360 :     Fast paths (quotient > 0 or remainder = 0) are not affected, though.
361 :     The problem was with quotient = 0, because that alone does not tell
362 :     us which way the rounding went. One then has to look at whether
363 :     remainder and divisor have the same sign... :(
364 :    
365 :     Anyway, I replaced the bootfiles with fresh ones...
366 :    
367 :     ----------------------------------------------------------------------
368 :     Name: Matthias Blume
369 : blume 1183 Date: 2002/03/29 14:10:00 EST
370 :     Tag: blume-20020329-inlprims
371 :     Description:
372 :    
373 :     NEW BOOTFILES!!! Version number bumped to 110.39.3.
374 :    
375 :     Primops have changed. This means that the bin/boot-file formats have
376 :     changed as well.
377 :    
378 :     To make sure that there is no confusion, I made a new version.
379 :    
380 :    
381 :     CHANGES:
382 :    
383 :     * removed REMT from mltree (remainder should never overflow).
384 :    
385 :     * added primops to deal with divisions of all flavors to the frontend
386 :    
387 :     * handled these primops all the way through so they map to their respective
388 :     MLRISC support
389 :    
390 :     * used these primops in the implementation of Int, Int32, Word, Word32
391 :    
392 :     * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
393 :    
394 :     * parameterized INLMIN, INLMAX, and INLABS by a numkind
395 :    
396 :     * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
397 :     floating point
398 :    
399 :     * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
400 :     and Real (but Real.abs maps to a separate floating-point-only primop)
401 :    
402 :    
403 :     TODO items:
404 :    
405 :     * Hacked Alpha32 instruction selection, disabling the selection of REMx
406 :     instructions because the machine instruction encoder cannot handle
407 :     them. (Hppa, PPC, and Sparc instruction selection did not handle
408 :     REM in the first place, and REM is supported by the x86 machine coder.)
409 :    
410 :     * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
411 :     selection phase. (The two can be streamlined because the hardware
412 :     delivers both quotient and remainder at the same time anyway.)
413 :    
414 :     * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
415 :     (Currently the behavior is inconsistent both across architectures and
416 :     wrt. the draft Basis spec.)
417 :    
418 :     * Word8 should eventually be handled natively, too.
419 :    
420 :     * There seems to be one serious bug in mltree-gen.sml. It appears, though,
421 :     as if there currently is no execution path that could trigger it in
422 :     SML/NJ. (The assumptions underlying functions arith and promotable do not
423 :     hold for things like multiplication and division.)
424 :    
425 :     ----------------------------------------------------------------------
426 :     Name: Matthias Blume
427 : blume 1181 Date: 2002/03/27 16:27:00 EST
428 :     Tag: blume-20020327-mlrisc-divisions
429 :     Description:
430 :    
431 :     Added support for all four division operations (ML's div, mod, quot,
432 :     and rem) to MLRISC. In the course of doing so, I also rationalized
433 :     the naming (no more annoying switch-around of DIV and QUOT), by
434 :     parameterizing the operation by div_rounding_mode (which can be either
435 :     DIV_TO_ZERO or DIV_TO_NEGINF).
436 :    
437 :     The generic MLTreeGen functor takes care of compiling all four
438 :     operations down to only round-to-zero div.
439 :    
440 :     Missing pieces:
441 :    
442 :     * Doing something smarter than relying on MLTreeGen on architectures
443 :     like, e.g., the x86 where hardware division delivers both quotient and
444 :     remainder at the same time. With this, the implementation of the
445 :     round-to-neginf operations could be further streamlined.
446 :    
447 :     * Remove inlining support for div/mod/rem from the frontend and replace it
448 :     with primops that get carried through to the backend. Do this for all
449 :     int and word types.
450 :    
451 :     ----------------------------------------------------------------------
452 :     Name: Matthias Blume
453 : blume 1180 Date: 2002/03/25 17:25:00 EST
454 :     Tag: blume-20020325-divmod
455 :     Description:
456 :    
457 :     I improved (hopefully without breaking them) the implementation of Int.div,
458 :     Int.mod, and Int.rem. For this, the code in translate.sml now takes
459 :     advantage of the following observations:
460 :    
461 :     Let q = x quot y r = x rem y
462 :     d = x div y m = x mod y
463 :    
464 :     where "quot" is the round-to-zero version of integer division that
465 :     hardware usually provides. Then we have:
466 :    
467 :     r = x - q * y where neither the * nor the - will overflow
468 :     d = if q >= 0 orelse x = q * y then q else q - 1
469 :     where neither the * nor the - will overflow
470 :     m = if q >= 0 orelse r = 0 then r else r + y
471 :     where the + will not overflow
472 :    
473 :     This results in substantial simplification of the generated code.
474 :     The following table shows the number of CFG nodes and edges generated
475 :     for
476 :     fun f (x, y) = x OPER y
477 :     (* with OPER \in div, mod, quot, rem *)
478 :    
479 :    
480 :     OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
481 :     --------------------------------------------------------
482 :     div | 24 | 39 | 12 | 16
483 :     mod | 41 | 71 | 12 | 16
484 :     quot | 8 | 10 | 8 | 10
485 :     rem | 10 | 14 | 8 | 10
486 :    
487 :    
488 :     ----------------------------------------------------------------------
489 :     Name: Matthias Blume
490 : blume 1179 Date: 2002/03/25 22:06:00 EST
491 :     Tag: blume-20020325-cprotobug
492 :     Description:
493 :    
494 :     Fixed a bug in cproto (c prototype decoder).
495 :    
496 :     ----------------------------------------------------------------------
497 :     Name: Matthias Blume
498 : blume 1178 Date: 2002/03/25 16:00:00 EST
499 :     Tag: blume-20020325-raw-primops
500 :     Description:
501 :    
502 :     I did some cleanup to Allen's new primop code and
503 :     replaced yesterday's bootfiles with new ones.
504 :     (But they are stored in the same place.)
505 :    
506 :     ----------------------------------------------------------------------
507 :     Name: Matthias Blume
508 : blume 1175 Date: 2002/03/24 22:40:00 EST
509 :     Tag: blume-20020324-bootfiles
510 :     Description:
511 :    
512 :     Made the bootfiles that Allen asked for.
513 :    
514 :     ----------------------------------------------------------------------
515 : leunga 1174 Name: Allen Leung
516 :     Date: 2002/03/23 15:50:00 EST
517 :     Tag: leunga-20020323-flint-cps-rcc-primops
518 :     Description:
519 :    
520 :     1. Changes to FLINT primops:
521 :    
522 :     (* make a call to a C-function;
523 :     * The primop carries C function prototype information and specifies
524 :     * which of its (ML-) arguments are floating point. C prototype
525 :     * information is for use by the backend, ML information is for
526 :     * use by the CPS converter. *)
527 :     | RAW_CCALL of { c_proto: CTypes.c_proto,
528 :     ml_args: ccall_type list,
529 :     ml_res_opt: ccall_type option,
530 :     reentrant : bool
531 :     } option
532 :     (* Allocate uninitialized storage on the heap.
533 :     * The record is meant to hold short-lived C objects, i.e., they
534 :     * are not ML pointers. With the tag, the representation is
535 :     * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
536 :     *)
537 :     | RAW_RECORD of {tag:bool,sz:int}
538 :     and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
539 :    
540 :     2. These CPS primops are now overloaded:
541 :    
542 :     rawload of {kind:numkind}
543 :     rawstore of {kind:numkind}
544 :    
545 :     The one argument form is:
546 :    
547 :     rawload {kind} address
548 :    
549 :     The two argument form is:
550 :    
551 :     rawload {kind} [ml object, byte-offset]
552 :    
553 :     3. RAW_CCALL/RCC now takes two extra arguments:
554 :    
555 :     a. The first is whether the C call is reentrant, i.e., whether
556 :     ML state should be saved and restored.
557 :     b. The second argument is a string argument specifying the name of
558 :     library and the C function.
559 :    
560 :     These things are currently not handled in the code generator, yet.
561 :    
562 :     4. In CProto,
563 :    
564 :     An encoding type of "bool" means "ml object" and is mapped into
565 :     C prototype of PTR. Note that "bool" is different than "string",
566 :     even though "string" is also mapped into PTR, because "bool"
567 :     is assigned an CPS type of BOGt, while "string" is assigned INT32t.
568 :    
569 :     5. Pickler/unpicker
570 :    
571 :     Changed to handle RAW_RECORD and newest RAW_CCALL
572 :    
573 :     6. MLRiscGen,
574 :    
575 :     1. Changed to handle the new rawload/rawstore/rawrecord operators.
576 :     2. Code for handling C Calls has been moved to a new module CPSCCalls,
577 :     in the file CodeGen/cpscompile/cps-c-calls.sml
578 :    
579 :     7. Added the conditional move operator
580 :    
581 :     condmove of branch
582 :    
583 :     to cps. Generation of this is still buggy so it is currently
584 :     disabled.
585 :    
586 :     ----------------------------------------------------------------------
587 : george 1168 Name: Lal George
588 :     Date: 2002/03/22 14:18:25 EST
589 : george 1171 Tag: george-20020322-cps-branch-prob
590 : george 1168 Description:
591 :    
592 : george 1171 Implemented the Ball-Larus branch prediction-heuristics, and
593 : george 1168 incorporated graphical viewers for control flow graphs.
594 :    
595 : george 1171 Ball-Larus Heuristics:
596 : george 1168 ---------------------
597 :     See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
598 :    
599 :     By design it uses the Dempster-Shafer theory for combining
600 :     probabilities. For example, in the function:
601 :    
602 :     fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
603 :    
604 :     the ball-larus heuristics predicts that the n=0 is unlikely
605 :     (OH-heuristic), and the 'then' branch is unlikely because of the
606 :     RH-heuristic -- giving the 'then' branch an even lower combined
607 : george 1171 probability using the Dempster-Shafer theory.
608 : george 1168
609 :     Finally, John Reppy's loop analysis in MLRISC, further lowers the
610 :     probability of the 'then' branch because of the loop in the else
611 :     branch.
612 :    
613 :    
614 :     Graphical Viewing:
615 :     ------------------
616 :     I merely plugged in Allen's graphical viewers into the compiler. The
617 :     additional code is not much. At the top level, saying:
618 :    
619 :     Control.MLRISC.getFlag "cfg-graphical-view" := true;
620 :    
621 :     will display the graphical view of the control flow graph just before
622 :     back-patching. daVinci must be in your path for this to work. If
623 :     daVinci is not available, then the default viewer can be changed
624 :     using:
625 :    
626 :     Control.MLRISC.getString "viewer"
627 :    
628 :     which can be set to "dot" or "vcg" for the corresponding viewers. Of
629 :     course, these viewers must be in your path.
630 :    
631 :     The above will display the compilation unit at the level of clusters,
632 :     many of which are small, boring, and un-interesting. Also setting:
633 :    
634 :     Control.MLRISC.getInt "cfg-graphical-view_size"
635 :    
636 :     will display clusters that are larger than the value set by the above.
637 :    
638 :    
639 :     ----------------------------------------------------------------------
640 : blume 1159 Name: Matthias Blume
641 :     Date: 2002/03/21 22:20:00 EST
642 :     Tag: blume-20020321-kmp-bugfix
643 :     Description:
644 :    
645 :     Changed the interface to the KMP routine in PreString and fixed
646 :     a minor bug in one place where it was used.
647 :    
648 :     ----------------------------------------------------------------------
649 : leunga 1156 Name: Allen Leung
650 : leunga 1158 Date: 2002/03/21 20:30:00 EST
651 :     Tag: leunga-20020321-cfg
652 :     Description:
653 :    
654 :     Fixed a potential problem in cfg edge splitting.
655 :    
656 :     ----------------------------------------------------------------------
657 :     Name: Allen Leung
658 : leunga 1156 Date: 2002/03/21 17:15:00 EST
659 :     Tag: leunga-20020321-x86-fp-cfg
660 :     Description:
661 :    
662 :     1. Recoded the buggy parts of x86-fp.
663 :    
664 :     a. All the block reordering code has been removed.
665 :     We now depend on the block placement phases to do this work.
666 :    
667 :     b. Critical edge splitting code has been simplified and moved into the
668 :     CFG modules, as where they belong.
669 :    
670 :     Both of these were quite buggy and complex. The code is now much, much
671 :     simpler.
672 :    
673 :     2. X86 backend.
674 :    
675 :     a. Added instructions for 64-bit support. Instruction selection for
676 :     64-bit has not been committed, however, since that
677 :     requires changes to MLTREE which haven't been approved by
678 :     Lal and John.
679 :    
680 :     b. Added support for FUCOMI and FUCOMIP when generating code for
681 :     PentiumPro and above. We only generate these instructions in
682 :     the fast-fp mode.
683 :    
684 :     c. Added cases for JP and JNP in X86FreqProps.
685 :    
686 :     3. CFG
687 :    
688 :     CFG now has a bunch of methods for edge splitting and merging.
689 :    
690 :     4. Machine description.
691 :    
692 :     John's simplification of MLTREE_BASIS.fcond broke a few machine
693 :     description things:
694 :    
695 :     rtl-build.{sig,sml} and hppa.mdl fixed.
696 :    
697 :     NOTE: the machine description stuff in the repository is still broken.
698 :     Again, I can't put my fixes in because that involves
699 :     changes to MLTREE.
700 :    
701 :     ----------------------------------------------------------------------
702 : blume 1145 Name: Matthias Blume
703 : blume 1155 Date: 2002/03/20 15:55:00 EST
704 :     Tag: blume-20020320-kmp
705 :     Description:
706 :    
707 :     Implemented Knuth-Morris-Pratt string matching in PreString and used
708 :     it for String.isSubstring, Substring.isSubstring, and
709 :     Substring.position.
710 :    
711 :     (Might need some stress-testing. Simple examples worked fine.)
712 :    
713 :     ----------------------------------------------------------------------
714 :     Name: Matthias Blume
715 : blume 1152 Date: 2002/03/19 16:37:00 EST
716 :     Tag: blume-20020319-witnesses
717 :     Description:
718 :    
719 :     Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
720 :    
721 :     This implements a generic mechanism for changing constness qualifiers
722 :     anywhere within big C types without resorting to outright "casts".
723 :     (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
724 :     modify the constness at the outermost level.)
725 :     The implementation of "convert" is based on the idea of "witness"
726 :     values -- values that are not used by the operation but whose types
727 :     "testify" to their applicability. On the implementation side, "convert"
728 :     is simply a projection (returning its second curried argument). With
729 :     cross-module inlining, it should not result in any machine code being
730 :     generated.
731 :    
732 :     ----------------------------------------------------------------------
733 :     Name: Matthias Blume
734 : blume 1148 Date: 2002/03/15 16:40:00 EST
735 :     Tag: blume-20020315-basis
736 :     Description:
737 :    
738 :     Provided (preliminary?) implementations for
739 :    
740 :     {String,Substring}.{concatWith,isSuffix,isSubstring}
741 :    
742 :     and
743 :    
744 :     Substring.full
745 :    
746 :     Those are in the Basis spec but they were missing in SML/NJ.
747 :    
748 :     ----------------------------------------------------------------------
749 :     Name: Matthias Blume
750 : blume 1145 Date: 2002/03/14 21:30:00 EST
751 :     Tag: blume-20020314-controls
752 :     Description:
753 :    
754 :     Controls:
755 :     ---------
756 :    
757 :     1. Factored out the recently-added Controls : CONTROLS stuff and put
758 :     it into its own library $/controls-lib.cm. The source tree for
759 :     this is under src/smlnj-lib/Controls.
760 :    
761 :     2. Changed the names of types and functions in this interface, so they
762 :     make a bit more "sense":
763 :    
764 :     module -> registry
765 :     'a registry -> 'a group
766 :    
767 :     3. The interface now deals in ref cells only. The getter/setter interface
768 :     is (mostly) gone.
769 :    
770 :     4. Added a function that lets one register an already-existing ref cell.
771 :    
772 :     5. Made the corresponding modifications to the rest of the code so that
773 :     everything compiles again.
774 :    
775 :     6. Changed the implementation of Controls.MLRISC back to something closer
776 :     to the original. In particular, this module (and therefore MLRISC)
777 :     does not depend on Controls. There now is some link-time code in
778 :     int-sys.sml that registers the MLRISC controls with the Controls
779 :     module.
780 :    
781 :     CM:
782 :     ---
783 :    
784 :     * One can now specify the lambda-split aggressiveness in init.cmi.
785 :    
786 :     ----------------------------------------------------------------------
787 : leunga 1142 Name: Allen Leung
788 :     Date: 2002/03/13 17:30:00 EST
789 :     Tag: leunga-20020313-x86-fp-unary
790 :     Description:
791 :    
792 :     Bug fix for:
793 :    
794 :     > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
795 :     > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
796 :     > - fun f(x,(y,z)) = Real.~ y;
797 :     > [autoloading]
798 :     > [autoloading done]
799 :     > fchsl (%eax), 184(%esp)
800 :     > Error: MLRisc bug: X86MCEmitter.emitInstr
801 :     >
802 :     > uncaught exception Error
803 :     > raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
804 :    
805 :     The problem was that the code generator did not generate any fp registers
806 :     in this case, and the ra didn't know that it needed to run the X86FP phase to
807 :     translate the pseudo fp instruction. This only happened with unary fp
808 :     operators in certain situations.
809 :    
810 :     ----------------------------------------------------------------------
811 : blume 1137 Name: Matthias Blume
812 : blume 1140 Date: 2002/03/13 14:00:00 EST
813 :     Tag: blume-20020313-overload-etc
814 :     Description:
815 :    
816 :     1. Added _overload as a synonym for overload for backward compatibility.
817 :     (Control.overloadKW must be true for either version to be accepted.)
818 :    
819 :     2. Fixed bug in install script that caused more things to be installed
820 :     than what was requested in config/targets.
821 :    
822 :     3. Made CM aware of the (_)overload construct so that autoloading
823 :     works.
824 :    
825 :     ----------------------------------------------------------------------
826 :     Name: Matthias Blume
827 : blume 1138 Date: 2002/03/12 22:03:00 EST
828 :     Tag: blume-20020312-url
829 :     Description:
830 :    
831 :     Forgot to update BOOT and srcarchiveurl.
832 :    
833 :     ----------------------------------------------------------------------
834 :     Name: Matthias Blume
835 : blume 1137 Date: 2002/03/12 17:30:00 EST
836 :     Tag: blume-20020312-version110392
837 :     Description:
838 :    
839 :     Yet another version number bump (because of small changes to the
840 :     binfile format). Version number is now 110.39.2. NEW BOOTFILES!
841 :    
842 :     Changes:
843 :    
844 :     The new pid generation scheme described a few weeks ago was overly
845 :     complicated. I implemented a new mechanism that is simpler and
846 :     provides a bit more "stability": Once CM has seen a compilation
847 :     unit, it keeps its identity constant (as long as you do not delete
848 :     those crucial CM/GUID/* files). This means that when you change
849 : blume 1145 an interface, compile, then go back to the old interface, and
850 : blume 1137 compile again, you arrive at the original pid.
851 :    
852 :     There now also is a mechanism that instructs CM to use the plain
853 :     environment hash as a module's pid (effectively making its GUID
854 :     the empty string). For this, "noguid" must be specified as an
855 :     option to the .sml file in question within its .cm file.
856 :     This is most useful for code that is being generated by tools such
857 :     as ml-nlffigen (because during development programmers tend to
858 :     erase the tool's entire output directory tree including CM's cached
859 :     GUIDs). "noguid" is somewhat dangerous (since it can be used to locally
860 :     revert to the old, broken behavior of SML/NJ, but in specific cases
861 :     where there is no danger of interface confusion, its use is ok
862 :     (I think).
863 :    
864 :     ml-nlffigen by default generates "noguid" annotations. They can be
865 :     turned off by specifying -guid in its command line.
866 :    
867 :     ----------------------------------------------------------------------
868 : george 1136 Name: Lal George
869 :     Date: 2002/03/12 12 14:42:36 EST
870 :     Tag: george-20020312-frequency-computation
871 :     Description:
872 :    
873 :     Integrated jump chaining and static block frequency into the
874 :     compiler. More details and numbers later.
875 : leunga 1025
876 : george 1003 ----------------------------------------------------------------------
877 : george 1133 Name: Lal George
878 :     Date: 2002/03/11 11 22:38:53 EST
879 :     Tag: george-20020311-jump-chain-elim
880 :     Description:
881 :    
882 :     Tested the jump chain elimination on all architectures (except the
883 :     hppa). This is on by default right now and is profitable for the
884 :     alpha and x86, however, it may not be profitable for the sparc and ppc
885 :     when compiling the compiler.
886 :    
887 :     The gc test will typically jump to a label at the end of the cluster,
888 :     where there is another jump to an external cluster containing the actual
889 :     code to invoke gc. This is to allow factoring of common gc invocation
890 :     sequences. That is to say, we generate:
891 :    
892 :     f:
893 :     testgc
894 :     ja L1 % jump if above to L1
895 :    
896 :     L1:
897 :     jmp L2
898 :    
899 :    
900 :     After jump chain elimination the 'ja L1' instructions is converted to
901 :     'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
902 :     up being implemented in their long form (if L2 is far away) using:
903 :    
904 :     jbe L3 % jump if below or equal to L3
905 :     jmp L2
906 :     L3:
907 :     ...
908 :    
909 :    
910 :     For large compilation units L2 may be far away.
911 :    
912 :    
913 :     ----------------------------------------------------------------------
914 : blume 1132 Name: Matthias Blume
915 :     Date: 2002/03/11 13:30:00 EST
916 :     Tag: blume-20020311-mltreeeval
917 :     Description:
918 :    
919 :     A functor parameter was missing.
920 :    
921 :     ----------------------------------------------------------------------
922 : leunga 1130 Name: Allen Leung
923 : leunga 1131 Date: 2002/03/11 10:30:00 EST
924 : leunga 1142 Tag: leunga-20020311-runtime-string0
925 : leunga 1131 Description:
926 :    
927 :     The representation of the empty string now points to a
928 :     legal null terminated C string instead of unit. It is now possible
929 :     to convert an ML string into C string with InlineT.CharVector.getData.
930 :     This compiles into one single machine instruction.
931 :    
932 :     ----------------------------------------------------------------------
933 :     Name: Allen Leung
934 : leunga 1130 Date: 2002/03/10 23:55:00 EST
935 :     Tag: leunga-20020310-x86-call
936 :     Description:
937 :    
938 :     Added machine generation for CALL instruction (relative displacement mode)
939 :    
940 :     ----------------------------------------------------------------------
941 : blume 1128 Name: Matthias Blume
942 :     Date: 2002/03/08 16:05:00
943 :     Tag: blume-20020308-entrypoints
944 :     Description:
945 :    
946 :     Version number bumped to 110.39.1. NEW BOOTFILES!
947 :    
948 :     Entrypoints: non-zero offset into a code object where execution should begin.
949 :    
950 :     - Added the notion of an entrypoint to CodeObj.
951 :     - Added reading/writing of entrypoint info to Binfile.
952 :     - Made runtime system bootloader aware of entrypoints.
953 :     - Use the address of the label of the first function given to mlriscGen
954 :     as the entrypoint. This address is currently always 0, but it will
955 :     not be 0 once we turn on block placement.
956 :     - Removed the linkage cluster code (which was The Other Way(tm) of dealing
957 :     with entry points) from mlriscGen.
958 :    
959 :     ----------------------------------------------------------------------
960 : leunga 1127 Name: Allen Leung
961 :     Date: 2002/03/07 20:45:00 EST
962 :     Tag: leunga-20020307-x86-cmov
963 :     Description:
964 :    
965 :     Bug fixes for CMOVcc on x86.
966 :    
967 :     1. Added machine code generation for CMOVcc
968 :     2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
969 :     3. CMOVcc cannot have an immediate operand as argument.
970 :    
971 :     ----------------------------------------------------------------------
972 : blume 1126 Name: Matthias Blume
973 :     Date: 2002/03/07 16:15:00 EST
974 :     Tag: blume-20020307-controls
975 :     Description:
976 :    
977 :     This is a very large but mostly boring patch which makes (almost)
978 :     every tuneable compiler knob (i.e., pretty much everything under
979 :     Control.* plus a few other things) configurable via both the command
980 :     line and environment variables in the style CM did its configuration
981 :     until now.
982 :    
983 :     Try starting sml with '-h' (or, if you are brave, '-H')
984 :    
985 :     To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
986 :     implements the underlying generic mechanism.
987 :    
988 :     The interface to some of the existing such facilities has changed somewhat.
989 :     For example, the MLRiscControl module now provides mkFoo instead of getFoo.
990 :     (The getFoo interface is still there for backward-compatibility, but its
991 :     use is deprecated.)
992 :    
993 :     The ml-build script passes -Cxxx=yyy command-line arguments through so
994 :     that one can now twiddle the compiler settings when using this "batch"
995 :     compiler.
996 :    
997 :     TODO items:
998 :    
999 :     We should go through and throw out all controls that are no longer
1000 :     connected to anything. Moreover, we should go through and provide
1001 :     meaningful (and correct!) documentation strings for those controls
1002 :     that still are connected.
1003 :    
1004 :     Currently, multiple calls to Controls.new are accepted (only the first
1005 :     has any effect). Eventually we should make sure that every control
1006 :     is being made (via Controls.new) exactly once. Future access can then
1007 :     be done using Controls.acc.
1008 :    
1009 :     Finally, it would probably be a good idea to use the getter-setter
1010 :     interface to controls rather than ref cells. For the time being, both
1011 :     styles are provided by the Controls module, but getter-setter pairs are
1012 :     better if thread-safety is of any concern because they can be wrapped.
1013 :    
1014 :     *****************************************
1015 :    
1016 :     One bug fix: The function blockPlacement in three of the MLRISC
1017 :     backpatch files used to be hard-wired to one of two possibilities at
1018 :     link time (according to the value of the placementFlag). But (I
1019 :     think) it should rather sense the flag every time.
1020 :    
1021 :     *****************************************
1022 :    
1023 :     Other assorted changes (by other people who did not supply a HISTORY entry):
1024 :    
1025 :     1. the cross-module inliner now works much better (Monnier)
1026 :     2. representation of weights, frequencies, and probabilities in MLRISC
1027 :     changed in preparation of using those for weighted block placement
1028 :     (Reppy, George)
1029 :    
1030 :     ----------------------------------------------------------------------
1031 : george 1116 Name: Lal George
1032 : george 1124 Date: 2002/03/07 14:44:24 EST 2002
1033 :     Tag: george-20020307-weighted-block-placement
1034 :    
1035 :     Tested the weighted block placement optimization on all architectures
1036 :     (except the hppa) using AMPL to generate the block and edge frequencies.
1037 :     Changes were required in the machine properties to correctly
1038 :     categorize trap instructions. There is an MLRISC flag
1039 :     "weighted-block-placement" that can be used to enable weighted block
1040 :     placement, but this will be ineffective without block/edge
1041 :     frequencies (coming soon).
1042 :    
1043 :    
1044 :     ----------------------------------------------------------------------
1045 :     Name: Lal George
1046 : george 1116 Date: 2002/03/05 17:24:48 EST
1047 :     Tag: george-20020305-linkage-cluster
1048 :    
1049 : george 1124 In order to support the block placement optimization, a new cluster
1050 :     is generated as the very first cluster (called the linkage cluster).
1051 :     It contains a single jump to the 'real' entry point for the compilation
1052 :     unit. Block placement has no effect on the linkage cluster itself, but
1053 :     all the other clusters have full freedom in the manner in which they
1054 :     reorder blocks or functions.
1055 : george 1116
1056 :     On the x86 the typical linkage code that is generated is:
1057 :     ----------------------
1058 :     .align 2
1059 :     L0:
1060 :     addl $L1-L0, 72(%esp)
1061 : george 1124 jmp L1
1062 : george 1116
1063 :    
1064 :     .align 2
1065 :     L1:
1066 :     ----------------------
1067 :    
1068 :     72(%esp) is the memory location for the stdlink register. This
1069 :     must contain the address of the CPS function being called. In the
1070 : george 1124 above example, it contains the address of L0; before
1071 : george 1116 calling L1 (the real entry point for the compilation unit), it
1072 :     must contain the address for L1, and hence
1073 :    
1074 :     addl $L1-L0, 72(%esp)
1075 :    
1076 : george 1124 I have tested this on all architectures except the hppa.The increase
1077 :     in code size is of course negligible
1078 : george 1116
1079 :     ----------------------------------------------------------------------
1080 : leunga 1115 Name: Allen Leung
1081 :     Date: 2002/03/03 13:20:00 EST
1082 :     Tag: leunga-20020303-mlrisc-tools
1083 :    
1084 :     Added #[ ... ] expressions to mlrisc tools
1085 :    
1086 :     ----------------------------------------------------------------------
1087 : blume 1096 Name: Matthias Blume
1088 : blume 1098 Date: 2002/02/27 12:29:00 EST
1089 :     Tag: blume-20020227-cdebug
1090 :     Description:
1091 :    
1092 :     - made types in structure C and C_Debug to be equal
1093 :     - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
1094 :     - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
1095 :    
1096 :     ----------------------------------------------------------------------
1097 :     Name: Matthias Blume
1098 : blume 1096 Date: 2002/02/26 12:00:00 EST
1099 :     Tag: blume-20020226-ffi
1100 :     Description:
1101 :    
1102 :     1. Fixed a minor bug in CM's "noweb" tool:
1103 :     If numbering is turned off, then truly don't number (i.e., do not
1104 :     supply the -L option to noweb). The previous behavior was to supply
1105 :     -L'' -- which caused noweb to use the "default" line numbering scheme.
1106 :     Thanks to Chris Richards for pointing this out (and supplying the fix).
1107 :    
1108 :     2. Once again, I reworked some aspects of the FFI:
1109 :    
1110 :     A. The incomplete/complete type business:
1111 :    
1112 :     - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
1113 :     gone!
1114 :     - ML types representing an incomplete type are now *equal* to
1115 :     ML types representing their corresponding complete types (just like
1116 :     in C). This is still safe because ml-nlffigen will not generate
1117 :     RTTI for incomplete types, nor will it generate functions that
1118 :     require access to such RTTI. But when ML code generated from both
1119 :     incomplete and complete versions of the C type meet, the ML types
1120 :     are trivially interoperable.
1121 :    
1122 :     NOTE: These changes restore the full generality of the translation
1123 :     (which was previously lost when I eliminated functorization)!
1124 :    
1125 :     B. Enum types:
1126 :    
1127 :     - Structure C now has a type constructor "enum" that is similar to
1128 :     how the "su" constructor works. However, "enum" is not a phantom
1129 :     type because each "T enum" has values (and is isomorphic to
1130 :     MLRep.Signed.int).
1131 :     - There are generic access operations for enum objects (using
1132 :     MLRep.Signed.int).
1133 :     - ml-nlffigen will generate a structure E_foo for each "enum foo".
1134 :     * The structure contains the definition of type "mlrep" (the ML-side
1135 :     representation type of the enum). Normally, mlrep is the same
1136 :     as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
1137 :     then mlrep will be defined as a datatype -- thus facilitating
1138 :     pattern matching on mlrep values.
1139 :     ("-ec" will be suppressed if there are duplicate values in an
1140 :     enumeration.)
1141 :     * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
1142 :     will be generated for each C enum constant xxx.
1143 :     * Conversion functions m2i and i2m convert between mlrep and
1144 :     MLRep.Signed.int. (Without "-ec", these functions are identities.)
1145 :     * Coversion functions c and ml convert between mlrep and "tag enum".
1146 :     * Access functions (get/set) fetch and store mlrep values.
1147 :     - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
1148 :     enumerations are merged into one single enumeration represented by
1149 :     structure E_'.
1150 :    
1151 :     ----------------------------------------------------------------------
1152 : leunga 1080 Name: Allen Leung
1153 : leunga 1094 Date: 2002/02/25 04:45:00 EST
1154 :     Tag: leunga-20020225-cps-spill
1155 :    
1156 :     This is a new implementation of the CPS spill phase.
1157 :     The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
1158 :     In case of problems, replace it with the old file spill.sml
1159 :    
1160 :     The current compiler runs into some serious performance problems when
1161 :     constructing a large record. This can happen when we try to compile a
1162 :     structure with many items. Even a very simple structure like the following
1163 :     makes the compiler slow down.
1164 :    
1165 :     structure Foo = struct
1166 :     val x_1 = 0w1 : Word32.int
1167 :     val x_2 = 0w2 : Word32.int
1168 :     val x_3 = 0w3 : Word32.int
1169 :     ...
1170 :     val x_N = 0wN : Word32.int
1171 :     end
1172 :    
1173 :     The following table shows the compile time, from N=1000 to N=4000,
1174 :     with the old compiler:
1175 :    
1176 :     N
1177 :     1000 CPS 100 spill 0.04u 0.00s 0.00g
1178 :     MLRISC ra 0.06u 0.00s 0.05g
1179 :     (spills = 0 reloads = 0)
1180 :     TOTAL 0.63u 0.07s 0.21g
1181 :    
1182 :     1100 CPS 100 spill 8.25u 0.32s 0.64g
1183 :     MLRISC ra 5.68u 0.59s 3.93g
1184 :     (spills = 0 reloads = 0)
1185 :     TOTAL 14.71u 0.99s 4.81g
1186 :    
1187 :     1500 CPS 100 spill 58.55u 2.34s 1.74g
1188 :     MLRISC ra 5.54u 0.65s 3.91g
1189 :     (spills = 543 reloads = 1082)
1190 :     TOTAL 65.40u 3.13s 6.00g
1191 :    
1192 :     2000 CPS 100 spill 126.69u 4.84s 3.08g
1193 :     MLRISC ra 0.80u 0.10s 0.55g
1194 :     (spills = 42 reloads = 84)
1195 :     TOTAL 129.42u 5.10s 4.13g
1196 :    
1197 :     3000 CPS 100 spill 675.59u 19.03s 11.64g
1198 :     MLRISC ra 2.69u 0.27s 1.38g
1199 :     (spills = 62 reloads = 124)
1200 :     TOTAL 682.48u 19.61s 13.99g
1201 :    
1202 :     4000 CPS 100 spill 2362.82u 56.28s 43.60g
1203 :     MLRISC ra 4.96u 0.27s 2.72g
1204 :     (spills = 85 reloads = 170)
1205 :     TOTAL 2375.26u 57.21s 48.00g
1206 :    
1207 :     As you can see the old cps spill module suffers from some serious
1208 : leunga 1115 performance problem. But since I cannot decipher the old code fully,
1209 :     instead of patching the problems up, I'm reimplementing it
1210 : leunga 1094 with a different algorithm. The new code is more modular,
1211 :     smaller when compiled, and substantially faster
1212 :     (O(n log n) time and O(n) space). Timing of the new spill module:
1213 :    
1214 :     4000 CPS 100 spill 0.02u 0.00s 0.00g
1215 :     MLRISC ra 0.25u 0.02s 0.15g
1216 :     (spills=1 reloads=3)
1217 :     TOTAL 7.74u 0.34s 1.62g
1218 :    
1219 :     Implementation details:
1220 :    
1221 :     As far as I can tell, the purpose of the CPS spill module is to make sure the
1222 :     number of live variables at any program point (the bandwidth)
1223 :     does not exceed a certain limit, which is determined by the
1224 :     size of the spill area.
1225 :    
1226 :     When the bandwidth is too large, we decrease the register pressure by
1227 :     packing live variables into spill records. How we achieve this is
1228 :     completely different than what we did in the old code.
1229 :    
1230 :     First, there is something about the MLRiscGen code generator
1231 :     that we should be aware of:
1232 :    
1233 :     o MLRiscGen performs code motion!
1234 :    
1235 :     In particular, it will move floating point computations and
1236 :     address computations involving only the heap pointer to
1237 :     their use sites (if there is only a single use).
1238 :     What this means is that if we have a CPS record construction
1239 :     statement
1240 :    
1241 :     RECORD(k,vl,w,e)
1242 :    
1243 :     we should never count the new record address w as live if w
1244 :     has only one use (which is often the case).
1245 :    
1246 :     We should do something similar to floating point, but the transformation
1247 :     there is much more complex, so I won't deal with that.
1248 :    
1249 :     Secondly, there are now two new cps primops at our disposal:
1250 :    
1251 :     1. rawrecord of record_kind option
1252 :     This pure operator allocates some uninitialized storage from the heap.
1253 :     There are two forms:
1254 :    
1255 :     rawrecord NONE [INT n] allocates a tagless record of length n
1256 :     rawrecord (SOME rk) [INT n] allocates a tagged record of length n
1257 :     and initializes the tag.
1258 :    
1259 :     2. rawupdate of cty
1260 :     rawupdate cty (v,i,x)
1261 :     Assigns to x to the ith component of record v.
1262 :     The storelist is not updated.
1263 :    
1264 :     We use these new primops for both spilling and increment record construction.
1265 :    
1266 :     1. Spilling.
1267 :    
1268 :     This is implemented with a linear scan algorithm (but generalized
1269 :     to trees). The algorithm will create a single spill record at the
1270 :     beginning of the cps function and use rawupdate to spill to it,
1271 :     and SELECT or SELp to reload from it. So both spills and reloads
1272 :     are fine-grain operations. In contrast, in the old algorithm
1273 :     "spills" have to be bundled together in records.
1274 :    
1275 :     Ideally, we should sink the spill record construction to where
1276 :     it is needed. We can even split the spill record into multiple ones
1277 :     at the places where they are needed. But CPS is not a good
1278 :     representation for global code motion, so I'll keep it simple and
1279 :     am not attempting this.
1280 :    
1281 :     2. Incremental record construction (aka record splitting).
1282 :    
1283 :     Long records with many component values which are simulatenously live
1284 :     (recall that single use record addresses are not considered to
1285 :     be live) are constructed with rawrecord and rawupdate.
1286 :     We allocate space on the heap with rawrecord first, then gradually
1287 :     fill it in with rawupdate. This is the technique suggested to me
1288 :     by Matthias.
1289 :    
1290 :     Some restrictions on when this is applicable:
1291 :     1. It is not a VECTOR record. The code generator currently does not handle
1292 :     this case. VECTOR record uses double indirection like arrays.
1293 :     2. All the record component values are defined in the same "basic block"
1294 :     as the record constructor. This is to prevent speculative
1295 :     record construction.
1296 :    
1297 :     ----------------------------------------------------------------------
1298 :     Name: Allen Leung
1299 : leunga 1086 Date: 2002/02/22 01:02:00 EST
1300 :     Tag: leunga-20020222-mlrisc-tools
1301 :    
1302 :     Minor bug fixes in the parser and rewriter
1303 :    
1304 :     ----------------------------------------------------------------------
1305 :     Name: Allen Leung
1306 : leunga 1085 Date: 2002/02/21 20:20:00 EST
1307 :     Tag: leunga-20020221-peephole
1308 :    
1309 :     Regenerated the peephole files. Some contained typos in the specification
1310 :     and some didn't compile because of pretty printing bugs in the old version
1311 :     of 'nowhere'.
1312 :    
1313 :     ----------------------------------------------------------------------
1314 :     Name: Allen Leung
1315 : leunga 1079 Date: 2002/02/19 20:20:00 EST
1316 :     Tag: leunga-20020219-mlrisc-tools
1317 :     Description:
1318 :    
1319 :     Minor bug fixes to the mlrisc-tools library:
1320 :    
1321 :     1. Fixed up parsing colon suffixed keywords
1322 :     2. Added the ability to shut the error messages up
1323 :     3. Reimplemented the pretty printer and fixed up/improved
1324 :     the pretty printing of handle and -> types.
1325 :     4. Fixed up generation of literal symbols in the nowhere tool.
1326 :     5. Added some SML keywords to to sml.sty
1327 :    
1328 :     ----------------------------------------------------------------------
1329 :     Name: Matthias Blume
1330 : blume 1078 Date: 2002/02/19 16:20:00 EST
1331 :     Tag: blume-20020219-cmffi
1332 :     Description:
1333 :    
1334 :     A wild mix of changes, some minor, some major:
1335 :    
1336 :     * All C FFI-related libraries are now anchored under $c:
1337 :     $/c.cm --> $c/c.cm
1338 :     $/c-int.cm --> $c/internals/c-int.cm
1339 :     $/memory.cm --> $c/memory/memory.cm
1340 :    
1341 :     * "make" tool (in CM) now treats its argument pathname slightly
1342 :     differently:
1343 :     1. If the native expansion is an absolute name, then before invoking
1344 :     the "make" command on it, CM will apply OS.Path.mkRelative
1345 :     (with relativeTo = OS.FileSys.getDir()) to it.
1346 :     2. The argument will be passed through to subsequent phases of CM
1347 :     processing without "going native". In particular, if the argument
1348 :     was an anchored path, then "make" will not lose track of that anchor.
1349 :    
1350 :     * Compiler backends now "know" their respective C calling conventions
1351 :     instead of having to be told about it by ml-nlffigen. This relieves
1352 :     ml-nlffigen from one of its burdens.
1353 :    
1354 :     * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
1355 :    
1356 :     * Export C_DEBUG and C_Debug from $c/c.cm.
1357 :    
1358 :     * C type encoding in ml-nlffi-lib has been improved to model the conceptual
1359 :     subtyping relationship between incomplete pointers and their complete
1360 :     counterparts. For this, ('t, 'c) ptr has been changed to 'o ptr --
1361 :     with the convention of instantiating 'o with ('t, 'c) obj whenever
1362 :     the pointer target type is complete. In the incomplete case, 'o
1363 :     will be instantiated with some "'c iobj" -- a type obtained by
1364 :     using one of the functors PointerToIncompleteType or PointerToCompleteType.
1365 :    
1366 :     Operations that work on both incomplete and complete pointer types are
1367 :     typed as taking an 'o ptr while operations that require the target to
1368 :     be known are typed as taking some ('t, 'c) obj ptr.
1369 :    
1370 :     voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
1371 :     where void is an eqtype without any values. This makes it possible
1372 :     to work on voidptr values using functions meant to operate on light
1373 :     incomplete pointers.
1374 :    
1375 :     * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
1376 :     been vastly simplified.
1377 :    
1378 :     ----------------------------------------------------------------------
1379 :     Name: Matthias Blume
1380 : blume 1076 Date: 2002/02/19 10:48:00 EST
1381 :     Tag: blume-20020219-pqfix
1382 :     Description:
1383 :    
1384 :     Applied Chris Okasaki's bug fix for priority queues.
1385 :    
1386 :     ----------------------------------------------------------------------
1387 :     Name: Matthias Blume
1388 : blume 1073 Date: 2002/02/15 17:05:00
1389 : blume 1069 Tag: Release_110_39
1390 :     Description:
1391 :    
1392 : blume 1073 Last-minute retagging is becoming a tradition... :-(
1393 :    
1394 :     This is the working release 110.39.
1395 :    
1396 :     ----------------------------------------------------------------------
1397 :     Name: Matthias Blume
1398 :     Date: 2002/02/15 16:00:00 EST
1399 :     Tag: Release_110_39-orig
1400 :     Description:
1401 :    
1402 : blume 1069 Working release 110.39. New bootfiles.
1403 :    
1404 : blume 1073 (Update: There was a small bug in the installer so it wouldn't work
1405 :     with all shells. So I retagged. -Matthias)
1406 :    
1407 : blume 1069 ----------------------------------------------------------------------
1408 :     Name: Matthias Blume
1409 : blume 1068 Date: 2002/02/15 14:17:00 EST
1410 :     Tag: blume-20020215-showbindings
1411 :     Description:
1412 :    
1413 :     Added EnvRef.listBoundSymbols and CM.State.showBindings. Especially
1414 :     the latter can be useful for exploring what bindings are available at
1415 :     the interactive prompt. (The first function returns only the list
1416 :     of symbols that are really bound, the second prints those but also the
1417 :     ones that CM's autoloading mechanism knows about.)
1418 :    
1419 :     ----------------------------------------------------------------------
1420 :     Name: Matthias Blume
1421 : blume 1067 Date: 2002/02/15 12:08:00 EST
1422 :     Tag: blume-20020215-iptrs
1423 :     Description:
1424 :    
1425 :     Two improvements to ml-nlffigen:
1426 :    
1427 :     1. Write files only if they do not exist or if their current contents
1428 :     do not coincide with what's being written. (That is, avoid messing
1429 :     with the time stamps unless absolutely necessary.)
1430 :    
1431 :     2. Implement a "repository" mechanism for generated files related
1432 :     to "incomplete pointer types". See the README file for details.
1433 :    
1434 :     ----------------------------------------------------------------------
1435 :     Name: Matthias Blume
1436 : blume 1066 Date: 2002/02/14 11:50:00 EST
1437 :     Tag: blume-20020214-quote
1438 :     Description:
1439 :    
1440 :     Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm). This is required
1441 :     because of the new and improved tag generation scheme. (Thanks to Allen
1442 :     Leung for pointing it out.)
1443 :    
1444 :     ----------------------------------------------------------------------
1445 : george 1065 Name: Lal George
1446 :     Date: 2002/02/14 09:55:27 EST 2002
1447 :     Tag: george-20020214-isabelle-bug
1448 :     Description:
1449 :    
1450 :     Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
1451 :     of Isabelle on the x86.
1452 :    
1453 :     From Allen:
1454 :     -----------
1455 :     I've found the problem:
1456 :    
1457 :     in ra-core.sml, I use the counter "blocked" to keep track of the
1458 :     true number of elements in the freeze queue. When the counter goes
1459 :     to zero, I skip examining the queue. But I've messed up the
1460 :     bookkeeping in combine():
1461 :    
1462 :     else ();
1463 :     case !ucol of
1464 :     PSEUDO => (if !cntv > 0 then
1465 :     (if !cntu > 0 then blocked := !blocked - 1 else ();
1466 :     ^^^^^^^^^^^^^^^^^^^^^^^
1467 :     moveu := mergeMoveList(!movev, !moveu)
1468 :     )
1469 :     else ();
1470 :    
1471 :     combine() is called to coalesce two nodes u and v.
1472 :     I think I was thinking that if the move counts of u and v are both
1473 :     greater than zero then after they are coalesced then one node is
1474 :     removed from the freeze queue. Apparently I was thinking that
1475 :     both u and v are of low degree, but that's clearly not necessarily true.
1476 :    
1477 :    
1478 :     02/12/2002:
1479 :     Here's the patch. HOL now compiles.
1480 :    
1481 :     I don't know how this impact on performance (compile
1482 :     time or runtime). This bug caused the RA (especially on the x86)
1483 :     to go thru the potential spill phase when there are still nodes on the
1484 :     freeze queue.
1485 :    
1486 :    
1487 :    
1488 :    
1489 :     ----------------------------------------------------------------------
1490 : blume 1058 Name: Matthias Blume
1491 : blume 1064 Date: 2002/02/13 22:40:00 EST
1492 :     Tag: blume-20020213-fptr-rtti
1493 :     Description:
1494 :    
1495 :     Fixed a bug in ml-nlffigen that was introduced with one of the previous
1496 :     updates.
1497 :    
1498 :     ----------------------------------------------------------------------
1499 :     Name: Matthias Blume
1500 : blume 1063 Date: 2002/02/13 16:41:00 EST
1501 :     Tag: blume-20020213-cmlpq
1502 :     Description:
1503 :    
1504 :     Added new priority queue export symbols (which have just been added to
1505 :     smlnj-lib.cm) to CML's version of smlnj-lib.cm. (Otherwise CML would
1506 :     not compile and the installer would choke.)
1507 :    
1508 :     ----------------------------------------------------------------------
1509 :     Name: Matthias Blume
1510 : blume 1062 Date: 2002/02/13 16:15:00 EST
1511 :     Tag: blume-20020213-various
1512 :     Description:
1513 :    
1514 :     1. More tweaks to ml-nlffigen:
1515 :    
1516 :     - better internal datastructures (resulting in slight speedup)
1517 :     - "-match" option requires exact match
1518 :     - "localized" gensym counters (untagged structs/unions nested within
1519 :     other structs/unions or within typedefs get a fresh counter; their
1520 :     tag will be prefixed by a concatenation of their parents' tags)
1521 :     - bug fixes (related to calculation of transitive closure of types
1522 :     to be included in the output)
1523 :    
1524 :     2. Minor Basis updates:
1525 :    
1526 :     - added implementations for List.collate and Option.app
1527 :    
1528 :     ----------------------------------------------------------------------
1529 :     Name: Matthias Blume
1530 : blume 1060 Date: 2002/02/11 15:55:00 EST
1531 :     Tag: blume-20020211-gensym
1532 :     Description:
1533 :    
1534 :     Added a "-gensym" option to command line of ml-nlffigen. This can be
1535 :     used to specify a "stem" -- a string that is inserted in all "gensym'd"
1536 :     names (ML structure names that correspond to unnamed C structs, unions,
1537 :     and enums), so that separate runs of ml-nlffigen do not clash.
1538 :    
1539 :     ----------------------------------------------------------------------
1540 :     Name: Matthias Blume
1541 : blume 1059 Date: 2002/02/11 12:05:00 EST
1542 :     Tag: blume-20020211-gensml
1543 :     Description:
1544 :    
1545 :     A quick fix for a problem with GenSML (in the pgraph-util library):
1546 :     Make generation of toplevel "local" optional. (Strictly speaking,
1547 :     signature definitions within "local" are not legal SML.)
1548 :    
1549 :     Other than that: updates to INSTALL and cm/TODO.
1550 :    
1551 :     ----------------------------------------------------------------------
1552 :     Name: Matthias Blume
1553 : blume 1058 Date: 2002/02/08 15:00:00 EST
1554 :     Tag: blume-20020208-uniquepid
1555 :     Description:
1556 :    
1557 :     0. Version number has been bumped to 110.38.1. NEW BOOTFILES!!!
1558 :    
1559 :     1. The installer (config/install.sh) has gotten smarter:
1560 :    
1561 :     - Configuration options are a bit easier to specify now
1562 :     (in config/targets).
1563 :     - Bug in recognizing .tar.bz2 files fixed.
1564 :     - Installer automatically resolves dependencies between
1565 :     configuration options (e.g., if you ask for eXene, you will
1566 :     also get cml -- regardless whether you asked for it or not).
1567 :     - Installer can run in "quieter mode" by setting the environment
1568 :     variable INSTALL_QUIETLY to "true". "Quieter" does not mean
1569 :     "completely silent", though.
1570 :     - Build HashCons library as part of smlnj-lib.
1571 :    
1572 :     2. A new scheme for assigning persistent identifiers to compilation
1573 :     units (and, by extension, to types etc.) has been put into place.
1574 :     This fixes a long-standing bug where types and even dynamic values
1575 :     can get internally confused, thereby compromising type safety
1576 :     (abstraction) and dynamic correctness. See
1577 :    
1578 :     http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
1579 :    
1580 :     for an example of how things could go wrong until now.
1581 :    
1582 :     The downside of the new scheme is that pids are not quite as
1583 :     persistent as they used to be: CM will generate a fresh pid
1584 :     for every compilation unit that it thinks it sees for the first
1585 :     time. That means that if you compile starting from a clean, fresh
1586 :     source tree at two different times, you end up with different
1587 :     binaries.
1588 :    
1589 :     Cutoff recompilation, however, has not been compromised because
1590 :     CM keeps pid information in special caches between runs.
1591 :    
1592 :     ----------------------------------------------------------------------
1593 : george 1053 Name: Lal George
1594 : george 1055 Date: 2002/02/07 15:34:13 EST 2002
1595 :     Tag: <none>
1596 :     Description:
1597 :    
1598 :     Compilers that generate assembly code may produce global labels
1599 :     whose value is resolved at link time. The various peephole optimization
1600 :     modules did not take this in account.
1601 :    
1602 :     TODO. The Labels.addrOf function should really return an option
1603 :     type so that clients are forced to deal with this issue, rather
1604 :     than an exception being raised.
1605 :    
1606 :     ----------------------------------------------------------------------
1607 :     Name: Lal George
1608 : george 1053 Date: 2002/02/06 13:55:02 EST
1609 :     Tag: george-20020206-ra-breakup
1610 :     Description:
1611 :    
1612 :     1. A bug fix from Allen.
1613 :    
1614 :     A typo causes extra fstp %st(0)'s to be generated at compensation
1615 :     edges, which might cause stack underflow traps at runtime. This
1616 :     occurs in fft where there are extraneous fstps right before the 'into'
1617 :     trap instruction (in this case they are harmless since none of the
1618 :     integers overflow.)
1619 :    
1620 : george 1054 2. Pulled out various utility modules that were embedded in the modules
1621 : george 1053 of the register allocator. I need these modules for other purposes, but
1622 : george 1054 they are not complete enough to put into a library (just yet).
1623 : george 1053 ----------------------------------------------------------------------
1624 : blume 1049 Name: Matthias Blume
1625 :     Date: 2002/01/31 16:05:00 EST
1626 :     Tag: blume-20020131-sparc-ccalls
1627 :     Description:
1628 :    
1629 :     1. C-calls on Sparc needlessly allocated a huge chunk (96 bytes)
1630 :     of extra stack space by mistake. Fixed.
1631 :    
1632 :     2. Bug in logic of handling of command-line options in ml-nlffigen fixed.
1633 :    
1634 :     ----------------------------------------------------------------------
1635 : leunga 1048 Name: Allen Leung
1636 :     Date: 2002/01/30
1637 :     Tag: leunga-20020130-nowhere-bug-fix
1638 :     Description:
1639 :    
1640 :     MLRISC bug fixes:
1641 :     1. Fixed a bindings computation bug in the 'nowhere' program generator tool.
1642 :     2. MachineInt.fromString was negating its value.
1643 :    
1644 :     ----------------------------------------------------------------------
1645 : blume 1039 Name: Matthias Blume
1646 : blume 1047 Date: 2002/01/29
1647 :     Tag: blume-20020129-INSTALL
1648 :     Description:
1649 :    
1650 :     - Added somewhat detailed installation instructions (file INSTALL).
1651 :     - Fixed curl-detection bug in config/install.sh.
1652 :     - It is now possible to select the URL getter using the URLGETTER
1653 :     environment variable:
1654 :    
1655 :     not set / "unknown" --> automatic detection (script tries wget,
1656 :     curl, and lynx)
1657 :     "wget" / "curl" / "lynx" --> use the specified program (script "knows"
1658 :     how to properly invoke them)
1659 :     other --> use $URLGETTER directly, it must take
1660 :     precisely two command-line arguments
1661 :     (source URL and destination file name)
1662 :    
1663 :     ----------------------------------------------------------------------
1664 :     Name: Matthias Blume
1665 : blume 1044 Date: 2002/01/28
1666 : blume 1045 Tag: blume-20020128-sparc-ccalls
1667 :     Description:
1668 :    
1669 :     - Fixed problem with calculation of "used" registers in sparc-c-calls.
1670 :     - Make use of the allocParam argument in sparc-c-calls.
1671 :    
1672 :     ----------------------------------------------------------------------
1673 :     Name: Matthias Blume
1674 :     Date: 2002/01/28
1675 : blume 1044 Tag: blume-20020128-allocParam
1676 :     Description:
1677 :    
1678 :     John Reppy: Changes c-calls API to accept client-callback for
1679 :     allocating extra stack space.
1680 :     me: Corresponding changes to mlriscGen (using a dummy argument that
1681 :     does not change the current behavior).
1682 :    
1683 :     ----------------------------------------------------------------------
1684 :     Name: Matthias Blume
1685 : blume 1040 Date: 2002/01/28 12:00:00
1686 :     Tag: Release_110_38
1687 :     Description:
1688 :    
1689 :     This time for real!!!
1690 :    
1691 :     ----------------------------------------------------------------------
1692 :     Name: Matthias Blume
1693 : blume 1039 Date: 2002/01/28 10:56:00 EST
1694 :     Tag: blume-20020128-retraction
1695 :     Description:
1696 :    
1697 :     0. Retracted earlier 110.38. (The Release_110_38 tag has been replaced
1698 :     with blume-Release_110_38-retracted.)
1699 :    
1700 :     1. Fixed a problem with incorrect rounding modes in real64.sml.
1701 :     (Thanks to Andrew Mccreight <andrew.mccreight@yale.edu>.)
1702 :    
1703 :     2. A bug in ml-nlffigen related to the handling of unnamed structs, unions,
1704 :     and enums fixed. The naming of corresponding ML identifiers should
1705 :     now be consistent again.
1706 :    
1707 :     ----------------------------------------------------------------------
1708 : leunga 1038 Name: Allen Leung
1709 :     Date: 2002/01/27
1710 :     Tag: leunga-20020127-nowhere
1711 :     Description:
1712 :    
1713 :     Added a target called nowhere in the configuration scripts.
1714 :     Enabling this will build the MLRISC 'nowhere' tool (for translating
1715 :     programs with where-clauses into legal SML code) during installation.
1716 :    
1717 :     ----------------------------------------------------------------------
1718 : blume 1036 Name: Matthias Blume
1719 : leunga 1038 Date: 2002/01/25 21:27:00 EST
1720 : blume 1039 Tag: blume-Release_110_38-retracted
1721 : blume 1037 Description:
1722 :    
1723 :     Call it a (working) release! Version is 110.38. Bootfiles are ready.
1724 :    
1725 :     README will be added later.
1726 :    
1727 : blume 1039 !!! NOTE: Re-tagged as blume-Release_110_38-retracted. Original tag
1728 :     (Release_110_38) removed. Reason: Last-minute bug fixes.
1729 :    
1730 : blume 1037 ----------------------------------------------------------------------
1731 :     Name: Matthias Blume
1732 : blume 1036 Date: 2002/01/25
1733 :     Tag: blume-20020125-ffi
1734 :     Description:
1735 :    
1736 :     A large number of tweaks and improvements to ml-nlffi-lib and
1737 :     ml-nlffigen:
1738 :    
1739 :     - ML represenation types have been streamlined
1740 :     - getter and setter functions work with concrete values, not abstract
1741 :     ones where possible
1742 :     - ml-nlffigen command line more flexible (see README file there)
1743 :     - some bugs have been fixed (hopefully)
1744 :    
1745 :     ----------------------------------------------------------------------
1746 : george 1034 Name: Lal George
1747 :     Date: 2002/01/24
1748 :     Tag: george-20020124-risc-ra-interface
1749 :     Description:
1750 :    
1751 :     There is a dramatic simplification in the interface to the
1752 :     register allocator for RISC architectures as a result of making
1753 :     parallel copy instructions explicit.
1754 :    
1755 :     ----------------------------------------------------------------------
1756 : blume 1028 Name: Matthias Blume
1757 : blume 1032 Date: 2002/01/22
1758 :     Tag: blume-20020122-x86-ccalls
1759 :     Description:
1760 :    
1761 :     Bug fix for c-calls on x86 (having to do with how char- and
1762 :     short-arguments are being handled).
1763 :    
1764 :     ----------------------------------------------------------------------
1765 :     Name: Matthias Blume
1766 : blume 1031 Date: 2002/01/21
1767 :     Tag: blume-20020121-ff
1768 :     Description:
1769 :    
1770 :     Another day of fiddling with the FFI...
1771 :    
1772 :     1. Bug fix/workaround: CKIT does not complain about negative array
1773 :     dimensions, so ml-nlffigen has to guard itself against this possibility.
1774 :     (Otherwise a negative dimension would send it into an infinite loop.)
1775 :    
1776 :     2. Some of the abstract types (light objects, light pointers, most "base"
1777 :     types) in structure C are now eqtypes.
1778 :    
1779 :     3. Added constructors and test functions for NULL function pointers.
1780 :    
1781 :     ----------------------------------------------------------------------
1782 :     Name: Matthias Blume
1783 : blume 1028 Date: 2002/01/18
1784 : blume 1030 Tag: blume-20020118-ready-for-new-release
1785 :     Description:
1786 :    
1787 :     Made config/srcarchiveurl point to a new place. (Will provide boot
1788 :     files shortly.)
1789 :    
1790 :     Maybe we christen this to be 110.38?
1791 :    
1792 :     ----------------------------------------------------------------------
1793 :     Name: Matthias Blume
1794 :     Date: 2002/01/18
1795 : blume 1029 Tag: blume-20020118-more-ffifiddle
1796 :     Description:
1797 :    
1798 :     Today's FFI fiddling:
1799 :    
1800 :     - Provided a structure CGetSet with "convenient" versions of C.Get.* and
1801 :     C.Set.* that use concrete (MLRep.*) arguments and results instead
1802 :     of abstract ones.
1803 :    
1804 :     - Provided word-style bit operations etc. for "int" representation
1805 :     types in MLRep.S<Foo>Bitops where <Foo> ranges over Char, Int, Short,
1806 :     and Long.
1807 :    
1808 :     ----------------------------------------------------------------------
1809 :     Name: Matthias Blume
1810 :     Date: 2002/01/18
1811 : blume 1028 Tag: blume-20020118-use-x86-fp
1812 :     Description:
1813 :    
1814 :     Now that x86-fast-fp seems to be working, I turned it back on again
1815 :     by default. (Seems to work fine now, even with the FFI.)
1816 :    
1817 :     Other than that, I added some documentation about the FFI to
1818 :     src/ml-nlffigen/README and updated the FFI test examples in
1819 :     src/ml-nlffi-lib/Tests/*.
1820 :    
1821 :     ----------------------------------------------------------------------
1822 : leunga 1025 Name: Allen Leung
1823 :     Date: 2002/01/17
1824 :     Tag: leunga-20020117-x86-fast-fp-call
1825 :     Description:
1826 :    
1827 :     1. Fixed a problem with handling return fp values when x86's fast fp
1828 :     mode is turned on.
1829 :    
1830 :     2. Minor pretty printing fix for cellset. Print %st(0) as %st(0) instead
1831 :     of %f32.
1832 :    
1833 :     3. Added a constructor INT32lit to the ast of MLRISC tools.
1834 :    
1835 :     ----------------------------------------------------------------------
1836 : blume 1021 Name: Matthias Blume
1837 :     Date: 2002/01/16
1838 :     Tag: blume-20020116-ffifiddle
1839 :     Description:
1840 :    
1841 :     More fiddling with the FFI interface:
1842 :    
1843 :     - Make constness 'c instead of rw wherever possible. This eliminates
1844 :     the need for certain explicit coercions. (However, due to ML's
1845 :     value polymorphism, there will still be many cases where explicit
1846 :     coercions are necessary. Phantom types are not the whole answer
1847 :     to modeling a subtyping relationship in ML.)
1848 :    
1849 :     - ro/rw coersions for pointers added. (Avoids the detour through */&.)
1850 :    
1851 :     - "printf" test example added to src/ml-nlffi-lib/Tests. (Demonstrates
1852 :     clumsy workaround for varargs problem.)
1853 :    
1854 :     ----------------------------------------------------------------------
1855 : george 1016 Name: Lal George
1856 :     Date: 2002/01/15
1857 :     Tag: <none>
1858 :     Description:
1859 :    
1860 :     1. Since COPY instructions are no longer native to the architecture,
1861 :     a generic functor can be used to implement the expandCopies function.
1862 :    
1863 :     2. Allowed EXPORT and IMPORT pseudo-op declarations to appear inside a
1864 :     TEXT segment.
1865 :    
1866 :     ----------------------------------------------------------------------
1867 : blume 1011 Name: Matthias Blume
1868 : blume 1015 Date: 2002/01/15
1869 :     Tag: blume-20020115-ffiupdates
1870 :     Description:
1871 :    
1872 :     1. Fix for bug resulting in single-precision float values being returned
1873 :     incorrectly from FFI calls.
1874 :    
1875 :     2. Small modifications to C FFI API:
1876 :    
1877 :     - memory-allocation routines return straight objects (no options)
1878 :     and raise an exception in out-of-memory situations
1879 :     - unsafe extensions to cast between function pointers and pointers
1880 :     from/to ints
1881 :     - added structure C_Debug as an alternative to structure C where
1882 :     pointer-dereferencing (|*| and |*!) always check for null-pointers
1883 :     - added open_lib' to DynLinkage; open_lib' works like open_lib
1884 :     but also takes a (possibly empty) list of existing library handles
1885 :     that the current library depends on
1886 :    
1887 :     ----------------------------------------------------------------------
1888 :     Name: Matthias Blume
1889 : blume 1011 Date: 2002/01/10
1890 :     Tag: blume-20020110-newffigen
1891 :     Description:
1892 :    
1893 :     1. Updates to portable graph code.
1894 :    
1895 :     2. Major update to ml-nlffigen and ml-nlffi-lib. Things are much
1896 :     more scalable now so that even huge interfaces such as the one
1897 :     for GTK compile in finite time and space. :-)
1898 :     See src/ml-nlffigen/README for details on what's new.
1899 :    
1900 :     ----------------------------------------------------------------------
1901 : george 1003 Name: Lal George
1902 : george 1009 Date: 2001/01/09 14:31:35 EST 2002
1903 :     Tag: george-20011206-rm-native-copy
1904 :     Description:
1905 :    
1906 :     Removed the native COPY and FCOPY instructions
1907 :     from all the architectures and replaced it with the
1908 :     explicit COPY instruction from the previous commit.
1909 :    
1910 :     It is now possible to simplify many of the optimizations
1911 :     modules that manipulate copies. This has not been
1912 :     done in this change.
1913 :    
1914 :     ----------------------------------------------------------------------
1915 :     Name: Lal George
1916 : george 1003 Date: 2001/12/06 16:50:13 EST 2001
1917 :     Tag: george-20011206-mlrisc-instruction
1918 :     Description:
1919 : george 824
1920 : george 1003 Changed the representation of instructions from being fully abstract
1921 :     to being partially concrete. That is to say:
1922 :    
1923 :     from
1924 :     type instruction
1925 :    
1926 :     to
1927 :     type instr (* machine instruction *)
1928 :    
1929 :     datatype instruction =
1930 :     LIVE of {regs: C.cellset, spilled: C.cellset}
1931 :     | KILL of {regs: C.cellset, spilled: C.cellset}
1932 :     | COPYXXX of {k: CB.cellkind, dst: CB.cell list, src: CB.cell list}
1933 :     | ANNOTATION of {i: instruction, a: Annotations.annotation}
1934 :     | INSTR of instr
1935 :    
1936 :     This makes the handling of certain special instructions that appear on
1937 :     all architectures easier and uniform.
1938 :    
1939 :     LIVE and KILL say that a list of registers are live or killed at the
1940 :     program point where they appear. No spill code is generated when an
1941 :     element of the 'regs' field is spilled, but the register is moved to
1942 :     the 'spilled' (which is present, more for debugging than anything else).
1943 :    
1944 :     LIVE replaces the (now deprecated) DEFFREG instruction on the alpha.
1945 :     We used to generate:
1946 :    
1947 :     DEFFREG f1
1948 :     f1 := f2 + f3
1949 :     trapb
1950 :    
1951 :     but now generate:
1952 :    
1953 :     f1 := f2 + f3
1954 :     trapb
1955 :     LIVE {regs=[f1,f2,f3], spilled=[]}
1956 :    
1957 :     Furthermore, the DEFFREG (hack) required that all floating point instruction
1958 :     use all registers mentioned in the instruction. Therefore f1 := f2 + f3,
1959 :     defines f1 and uses [f1,f2,f3]! This hack is no longer required resulting
1960 :     in a cleaner alpha implementation. (Hopefully, intel will not get rid of
1961 :     this architecture).
1962 :    
1963 :     COPYXXX is intended to replace the parallel COPY and FCOPY available on
1964 :     all the architectures. This will result in further simplification of the
1965 :     register allocator that must be aware of them for coalescing purposes, and
1966 :     will also simplify certain aspects of the machine description that provides
1967 :     callbacks related to parallel copies.
1968 :    
1969 :     ANNOTATION should be obvious, and now INSTR represents the honest to God
1970 :     machine instruction set!
1971 :    
1972 :     The <arch>/instructions/<arch>Instr.sml files define certain utility
1973 :     functions for making porting easier -- essentially converting upper case
1974 :     to lower case. All machine instructions (of type instr) are in upper case,
1975 :     and the lower case form generates an MLRISC instruction. For example on
1976 :     the alpha we have:
1977 :    
1978 :     datatype instr =
1979 :     LDA of {r:cell, b:cell, d:operand}
1980 :     | ...
1981 :    
1982 :     val lda : {r:cell, b:cell, d:operand} -> instruction
1983 :     ...
1984 :    
1985 :     where lda is just (INSTR o LDA), etc.
1986 :    
1987 : leunga 815 ----------------------------------------------------------------------
1988 : blume 986 Name: Matthias Blume
1989 : blume 993 Date: 2001/11/22 21:40:00 EST
1990 :     Tag: Release_110_37
1991 :     Description:
1992 :    
1993 :     Release 110.37. This time for real.
1994 :    
1995 :     ----------------------------------------------------------------------
1996 :     Name: Matthias Blume
1997 : blume 988 Date: 2001/11/21 16:35:00 EST
1998 :     Tag: blume-20011121-foot-in-mouth
1999 :     Description:
2000 :    
2001 :     Removed the "Release_110_37" tag because of a serious bug.
2002 :     This will be re-tagged once the bug is fixed.
2003 :    
2004 :     ----------------------------------------------------------------------
2005 :     Name: Matthias Blume
2006 : blume 987 Date: 2001/11/21 16:14:00 EST
2007 :     Tag: blume-20011121-forgottenfile
2008 :     Description:
2009 :    
2010 :     Forgot to add a file. (Just a .tex-file -- part of
2011 :     the CM manual source.)
2012 :    
2013 :     ----------------------------------------------------------------------
2014 :     Name: Matthias Blume
2015 : blume 986 Date: 2001/11/21 16:10:00 EST
2016 : blume 988 Tag: blume-20011121-invalid_110_37
2017 : blume 986 Description:
2018 :    
2019 : blume 988 Note: I removed the original tag "Release_110_37" from this commit
2020 :     because we found a serious bug in all non-x86 backends.
2021 :     - Matthias
2022 :    
2023 : blume 986 1. Modifications to the SML/NJ code generator and to the runtime system
2024 :     so that code object name strings are directly inserted into code
2025 :     objects at code generation time. The only business the runtime system
2026 :     has with this is now to read the name strings on occasions.
2027 :     (The encoding of the name string has also changed somewhat.)
2028 :    
2029 :     2. CM now implements a simple "set calculus" for specifying export lists.
2030 :     In particular, it is now possible to refer to the export lists of
2031 :     other libraries/groups/sources and form unions as well as differences.
2032 :     See the latest CM manual for details.
2033 :    
2034 :     3. An separate notion of "proxy" libraries has again be eliminated from
2035 :     CM's model. (Proxy libraries are now simply a special case of using
2036 :     the export list calculus.)
2037 :    
2038 :     4. Some of the existing libraries now take advantage of the new set
2039 :     calculus.
2040 :     (Notice that not all libraries have been converted because some
2041 :     of the existing .cm-files are supposed to be backward compatible
2042 :     with 110.0.x.)
2043 :    
2044 :     5. Some cleanup in stand-alone programs. (Don't use "exnMessage" -- use
2045 :     "General.exnMessage"! The former relies on a certain hook to be
2046 :     initialized, and that often does not happen in the stand-alone case.)
2047 :    
2048 :     ----------------------------------------------------------------------
2049 : george 984 Name: Lal George
2050 :     Date: 2001/11/21 13:56:18 EST
2051 :     Tag: george-2001121-pseudo-ops
2052 :     Description:
2053 :    
2054 :     Implemented a complete redesign of MLRISC pseudo-ops. Now there
2055 :     ought to never be any question of incompatabilities with
2056 :     pseudo-op syntax expected by host assemblers.
2057 :    
2058 :     For now, only modules supporting GAS syntax are implemented
2059 :     but more should follow, such as MASM, and vendor assembler
2060 :     syntax, e.g. IBM as, Sun as, etc.
2061 :    
2062 :     ----------------------------------------------------------------------
2063 : blume 977 Name: Matthias Blume
2064 :     Date: 2001/11/14 11:52:00 EST
2065 :     Tag: blume-20011114-srcname
2066 :     Description:
2067 :    
2068 :     1. Routed the name of the current source file to mlriscgen where it
2069 :     should be directly emitted into the code object. (This last part
2070 :     is yet to be done.)
2071 :    
2072 :     2. Some cleanup of the pgraph code to make it match the proposal that
2073 :     I put out the other day. (The proposal notwithstanding, things are
2074 :     still in flux here.)
2075 :    
2076 :     ----------------------------------------------------------------------
2077 : george 976 Name: Lal George
2078 :     Date: 2001/11/14 09:44:04 EST
2079 :     Tag:
2080 :     Description:
2081 :    
2082 :     Fix for a backpatching bug reported by Allen.
2083 :    
2084 :     Because the boundary between short and long span-dependent
2085 :     instructions is +/- 128, there are an astounding number of
2086 :     span-dependent instructions whose size is over estimated.
2087 :    
2088 :     Allen came up with the idea of letting the size of span
2089 :     dependent instructions be non-monotonic, for a maxIter
2090 :     number of times, after which the size must be monotonically
2091 :     increasing.
2092 :    
2093 :     This table shows the number of span-dependent instructions
2094 :     whose size was over-estimated as a function of maxIter, for the
2095 :     file Parse/parse/ml.grm.sml:
2096 :    
2097 :     maxIter # of instructions:
2098 :     10 687
2099 :     20 438
2100 :     30 198
2101 :     40 0
2102 :    
2103 :     In compiling the compiler, there is no significant difference in
2104 :     compilation speed between maxIter=10 and maxIter=40. Actually,
2105 :     my measurements showed that maxIter=40 was a tad faster than
2106 :     maxIter=10! Also 96% of the files in the compiler reach a fix
2107 :     point within 13 iterations, so fixing maxIter at 40, while high,
2108 :     is okay.
2109 :    
2110 :     ----------------------------------------------------------------------
2111 : blume 975 Name: Matthias Blume
2112 :     Date: 2001/10/31 15:25:00 EST
2113 :     Tag: blume-20011031-pgraph
2114 :     Description:
2115 :    
2116 :     CKIT:
2117 :     * Changed the "Function" constructor of type Ast.ctype to carry optional
2118 :     argument identifiers.
2119 :     * Changed the return type of TypeUtil.getFunction accordingly.
2120 :     * Type equality ignores the argument names.
2121 :     * TypeUtil.composite tries to preserve argument names but gives up quickly
2122 :     if there is a mismatch.
2123 :    
2124 :     installation script:
2125 :     * attempts to use "curl" if available (unless "wget" is available as well)
2126 :    
2127 :     CM:
2128 :     * has an experimental implementation of "portable graphs" which I will
2129 :     soon propose as an implementation-independent library format
2130 :     * there are also new libraries $/pgraph.cm and $/pgraph-util.cm
2131 :    
2132 :     NLFFI-LIB:
2133 :     * some cleanup (all cosmetic)
2134 :    
2135 :     NLFFIGEN:
2136 :     * temporarily disabled the mechanism that suppresses ML output for
2137 :     C definitions whose identifiers start with an underscore character
2138 :     * generate val bindings for enum constants
2139 :     * user can request that only one style (light or heavy) is being used;
2140 :     default is to use both (command-line arguments: -heavy and -light)
2141 :     * fixed bug in handling of function types involving incomplete pointers
2142 :     * generate ML entry points that take record arguments (i.e., using
2143 :     named arguments) for C functions that have a prototype with named
2144 :     arguments
2145 :     (see changes to CKIT)
2146 :    
2147 :     ----------------------------------------------------------------------
2148 : leunga 972 Name: Allen Leung
2149 :     Date: 2001/10/27 20:34:00 EDT
2150 :     Tag: leunga-20011027-x86-fast-fp-call
2151 :     Description:
2152 :    
2153 :     Fixed the bug described in blume-20010920-slowfp.
2154 :    
2155 :     The fix involves
2156 :     1. generating FCOPYs in FSTP in ia32-svid
2157 :     2. marking a CALL with the appropriate annotation
2158 :    
2159 :     ----------------------------------------------------------------------
2160 : blume 961 Name: Matthias Blume
2161 :     Date: 2001/10/16 11:32:00 EDT
2162 :     Tag: blume-20011016-netbsd
2163 :     Description:
2164 :    
2165 :     Underscore patch from Chris Richards (fixing problem with compiling
2166 :     runtime system under recent NetBSD).
2167 :    
2168 :     ----------------------------------------------------------------------
2169 : george 959 Name: Allen Leung
2170 :     Date: 2001/10/12 17:18:32 EDT 2001
2171 :     Tag: leung-20011012-x86-printflowgraph
2172 :     Description:
2173 :    
2174 :     X86RA now uses a valid (instead of dummy) PrintFlowgraph module.
2175 :    
2176 :     ----------------------------------------------------------------------
2177 : george 952 Name: Lal George
2178 : george 958 Date: 2001/10/11 23:51:34 EDT
2179 :     Tag: george-20011011-too-many-instrs
2180 :     Description:
2181 :    
2182 :     The representation of a program point never expected to see more
2183 :     than 65536 instructions in a basic block!
2184 :    
2185 :     ----------------------------------------------------------------------
2186 :     Name: Lal George
2187 : george 952 Date: 2001/10/09 09:41:37 EDT
2188 :     Tag: george-20011008-mlrisc-labels
2189 :     Description:
2190 :    
2191 :     Changed the machine description files to support printing of
2192 :     local and global labels in assembly code, based on host assembler
2193 :     conventions.
2194 :    
2195 :     ----------------------------------------------------------------------
2196 : blume 916 Name: Matthias Blume
2197 : blume 939 Date: 2001/09/25 15:25:00 EDT
2198 :     Tag: blume-20010925-exninfo
2199 :     Description:
2200 :    
2201 :     I provided a non-hook implementation of exnName (at the toplevel) and
2202 :     made the "dummy" implementation of exnMessage (at the toplevel) more
2203 :     useful: if nothing gets "hooked in", then at least you are going to
2204 :     see the exception name and a message indicating why you don't see more.
2205 :    
2206 :     [For the time being, programs that need exnMessage and want to use
2207 :     ml-build should either use General.exnMessage (strongly recommended) or
2208 :     refer to structure General at some other point so that CM sees a
2209 :     static dependency.]
2210 :    
2211 :     [Similar remarks go for "print" and "use": If you want to use their
2212 :     functionality in stand-alone programs generated by ml-build, then use
2213 :     TextIO.output and Backend.Interact.useFile (from $smlnj/compiler.cm).]
2214 :    
2215 :     ----------------------------------------------------------------------
2216 :     Name: Matthias Blume
2217 : blume 938 Date: 2001/09/20 17:28:00 EDT
2218 :     Tag: blume-20010920-slowfp
2219 :     Description:
2220 :    
2221 :     Allen says that x86-fast-fp is not safe yet, so I turned it off again...
2222 :    
2223 :     ----------------------------------------------------------------------
2224 :     Name: Matthias Blume
2225 : blume 937 Date: 2001/09/20 17:20:00 EDT
2226 :     Tag: blume-20010920-canonicalpaths
2227 :     Description:
2228 :    
2229 :     0. Updated the BOOT file (something that I forgot to do earlier).
2230 :    
2231 :     1. Small internal change to CM so that it avoids "/../" in filenames
2232 :     as much as possible (but only where it is safe).
2233 :    
2234 :     2. Changed config/_run-sml (resulting in a changed bin/.run-sml) so
2235 :     that arguments that contain delimiters are passed through correctly.
2236 :     This change also means that all "special" arguments of the form
2237 :     @SMLxxx... must come first.
2238 :    
2239 :     3. Changed install script to put relative anchor names for tool commands
2240 :     into pathconfig.
2241 :    
2242 :     ----------------------------------------------------------------------
2243 :     Name: Matthias Blume
2244 : blume 932 Date: 2001/09/18 15:35:00 EDT
2245 :     Tag: blume-20010918-readme11036
2246 :     Description:
2247 :    
2248 :     Added README files.
2249 :    
2250 :     ----------------------------------------------------------------------
2251 :     Name: Matthias Blume
2252 : blume 929 Date: 2001/09/18 11:45:00 EDT
2253 :     Tag: Release_110_36 (retag)
2254 :     Description:
2255 :    
2256 :     Fixed mistake in config/preloads. Retagged as 110.36.
2257 :    
2258 :     ----------------------------------------------------------------------
2259 :     Name: Matthias Blume
2260 : blume 928 Date: 2001/09/18 09:40:00 EDT
2261 : blume 929 Tag: Release_110_36_orig (tag changed)
2262 : blume 928 Description:
2263 :    
2264 :     New version (110.36). New bootfiles.
2265 :    
2266 :     ----------------------------------------------------------------------
2267 :     Name: Matthias Blume
2268 : blume 927 Date: 2001/09/14 16:15:00 EDT
2269 :     Tag: blume-20010914-x86fastfp
2270 :     Description:
2271 :    
2272 :     John committed some changes that Allen made, in particular a (hopefully)
2273 :     correctly working version of the x86-fp module.
2274 :    
2275 :     I changed the default setting of the Control.MLRISC.getFlag "x86-fast-fp"
2276 :     flag to "true". Everything seems to compile to a fixpoint ok, and
2277 :     "mandelbrot" speeds up by about 15%.
2278 :    
2279 :     ----------------------------------------------------------------------
2280 :     Name: Matthias Blume
2281 : blume 923 Date: 2001/09/13 11:20:00 EDT
2282 :     Tag: blume-20010913-minimal
2283 :     Description:
2284 :    
2285 :     1. Stefan Monnier's patch to fix a miscompilation problem that
2286 :     was brought to light by John Reppy's work on Moby.
2287 :    
2288 :     2. Implemented a minimal "structure Compiler" that contains just
2289 :     "version" and "architecture". The minimal version will be
2290 :     available when the full version is not. This is for backward-
2291 :     compatibility with code that wants to test Compiler.version.
2292 :    
2293 :     ----------------------------------------------------------------------
2294 :     Name: Matthias Blume
2295 : blume 920 Date: 2001/08/28 14:03:00 EDT
2296 :     Tag: blume-20010828-ml-lex
2297 :     Description:
2298 :    
2299 :     Fix for bug 1581, received from Neophytos Michael.
2300 :    
2301 :     ----------------------------------------------------------------------
2302 :     Name: Matthias Blume
2303 : blume 918 Date: 2001/08/27 11:20:00 EDT
2304 :     Tag: blume-20010827-readme11035
2305 :     Description:
2306 :    
2307 :     Fleshed out the README file for 110.35.
2308 :    
2309 :     ----------------------------------------------------------------------
2310 :     Name: Matthias Blume
2311 : blume 916 Date: 2001/08/24 17:10:00 EDT
2312 :     Tag: Release_110_35
2313 :     Description:
2314 :    
2315 :     New version number (110.35). New bootfiles.
2316 :    
2317 :     ----------------------------------------------------------------------
2318 : george 915 Name: Lal George
2319 : george 909 Date: 2001/08/24 13:47:18 EDT 2001
2320 :     Tag: george-20010824-MLRISC-graphs
2321 :     Description:
2322 :    
2323 :     removed clusters from MLRISC completely and replaced with graphs.
2324 :    
2325 :     ----------------------------------------------------------------------
2326 :     Name: Matthias Blume
2327 : blume 905 Date: 2001/08/23 17:50:00 EDT
2328 :     Tag: blume-20010823-toplevel
2329 :     Description:
2330 :    
2331 :     - some reorganization of the code that implements various kinds of
2332 :     environments in the compiler (static, dynamic, symbolic, combined)
2333 :     - re-implemented the EnvRef module so that evalStream works properly
2334 :     (if the stream contains references to "use", "CM.make", etc.)
2335 :     - cleaned up evalloop.sml and interact.sml (but they need more cleaning)
2336 :    
2337 :     ----------------------------------------------------------------------
2338 :     Name: Matthias Blume
2339 : blume 904 Date: 2001/08/20 15:50 EDT
2340 :     Tag: blume20010820-slipup
2341 :     Description:
2342 :    
2343 :     I forgot to commit a few files. Here they are...
2344 :    
2345 :     ----------------------------------------------------------------------
2346 :     Name: Matthias Blume
2347 : blume 903 Date: 2001/08/20 15:35:00 EDT
2348 :     Tag: blume-20010820-debugprof
2349 :     Description:
2350 :    
2351 :     !!!! NEW BOOTFILES !!!!
2352 :    
2353 :     This is another round of reorganizing the compiler sources. This
2354 :     time the main goal was to factor out all the "instrumentation"
2355 :     passes (for profiling and backtracing) into their own library.
2356 :     The difficulty was to do it in such a way that it does not depend
2357 :     on elaborate.cm but only on elabdata.cm.
2358 :    
2359 :     Therefore there have been further changes to both elaborate.cm and
2360 :     elabdata.cm -- more "generic" things have been moved from the former
2361 :     to the latter. As a result, I was forced to split the assignment
2362 :     of numbers indicating "primtyc"s into two portions: SML-generic and
2363 :     SML/NJ-specific. Since it would have been awkward to maintain,
2364 :     I bit the bullet and actually _changed_ the mapping between these
2365 :     numbers and primtycs. The bottom line of this is that you need
2366 :     a new set of bin- and bootfiles.
2367 :    
2368 :     I have built new bootfiles for all architectures, so doing a fresh
2369 :     checkout and config/install.sh should be all you need.
2370 :    
2371 :     The newly created library's name is
2372 :    
2373 :     $smlnj/viscomp/debugprof.cm
2374 :    
2375 :     and its sources live under
2376 :    
2377 :     src/compiler/DebugProf
2378 :    
2379 :     ----------------------------------------------------------------------
2380 :     Name: Matthias Blume
2381 : blume 902 Date: 2001/08/15 17:15:00 EDT
2382 :     Tag: blume-20010815-compreorg
2383 :     Description:
2384 :    
2385 :     This is a first cut at reorganizing the CM libraries that make up the
2386 :     core of the compiler. The idea is to separate out pieces that could
2387 :     be used independently by tools, e.g., the parser, the typechecker, etc.
2388 :    
2389 :     The current status is a step in this direction, but it is not quite
2390 :     satisfactory yet. Expect more changes in the future.
2391 :    
2392 :     Here is the current (new) organization...
2393 :    
2394 :     What used to be $smlnj/viscomp/core.cm is now divided into
2395 :     six CM libraries:
2396 :    
2397 :     $smlnj/viscomp/basics.cm
2398 :     /parser.cm
2399 :     /elabdata.cm
2400 :     /elaborate.cm
2401 :     /execute.cm
2402 :     /core.cm
2403 :    
2404 :     The CM files for these libraries live under src/system/smlnj/viscomp.
2405 :     All these libraries are proxy libraries that contain precisely
2406 :     one CM library component. Here are the locations of the components
2407 :     (all within the src/compiler tree):
2408 :    
2409 :     Basics/basics.cm
2410 :     Parse/parser.cm
2411 :     ElabData/elabdata.cm
2412 :     Elaborator/elaborate.cm
2413 :     Execution/execute.cm
2414 :     core.cm
2415 :    
2416 :     [This organization is the same that has been used already
2417 :     for a while for the architecture-specific parts of the visible
2418 :     compiler and for the old version of core.cm.]
2419 :    
2420 :     As you will notice, many source files have been moved from their
2421 :     respective original locations to a new home in one of the above
2422 :     subtrees.
2423 :    
2424 :     The division of labor between the new libraries is the following:
2425 :    
2426 :     basics.cm:
2427 :     - Simple, basic definitions that pertain to many (or all) of
2428 :     the other libraries.
2429 :     parser.cm:
2430 :     - The SML parser, producing output of type Ast.dec.
2431 :     - The type family for Ast is also defined and exported here.
2432 :     elabdata.cm:
2433 :     - The datatypes that describe input and output of the elaborator.
2434 :     This includes types, absyn, and static environments.
2435 :     elaborator.cm:
2436 :     - The SML/NJ type checker and elaborator.
2437 :     This maps an Ast.dec (with a given static environment) to
2438 :     an Absyn.dec (with a new static environment).
2439 :     - This libraries implements certain modules that used to be
2440 :     structures as functors (to remove dependencies on FLINT).
2441 :     execute.cm:
2442 :     - Everything having to do with executing binary code objects.
2443 :     - Dynamic environments.
2444 :     core.cm:
2445 :     - SML/NJ-specific instantiations of the elaborator and MLRISC.
2446 :     - Top-level modules.
2447 :     - FLINT (this should eventually become its own library)
2448 :    
2449 :     Notes:
2450 :    
2451 :     I am not 100% happy with the way I separated the elaborator (and its
2452 :     data structures) from FLINT. Two instances of the same problem:
2453 :    
2454 :     1. Data structures contain certain fields that carry FLINT-specific
2455 :     information. I hacked around this using exn and the property list
2456 :     module from smlnj-lib. But the fact that there are middle-end
2457 :     specific fields around at all is a bit annoying.
2458 :    
2459 :     2. The elaborator calculates certain FLINT-related information. I tried
2460 :     to make this as abstract as I could using functorization, but, again,
2461 :     the fact that the elaborator has to perform calculations on behalf
2462 :     of the middle-end at all is not nice.
2463 :    
2464 :     3. Having to used exn and property lists is unfortunate because it
2465 :     weakens type checking. The other alternative (parameterizing
2466 :     nearly *everything*) is not appealing, though.
2467 :    
2468 :     I removed the "rebinding =" warning hack because due to the new organization
2469 :     it was awkward to maintain it. As a result, the compiler now issues some of
2470 :     these warnings when compiling init.cmi during bootstrap compilation. On
2471 :     the plus side, you also get a warning when you do, for example:
2472 :     val op = = Int32.+
2473 :     which was not the case up to now.
2474 :    
2475 :     I placed "assign" and "deref" into the _Core structure so that the
2476 :     code that deals with the "lazy" keyword can find them there. This
2477 :     removes the need for having access to the primitive environment
2478 :     during elaboration.
2479 :    
2480 :     ----------------------------------------------------------------------
2481 :     Name: Matthias Blume
2482 : blume 898 Date: 2001/08/13
2483 :     Tag: blume-20010813-closures
2484 :     Description:
2485 :    
2486 :     This fix was sent to us by Zhong Shao. It is supposed to improve the
2487 :     performance of certain loops by avoiding needless closure allocation.
2488 :    
2489 :     ----------------------------------------------------------------------
2490 : george 890 Name: Lal George
2491 : george 896 Date: 2001/07/31 10:03:23 EDT 2001
2492 :     Tag: george-20010731-x86-fmalloc
2493 :     Description: Fixed bug in x86 calls
2494 :    
2495 :     There was a bug where call instructions would mysteriously
2496 :     vanish. The call instruction had to be one that returned
2497 :     a floating point value.
2498 :    
2499 :     ----------------------------------------------------------------------
2500 :     Name: Lal George
2501 : george 890 Date: 2001/07/19 16:36:29 EDT 2001
2502 :     Tag: george-20010719-simple-cells
2503 :     Description:
2504 :    
2505 :     I have dramatically simplified the interface for CELLS in MLRISC.
2506 :    
2507 :     In summary, the cells interface is broken up into three parts:
2508 :    
2509 :     1. CellsBasis : CELLS_BASIS
2510 :    
2511 :     CellsBasis is a top level structure and common for all
2512 :     architectures. it contains the definitions of basic datatypes
2513 :     and utility functions over these types.
2514 :    
2515 :     2. functor Cells() : CELLS
2516 :    
2517 :     Cells generates an interface for CELLS that incorporates the
2518 :     specific resources on the target architecture, such as the
2519 :     presence of special register classes, their number and size,
2520 :     and various useful substructures.
2521 :    
2522 :     3. <ARCH>CELLS
2523 :    
2524 :     e.g. SparcCells: SPARCCELLS
2525 :    
2526 :     <ARCH>CELLS usually contains additional bindings for special
2527 :     registers on the architecture, such as:
2528 :    
2529 :     val r0 : cell (* register zero *)
2530 :     val y : cell (* Y register *)
2531 :     val psr : cell (* processor status register *)
2532 :     ...
2533 :    
2534 :     The structure returned by applying the Cells functor is opened
2535 :     in this interface.
2536 :    
2537 :     The main implication of all this is that the datatypes for cells is
2538 :     split between CellsBasis and CELLS -- a fairly simple change for user
2539 :     code.
2540 :    
2541 :     In the old scheme the CELLS interface had a definitional binding of
2542 :     the form:
2543 :    
2544 :     signature CELLS = sig
2545 :    
2546 :     structure CellsBasis = CellsBasis
2547 :    
2548 :     ...
2549 :    
2550 :     end
2551 :    
2552 :     With all the sharing constraints that goes on in MLRISC, this old
2553 :     design quickly leads to errors such as:
2554 :    
2555 :     "structure definition spec inside of sharing ... "
2556 :    
2557 :    
2558 :     and appears to require an unacceptable amount of sharing and where
2559 :     constraint hackery.
2560 :    
2561 :     I think this error message (the interaction of definitional specs and
2562 :     sharing) requires more explanation on our web page.
2563 :    
2564 :     ----------------------------------------------------------------------
2565 : blume 826 Name: Matthias Blume
2566 : blume 879 Date: 2001/07/19 15:00:00 EDT
2567 :     Tag: blume-20010719-libreorg
2568 :     Description:
2569 :    
2570 :     This update puts together a fairly extensive but straightforward change
2571 :     to the way the libraries that implement the interactive system are
2572 :     organized:
2573 :    
2574 :     The biggest change is the elimination of structure Compiler. As a
2575 :     replacement for this structure, there is now a CM library
2576 :     (known as $smlnj/compiler.cm or $smlnj/compiler/current.cm)
2577 :     that exports all the substructures of the original structure Compiler
2578 :     directly. So instead of saying Compiler.Foo.bar one now simply
2579 :     says Foo.bar. (The CM libraries actually export a collection of
2580 :     structures that is richer than the collection of substructures of
2581 :     structure Compiler.)
2582 :    
2583 :     To make the transition smooth, there is a separate library called
2584 :     $smlnj/compiler/compiler.cm which puts together and exports the
2585 :     original structure Compiler (or at least something very close to it).
2586 :    
2587 :     There are five members of the original structure Compiler
2588 :     that are not exported directly but which instead became members
2589 :     of a new structure Backend (described by signature BACKEND). These are:
2590 :     structure Profile (: PROFILE), structure Compile (: COMPILE), structure
2591 :     Interact (: INTERACT), structure Machine (: MACHINE), and val
2592 :     architecture (: string).
2593 :    
2594 :     Structure Compiler.Version has become structure CompilerVersion.
2595 :    
2596 :     Cross-compilers for alpha32, hppa, ppc, sparc, and x86 are provided
2597 :     by $smlnj/compiler/<arch>.cm where <arch> is alpha32, hppa, ppc, sparc,
2598 :     or x86, respectively.
2599 :     Each of these exports the same frontend structures that
2600 :     $smlnj/compiler.cm exports. But they do not have a structure Backend
2601 :     and instead export some structure <Arch>Backend where <Arch> is Alpha32,
2602 :     Hppa, PPC, Sparc, or X86, respectively.
2603 :    
2604 :     Library $smlnj/compiler/all.cm exports the union of the exports of
2605 :     $smlnj/compiler/<arch>.cm
2606 :    
2607 :     There are no structures <Arch>Compiler anymore, use
2608 :     $smlnj/compiler/<arch>.cm instead.
2609 :    
2610 :     Library host-compiler-0.cm is gone. Instead, the internal library
2611 :     that instantiates CM is now called cm0.cm. Selection of the host
2612 :     compiler (backend) is no longer done here but. (Responsibility for it
2613 :     now lies with $smlnj/compiler/current.cm. This seems to be more
2614 :     logical.)
2615 :    
2616 :     Many individual files have been moved or renamed. Some files have
2617 :     been split into multiple files, and some "dead" files have been deleted.
2618 :    
2619 :     Aside from these changes to library organization, there are also changes
2620 :     to the way the code itself is organized:
2621 :    
2622 :     Structure Binfile has been re-implemented in such a way that it no
2623 :     longer needs any knowledge of the compiler. It exclusively deals
2624 :     with the details of binfile layout. It no longer invokes the
2625 :     compiler (for the purpose of creating new prospective binfile
2626 :     content), and it no longer has any knowledge of how to interpret
2627 :     pickles.
2628 :    
2629 :     Structure Compile (: COMPILE) has been stripped down to the bare
2630 :     essentials of compilation. It no longer deals with linking/execution.
2631 :     The interface has been cleaned up considerably.
2632 :    
2633 :     Utility routines for dealing with linking and execution have been
2634 :     moved into their own substructures.
2635 :    
2636 :     (The ultimate goal of these changes is to provide a light-weight
2637 :     binfile loader/linker (at least for, e.g., stable libraries) that
2638 :     does not require CM or the compiler to be present.)
2639 :    
2640 :     CM documentation has been updated to reflect the changes to library
2641 :     organization.
2642 :    
2643 :     ----------------------------------------------------------------------
2644 :     Name: Matthias Blume
2645 : blume 875 Date: 2001/07/10 17:30:00 EDT
2646 :     Tag: Release_110_34
2647 :     Description:
2648 :    
2649 : blume 876 Minor tweak to 110.34 (re-tagged):
2650 :    
2651 :     - README.html file added to CVS repository
2652 :     - runtime compiles properly under FreeBSD 3.X and 4.X
2653 :    
2654 :     ----------------------------------------------------------------------
2655 :     Name: Matthias Blume
2656 :     Date: 2001/07/10 17:30:00 EDT
2657 :     Tag: Release_110_34
2658 :     Description:
2659 :    
2660 : blume 875 New version number (110.34). New bootfiles.
2661 :    
2662 :     ----------------------------------------------------------------------
2663 :     Name: Matthias Blume
2664 : blume 874 Date: 2001/07/09 16:00:00 EDT
2665 :     Tag: blume-20010709-more-varargs
2666 :     Description:
2667 :    
2668 :     I changed the handling of varargs in ml-nlffigen again:
2669 :     The ellipsis ... will now simply be ignored (with an accompanying warning).
2670 :    
2671 :     The immediate effect is that you can actually call a varargs function
2672 :     from ML -- but you can't actually supply any arguments beyond the ones
2673 :     specified explicitly. (For example, you can call printf with its format
2674 :     string, but you cannot pass additional arguments.)
2675 :    
2676 :     This behavior is only marginally more useful than the one before, but
2677 :     it has the advantage that a function or, more importantly, a function
2678 :     type never gets dropped on the floor, thus avoiding follow-up problems with
2679 :     other types that refer to the offending one.
2680 :    
2681 :     ----------------------------------------------------------------------
2682 :     Name: Matthias Blume
2683 : blume 873 Date: 2001/07/09 11:25:00 EDT
2684 :     Tag: blume-20010709-varargs
2685 :     Description:
2686 :    
2687 :     1. ckit-lib.cm now exports structure Error
2688 :     2. ml-nlffigen reports occurences of "..." (i.e., varargs function types)
2689 :     with a warning accompanied by a source location. Moreover, it
2690 :     merely skips the offending function or type and proceeds with the
2691 :     rest of its work.u As a result, one can safely feed C code containing
2692 :     "..." to ml-nlffigen.
2693 :     3. There are some internal improvements to CM, providing slightly
2694 :     more general string substitutions in the tools subsystem.
2695 :    
2696 :     ----------------------------------------------------------------------
2697 :     Name: Matthias Blume
2698 : blume 854 Date: 2001/06/27 15:10:00 EDT
2699 :     Tag: blume-20010627-concur
2700 :     Description:
2701 :    
2702 :     Fixed a small bug in CM's handling of parallel compilation.
2703 :     (You could observe the bug by Control-C-interrupting an ordinary
2704 :     CMB.make or CM.stabilize and then attaching some compile servers.
2705 :     The result was that all of a sudden the previously interrupted
2706 :     compilation would continue on its own. This was because of
2707 :     an over-optimization: CM did not bother to clean out certain queues
2708 :     when no servers were attached "anyway", resulting in the contents
2709 :     of these queues to grab control when new servers did get attached.)
2710 :    
2711 :     There is also another minor update to the CM manual.
2712 :    
2713 :     ----------------------------------------------------------------------
2714 :     Name: Matthias Blume
2715 : blume 853 Date: 2001/06/26 16:15:00 EDT
2716 :     Tag: blume-20010626-cmdoc
2717 :     Description:
2718 :    
2719 :     Minor typo fixed in CM manual (syntax diagram for libraries).
2720 :    
2721 :     ----------------------------------------------------------------------
2722 :     Name: Matthias Blume
2723 : blume 852 Date: 2001/06/25 22:55:00 EDT
2724 :     Tag: blume-20010625-x86pc
2725 :     Description:
2726 :    
2727 :     Fixed a nasty bug in the X86 assembly code that caused signal
2728 :     handlers to fail (crash) randomly.
2729 :    
2730 :     ----------------------------------------------------------------------
2731 :     Name: Matthias Blume
2732 : blume 846 Date: 2001/06/25 12:05:00 EDT
2733 :     Tag: blume-20010625-nlffigen
2734 :     Description:
2735 :    
2736 :     This update fixes a number of minor bugs in ml-nlffigen as reported by
2737 :     Nick Carter <nbc@andrew.cmu.edu>.
2738 :    
2739 :     1. Silly but ok typedefs of the form "typedef void myvoid;" are now accepted.
2740 :     2. Default names for generated files are now derived from the name of
2741 :     the C file *without its directory*. In particular, this causes generated
2742 :     files to be placed locally even if the C file is in some system directory.
2743 :     3. Default names for generated signatures and structures are also derived
2744 :     from the C file name without its directory. This avoids silly things
2745 :     like "structure GL/GL".
2746 :     (Other silly names are still possible because ml-nlffigen does not do
2747 : blume 847 a thorough check of whether generated names are legal ML identifiers.
2748 :     When in doubt, use command line arguments to force particular names.)
2749 : blume 846
2750 :     ----------------------------------------------------------------------
2751 :     Name: Matthias Blume
2752 : blume 845 Date: 2001/06/21 12:25:00 EDT
2753 :     Tag: blume-20010621-eXene
2754 :     Description:
2755 :    
2756 :     eXene now compiles and (sort of) works again.
2757 :    
2758 :     The library name (for version > 110.33) is $/eXene.cm.
2759 :    
2760 :     I also added an new example in src/eXene/examples/nbody. See the
2761 :     README file there for details.
2762 :    
2763 :     ----------------------------------------------------------------------
2764 :     Name: Matthias Blume
2765 : blume 844 Date: 2001/06/20 16:40:00 EDT
2766 :     Tag: blume-20010620-cml
2767 :     Description:
2768 :    
2769 :     CML now compiles and works again.
2770 :    
2771 :     Libraries (for version > 110.33):
2772 :    
2773 :     $cml/cml.cm Main CML library.
2774 :     $cml/basis.cm CML's version of $/basis.cm.
2775 :     $cml/cml-internal.cm Internal helper library.
2776 :     $cml/core-cml.cm Internal helper library.
2777 :     $cml-lib/trace-cml.cm Tracing facility.
2778 :     $cml-lib/smlnj-lib.cm CML's version of $/smlnj-lib.cm
2779 :    
2780 :     The installer (config/install.sh) has been taught how to properly
2781 :     install this stuff.
2782 :    
2783 :     ----------------------------------------------------------------------
2784 :     Name: Matthias Blume
2785 : blume 843 Date: 2001/06/19 17:55:00 EDT
2786 :     Tag: blume-20010619-instantiate
2787 :     Description:
2788 :    
2789 :     This un-breaks the fix for bug 1432.
2790 :     (The bug was originally fixed in 110.9 but I broke it again some
2791 :     time after that.)
2792 :    
2793 :     ----------------------------------------------------------------------
2794 :     Name: Matthias Blume
2795 : blume 842 Date: 2001/06/19 17:25:00 EDT
2796 :     Tag: blume-20010619-signals
2797 :     Description:
2798 :    
2799 :     This should (hopefully) fix the long-standing signal handling bug.
2800 :     (The runtime system was constructing a continuation record with an
2801 :     incorrect descriptor which would cause the GC to drop data on the floor...)
2802 :    
2803 :     ----------------------------------------------------------------------
2804 :     Name: Matthias Blume
2805 : blume 840 Date: 2001/06/15 15:05:00 EDT
2806 : blume 841 Tag: blume-20010615-moresparc
2807 :     Description:
2808 :    
2809 :     Here is a short late-hour update related to Sparc c-calls:
2810 :    
2811 :     -- made handling of double-word arguments a bit smarter
2812 :    
2813 :     -- instruction selection phase tries to collapse certain clumsily
2814 :     constructed ML-Trees; typical example:
2815 :    
2816 :     ADD(ty,ADD(_,e,LI d1),LI d2) -> ADD(ty,e,LI(d1+d2))
2817 :    
2818 :     This currently has no further impact on SML/NJ since mlriscGen does
2819 :     not seem to generate such patterns in the first place, and c-calls
2820 :     (which did generate them in the beginning) has meanwhile been fixed
2821 :     so as to avoid them as well.
2822 :    
2823 :     ----------------------------------------------------------------------
2824 :     Name: Matthias Blume
2825 :     Date: 2001/06/15 15:05:00 EDT
2826 : blume 840 Tag: blume-20010615-sparc
2827 :     Description:
2828 :    
2829 :     The purpose of this update is to provide an implementation of NLFFI
2830 :     on Sparc machines.
2831 :    
2832 :     Here are the changes in detail:
2833 :    
2834 :     * src/MLRISC/sparc/c-calls/sparc-c-calls.sml is a new file containing
2835 :     the Sparc implementation of the c-calls API.
2836 :     * The Sparc backend of SML/NJ has been modified to uniformely use %fp
2837 :     for accessing the ML frame. Thus, we have a real frame pointer and
2838 :     can freely modify %sp without need for an omit-frame-ptr phase.
2839 :     The vfp logic in src/compiler/CodeGen/* has been changed to accomodate
2840 :     this case.
2841 :     * ml-nlffigen has been taught to produce code for different architectures
2842 :     and calling conventions.
2843 :     * In a way similar to what was done in the x86 case, the Sparc
2844 :     backend uses its own specific extension to mltree. (For example,
2845 :     it needs to be able to generate UNIMP instructions which are part
2846 :     of the calling convention.)
2847 :     * ml-nlffi-lib was reorganized to make it more modular (in particular,
2848 :     to make it easier to plug in new machine- and os-dependent parts).
2849 :    
2850 :     There are some other fairly unrelated bug fixes and cleanups as well:
2851 :    
2852 :     * I further hacked the .cm files for MLRISC tools (like MDLGen) so
2853 :     that they properly share their libraries with existing SML/NJ libraries.
2854 :     * I fixed a minor cosmetic bug in CM, supressing certain spurious
2855 :     follow-up error messages.
2856 :     * Updates to CM/CMB documentation.
2857 :    
2858 :     TODO items:
2859 :    
2860 :     * MLRISC should use a different register as its asmTemp on the Sparc.
2861 :     (The current %o2 is a really bad choice because it is part of the
2862 :     calling conventions, so things might interfere in unexpected ways.)
2863 :    
2864 :     ----------------------------------------------------------------------
2865 :     Name: Matthias Blume
2866 : blume 839 Date: 2001/06/07
2867 :     Tag: blume-20010607-calls
2868 :     Description:
2869 :    
2870 :     A number of internal changes related to C calls and calling conventions:
2871 :    
2872 :     1. ML-Tree CALL statements now carry a "pops" field. It indicates the
2873 :     number of bytes popped implicitly (by the callee). In most cases
2874 :     this field is 0 but on x86/win32 it is some non-zero value. This
2875 :     is information provided for the benefit of the "omit-frameptr" pass.
2876 :     2. The CALL instruction on the x86 carries a similar "pops" field.
2877 :     The instruction selection phase copies its value from the ML-Tree
2878 :     CALL statement.
2879 :     3. On all other architectures, the instruction selection phase checks
2880 :     whether "pops=0" and complains if not.
2881 :     4. The c-calls implementation for x86 now accepts two calling conventions:
2882 :     "ccall" and "stdcall". When "ccall" is selected, the caller cleans
2883 :     up after the call and pops is set to 0. For "stdcall", the caller
2884 :     does nothing, leaving the cleanup to the callee; pops is set to
2885 :     the number of bytes that were pushed onto the stack.
2886 :     5. The cproto decoder (compiler/Semant/types/cproto.sml) now can
2887 :     distinguish between "ccall" and "stdcall".
2888 :     6. The UNIMP instruction has been added to the supported Sparc instruction
2889 :     set. (This is needed for implementing the official C calling convention
2890 :     on this architecture.)
2891 :     7. I fixed some of the .cm files under src/MLRISC/Tools to make them
2892 :     work with the latest CM.
2893 :    
2894 :     ----------------------------------------------------------------------
2895 :     Name: Matthias Blume
2896 : blume 838 Date: 2001/06/05 15:10:00 EDT
2897 :     Tag: blume-20010605-cm-index
2898 :     Description:
2899 :    
2900 :     0. The "lambdasplit" parameter for class "sml" in CM has been documented.
2901 :    
2902 :     1. CM can now generate "index files". These are human-readable files
2903 :     that list on a per-.cm-file basis each toplevel symbol defined or
2904 :     imported. The location of the index file for
2905 :     <p>/<d>.cm is <p>/CM/INDEX/<d>.cm.
2906 :     To enable index-file generation, set CM.Control.generate_index to true
2907 :     or export an environment-symbol: export CM_GENERATE_INDEX=true.
2908 :    
2909 :     The CM manual has been updated accordingly.
2910 :    
2911 :     2. I made some slight modifications to the c-calls API in MLRISC.
2912 :    
2913 :     a) There is now a callback to support saving/restoring of
2914 :     dedicated but caller-save registers around the actual call
2915 :     instruction.
2916 :     b) One can optionally specify a comment-annotation for the
2917 :     call instruction.
2918 :    
2919 :     3. SML/NJ (mlriscGen.sml) uses this new API for the rawccall primop.
2920 :     (For example, the comment annotation shows the C prototype of
2921 :     the function being called.)
2922 :    
2923 :     ----------------------------------------------------------------------
2924 :     Name: Matthias Blume
2925 : blume 837 Date: 2001/06/01 13:30:00 EDT
2926 :     Tag: blume-20010601-nlffi-cleanup
2927 :     Description:
2928 :    
2929 :     This is mostly a cleanup of MLFFI stuff:
2930 :    
2931 :     - some signature files have been put into a more exposed place
2932 :     - the ugly 'f type parameter is gone (simplifies types tremendously!)
2933 :     - ml-nlffigen changed accordingly
2934 :     - tutorial updated
2935 :    
2936 :     Other changes:
2937 :    
2938 :     - author's affiliation in CM manual(s) updated
2939 :     - some more recognized keywords added to Allen's sml.sty
2940 :    
2941 :     ----------------------------------------------------------------------
2942 :     Name: Matthias Blume
2943 : blume 836 Date: 2001/05/25 15:30:00 EDT
2944 :     Tag: blume-20010525-iptr
2945 :     Description:
2946 :    
2947 :     - put the official 110.33-README (as it appears on the ftp server) under
2948 :     CVS
2949 :     - fixed a small bug related to incomplete pointer types in
2950 :     ml-nlffigen
2951 :     - small cosmetic change to the ml-nlffi-lib's "arr" type constructor
2952 :     (it does not need the 'f type parameter)
2953 :    
2954 :     ----------------------------------------------------------------------
2955 :     Name: Matthias Blume
2956 : blume 834 Date: 2001/05/23 14:30:00 EDT
2957 :     Tag: Release_110_33
2958 :     Description:
2959 :    
2960 :     New version number (110.33). New bootfiles.
2961 :    
2962 :     ----------------------------------------------------------------------
2963 :     Name: Matthias Blume
2964 : blume 833 Date: 2001/05/22 18:06:00 EDT
2965 :     Tag: blume-20010522-targets
2966 :     Description:
2967 :    
2968 :     Made install.sh use file config/targets.customized if it exists, falling
2969 :     back to config/targets if it doesn't. This way one can have a customized
2970 :     version of the targets file without touching the "real thing", thus
2971 :     eliminating the constant fear of accidentally checking something bogus
2972 :     back into the CVS repository... (File config/targets.customized must
2973 :     not be added to the repository!)
2974 :    
2975 :     ----------------------------------------------------------------------
2976 :     Name: Matthias Blume
2977 : blume 832 Date: 2001/05/22 16:30:00 EDT
2978 :     Tag: blume-20010522-minitut
2979 :     Description:
2980 :    
2981 :     1. Bug fix in ml-nlffigen; now (hopefully) correctly handling
2982 :     struct returns.
2983 :     2. Added src/ml-nlffi-lib/Doc/mini-tutorial.txt. This is some very
2984 :     incomplete, preliminary documentation for NLFFI.
2985 :    
2986 :     ----------------------------------------------------------------------
2987 :     Name: Matthias Blume
2988 : blume 829 Date: 2001/05/14 11:30:00 EDT
2989 :     Tag: blume-20010514-script
2990 :     Description:
2991 :    
2992 :     Some bugs in install script fixed.
2993 :    
2994 :     In addition to that I also made a slight change to the NLFFI API:
2995 :     Functors generated by ml-nlffigen now take the dynamic library as a
2996 :     straight functor argument, not as a suspended one. (The original
2997 :     functor code used to force the suspension right away anyway, so there
2998 :     was nothing gained by this complication of the interface.)
2999 :    
3000 :     ----------------------------------------------------------------------
3001 :     Name: Matthias Blume
3002 : blume 828 Date: 2001/05/11 14:35:00 EDT
3003 :     Tag: blume-20010511-ml-nlffi
3004 :     Description:
3005 :    
3006 :     I finally took the plunge and added my new FFI code to the main
3007 :     repository. For x86-linux it is now ready for prime-time.
3008 :    
3009 :     There are two new subdirectories of "src":
3010 :    
3011 :     - ml-nlffi-lib:
3012 :     The utility library for programs using the FFI interface.
3013 :     Here is the implementation of $/c.cm and its associated low-level
3014 :     partners $/c-int.cm and $/memory.cm.
3015 :     - ml-nlffigen:
3016 :     A stand-alone program for generating ML glue code from C source
3017 :     code.
3018 :    
3019 :     Building ml-nlffigen requires $/ckit-lib.cm.
3020 :    
3021 :     The config/install.sh script has been updates to do the Right Thing
3022 :     (hopefully).
3023 :    
3024 :     Notice that the source tree for the C-Kit will not be put under "src"
3025 :     but directly under the installation root directory. (This is the
3026 :     structure that currently exists on the CVS server when you check out
3027 :     module "sml".) Fortunately, config/install.sh knows about this oddity.
3028 :    
3029 :     Bugs: No documentation yet.
3030 :    
3031 :     ----------------------------------------------------------------------
3032 :     Name: Matthias Blume
3033 : blume 826 Date: 2001/05/09 16:35:00 EDT
3034 :     Tag: blume-20010509-cpscontract
3035 :     Description:
3036 :    
3037 :     Fixed a bug in the accounting code in cpsopt/contract.sml. (The
3038 :     wrapper/unwrapper elimination did not decrement usage counts and some
3039 :     dead variables got overlooked by the dead-up logic.)
3040 :    
3041 :     ----------------------------------------------------------------------
3042 : george 824 Name: Lal George
3043 :     Date: 2001/05/08 17:26:09 EDT
3044 :     Tag: george-20010508-omit-frameptr
3045 :     Description:
3046 :    
3047 :     Changes to implement the omit-frame-pointer optimization to support
3048 :     raw C calls. For now, there is only support on the Intel x86, but
3049 :     other architectures will follow as more experience is gained with this.
3050 :    
3051 :    
3052 :     ----------------------------------------------------------------------
3053 : blume 818 Name: Matthias Blume
3054 : blume 819 Date: 2001/05/07 14:40:00 EDT
3055 :     Tag: blume-20010507-proxies
3056 :     Description:
3057 :    
3058 :     I made into "proxy libraries" all libraries that qualify for such a
3059 :     change. (A qualifying library is a library that has another library or
3060 :     groups as its sole member and repeats that member's export list
3061 :     verbatim. A proxy library avoids this repetition by omitting its export
3062 :     list, effectively inheriting the list that its (only) member exports.
3063 :     See the CM manual for more explanation.)
3064 :     The main effect is that explicit export lists for these libraries
3065 :     do not have to be kepts in sync, making maintenance a bit easier.
3066 :    
3067 :     I also added copyright notices to many .cm-files.
3068 :    
3069 :     Last but not least, I made a new set of bootfiles.
3070 :    
3071 :     ----------------------------------------------------------------------
3072 :     Name: Matthias Blume
3073 : blume 818 Date: 2001/05/04 17:00:00 EDT
3074 :     Tag: blume-20010504-cm-lsplit
3075 :     Description:
3076 :    
3077 :     0. John merged pending changes to $/smlnj-lib.cm
3078 :    
3079 :     1. Allen's previous change accidentally backed out of one of Lal's
3080 :     earlier changes. I undid this mistake (re-introducing Lal's change).
3081 :    
3082 :     2. I used the new topOrder' function from graph-scc.sml (from $/smlnj-lib.cm)
3083 :     within the compiler where applicable. There is some code simplification
3084 :     because of that.
3085 :    
3086 :     3. The "split" phase (in FLINT) is now part of the default list of phases.
3087 :     Compiler.Control.LambdaSplitting.* can be used to globally control the
3088 :     lambda-splitting (cross-module-inlining) engine. In addition to that,
3089 :     it can now also be controlled on a per-source basis: CM has been taught
3090 :     a new tool parameter applicable to ML source files.
3091 :    
3092 :     - To turn lambda-splitting off completely:
3093 :     local open Compiler.Control.LambdaSplitting in
3094 :     val _ = set Off
3095 :     end
3096 :     - To make "no lambda-splitting" the global default (but allow per-source
3097 :     overriding); this is the initial setting:
3098 :     local open Compiler.Control.LambdaSplitting in
3099 :     val _ = set (Default NONE)
3100 :     end
3101 :     - To make "lambda-splitting with aggressiveness a" the global default
3102 :     (and allow per-source overriding):
3103 :     local open Compiler.Control.LambdaSplitting in
3104 :     val _ = set (Default (SOME a))
3105 :     end
3106 :    
3107 :     - To turn lambda-splitting off for a given ML souce file (say: a.sml)
3108 :     write (in the respective .cm-file):
3109 :     a.sml (lambdasplitting:off)
3110 :     - To turn lambda-splitting for a.sml on with minimal aggressiveness:
3111 :     a.sml (lambdasplitting:on)
3112 :     - To turn lambda-splitting for a.sml on with aggressiveness <a> (where
3113 :     <a> is a decimal non-negative integer):
3114 :     a.sml (lambdasplitting:<a>)
3115 :     - To turn lambda-splitting for a.sml on with maximal aggressiveness:
3116 :     a.sml (lambdasplitting:infinity)
3117 :     - To use the global default for a.sml:
3118 :     a.sml (lambdasplitting:default)
3119 :     or simply
3120 :     a.sml
3121 :    
3122 :     ----------------------------------------------------------------------
3123 : leunga 815 Name: Allen Leung
3124 :     Date: 2001/05/04 01:57:00 EDT
3125 :     Tag: leunga-20010504-sync
3126 :     Description:
3127 : leunga 788
3128 : leunga 815 MLRISC features.
3129 :    
3130 :     1. Fix to CMPXCHG instructions.
3131 :     2. Changed RA interface to allow annotations in callbacks.
3132 :     3. Added a new method to the stream interface to allow annotations updates.
3133 :    
3134 : george 749 ----------------------------------------------------------------------
3135 : blume 812 Name: Matthias Blume
3136 : blume 814 Date: 2001/05/01 11:45:00 EDT
3137 :     Tag: blume-20010501-pcedittmp
3138 :     Description:
3139 :    
3140 :     Changed install.sh to use the current working directory instead of
3141 :     /usr/tmp for a temporary file (pcedittmp). The previous choice
3142 :     of /usr/tmp caused trouble with MacOS X because of file premission
3143 :     problems.
3144 :    
3145 :     ----------------------------------------------------------------------
3146 :     Name: Matthias Blume
3147 : blume 812 Date: 2001/04/20 11:10:00 EDT
3148 :     Tag: blume-20010420-inMLflag
3149 :     Description:
3150 :    
3151 :     - added vp_limitPtrMask to vproc-state.h
3152 :     (for use by the raw-C-calls mechanism to implement proper interrupt
3153 :     handling)
3154 :     - made the ML compiler aware of various data-structure offsets so it
3155 :     can generate code for accessing the vp_inML flag and vp_limitPtrMask
3156 :     - tweaked mlriscGen.sml to have it emit interrupt-handling code for
3157 :     raw C-calls
3158 :    
3159 :     ----------------------------------------------------------------------
3160 : george 810 Name: Lal George
3161 :     Date: 2001/04/20 09:15:28 EDT
3162 :     Tag: george-20010420-macosX
3163 :     Description:
3164 :    
3165 :     - Changes to port to Mac OS X; Darwin.
3166 :    
3167 :     - In the process I found that sqrt was broken on the PPC, because the
3168 :     fsqrt instruction is not implemented.
3169 :    
3170 :     ----------------------------------------------------------------------
3171 : blume 799 Name: Matthias Blume
3172 : blume 808 Date: 2001/04/18 12:45:00 EDT
3173 :     Tag: blume-20010418-ccalls
3174 :     Description:
3175 :    
3176 :     - fixed two off-by-4 errors in the x86-specific c-calls implementation
3177 :     (this bug prevented structure arguments containing pointers from being
3178 :     passed correctly)
3179 :     - changed the raw-C-call code in mlriscGen.sml in such a way that
3180 :     structure arguments are represented as a pointer to the beginning
3181 :     of the structure (instead of having a series of synthesized arguments,
3182 :     one for each structure member)
3183 :    
3184 :     - made makeml script's verbosity level configurable via environment
3185 :     variable (MAKEML_VERBOSITY)
3186 :    
3187 : blume 813 - eliminated placeholder implementations for f32l, w16s, i16s, and f32s
3188 : blume 808 in rawmem-x86.sml; we are now using the real thing
3189 :    
3190 :     ----------------------------------------------------------------------
3191 :     Name: Matthias Blume
3192 : blume 806 Date: 2001/03/22 16:25:00 EST
3193 :     Tag: blume-20010322-bootfiles
3194 :     Description:
3195 :    
3196 :     Created a new set of bootfiles (for your automatic installation convenience).
3197 :    
3198 :     ----------------------------------------------------------------------
3199 :     Name: Matthias Blume
3200 : blume 805 Date: 2001/03/22 15:10:00 EST
3201 :     Tag: blume-20010322-rawmem-parcm
3202 :     Description:
3203 :    
3204 :     1. All "raw memory access" primitives for the new FFI are implemented now
3205 :     (at least on the x86).
3206 :     2. Some further cleanup of CM's parallel make mechanism.
3207 :    
3208 :     ----------------------------------------------------------------------
3209 :     Name: Matthias Blume
3210 : blume 801 Date: 2001/03/19 17:53:00 EST
3211 :     Tag: blume-20010319-parallel
3212 :     Description:
3213 :    
3214 :     Parallel make (using compile servers) now works again.
3215 :    
3216 :     To this end, CM.stabilize and CMB.make have been modified to work in
3217 :     two passes when compile servers are attached:
3218 :     1. Compile everything, do not perform stabilization; this pass
3219 :     uses compile servers
3220 :     2. Stabilize everything; this pass does not use compile servers
3221 :     If there are no compile servers, the two passes are combined into one
3222 :     (as before). Splitting the passes increases the inherent parallelism
3223 :     in the dependency graph because the entire graph including all
3224 :     libraries is available at the same time. This, in turn, improves
3225 :     server utilization. The downside is that the master process will
3226 :     have to do some extra work after compilation is done (because for
3227 :     technical reasons it must re-read all the binfiles during stabilization).
3228 :    
3229 :     ----------------------------------------------------------------------
3230 :     Name: Matthias Blume
3231 : blume 800 Date: 2001/03/16 12:22:00 EST
3232 :     Tag: blume-20010316-bootfiles
3233 :     Description:
3234 :    
3235 :     Created a new set of bootfiles (for your automatic installation convenience).
3236 :    
3237 :     ----------------------------------------------------------------------
3238 :     Name: Matthias Blume
3239 : blume 799 Date: 2001/03/16 11:00:00 EST
3240 :     Tag: blume-20010316-MLTREE-fixup
3241 :     Description:
3242 :    
3243 :     This is a minor fixup for an (untagged) earlier commit by Allen.
3244 :     (A file was missing).
3245 :    
3246 :     ----------------------------------------------------------------------
3247 : leunga 788 Name: Allen Leung
3248 : leunga 796 Date: Mon Mar 5 18:54:57 EST 2001
3249 :     Tag: leunga-20010305-cut-support
3250 :    
3251 :     1. New support for alternative control-flow in MLTREE.
3252 :     Currently we support
3253 :    
3254 :     FLOW_TO(CALL ...., [k1,...,kn])
3255 :    
3256 :     This is needed for 'cuts to' in C-- and try/handle-like constructs
3257 :     in Moby
3258 :    
3259 :     New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
3260 :    
3261 :     2. Register Allocator
3262 :    
3263 :     Changes in interface [from Fermin, John]
3264 :    
3265 :     3. Alpha 8-bit SLL support [Fermin]
3266 :    
3267 :     4. All architectures
3268 :    
3269 :     A new module (ClusterExpandCopies) for expanding parallel copies.
3270 :    
3271 :     ----------------------------------------------------------------------
3272 :     Name: Allen Leung
3273 : leunga 788 Date: 2001/02/27 23:07:00 EST
3274 :     Tag: leunga-20010227-minor-stuff
3275 :    
3276 :     1. Alpha bug fix for CMOVNE
3277 :     2. Handle mltree COND(..,FCMP ...,...)
3278 :     3. Bug fix in simplifier
3279 :    
3280 :     ----------------------------------------------------------------------
3281 : blume 777 Name: Matthias Blume
3282 : blume 781 Date: 2001/01/30 17:50:00 EST
3283 :     Tag: blume-20010130-sync
3284 :     Description:
3285 :    
3286 :     This is just a minor update to sync my devel branch with the main brach.
3287 :     The only visible change is the addition of some README files.
3288 :    
3289 :     ----------------------------------------------------------------------
3290 :     Name: Matthias Blume
3291 : blume 778 Date: 2001/01/12 23:30:00 JST
3292 :     Tag: blume-20010112-bootfiles
3293 :     Description:
3294 :    
3295 :     Made a new set of bootfiles that goes with the current state of the
3296 :     repository.
3297 :    
3298 :     ----------------------------------------------------------------------
3299 :     Name: Matthias Blume
3300 : blume 777 Date: 2001/01/12 21:20:00 JST
3301 :     Tag: blume-20010112-sync
3302 :     Description:
3303 :    
3304 :     I am just flushing out some minor changes that had accumulated in
3305 :     my private branch in order to sync with the main tree. (This is
3306 :     mainly because I had CVS trouble when trying to merge _into_ my
3307 :     private branch.)
3308 :    
3309 :     Most people should be completely unaffected by this.
3310 :    
3311 :     ----------------------------------------------------------------------
3312 : leunga 775 Name: Allen Leung
3313 :     Date: Thu Jan 11 21:03:00 EST 2001
3314 :     Tag: leunga-20010111-labexp=mltree
3315 :     Description:
3316 :    
3317 :     1. Removed the type LabelExp and replace it by MLTree.
3318 :     2. Rewritten mltree-simplify with the pattern matcher tool.
3319 :     3. There were some bugs in alpha code generator which would break
3320 :     64-bit code generation.
3321 :     4. Redo the tools to generate code with the
3322 :     5. The CM files in MLRISC (and in src/system/smlnj/MLRISC)
3323 :     are now generated by perl scripts.
3324 :    
3325 :     ----------------------------------------------------------------------
3326 : blume 771 Name: Matthias Blume
3327 : blume 774 Date: 2001/01/10 21:55:00 JST
3328 :     Tag: blume-20010110-rcc
3329 :     Description:
3330 :    
3331 :     The RCC stuff now seems to work (but only on the x86).
3332 :     This required hacking of the c-calls interface (and -implementation) in
3333 :     MLRISC.
3334 :    
3335 :     Normal compiler users should be unaffected.
3336 :    
3337 :     ----------------------------------------------------------------------
3338 :     Name: Matthias Blume
3339 : blume 773 Date: 2001/01/09 01:20:00 JST
3340 :     Tag: blume-20010109-rcc
3341 :     Description:
3342 :    
3343 :     This is a fairly big patch, flushing out a large number of pending
3344 :     changes that I made to my development copy over the last couple of days.
3345 :    
3346 :     Of practical relevance at this moment is a workaround for a pickling
3347 :     bug that Allen ran into the other day. The cause of the bug itself is
3348 :     still unknown and it might be hard to fix it properly, but the
3349 :     workaround has some merits of its own (namely somewhat reducing pickling
3350 :     overhead for certain libraries). Therefore, I think this solution should
3351 :     be satisfactory at this time.
3352 :    
3353 :     The rest of the changes (i.e., the vast majority) has to do with my
3354 :     ongoing efforts of providing direct support for C function calls from
3355 :     ML. At the moment there is a new primop "RAW_CCALL", typing magic
3356 :     in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
3357 :     case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
3358 :     translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
3359 :     handle RCC.
3360 :    
3361 :     The last part (the changes to mlriscGen.sml) are still known to be
3362 :     wrong on the x86 and not implemented on all other architectures. But
3363 :     the infrastructure is in place. I had to change a few functor
3364 :     signatures in the backend to be able to route the CCalls interface
3365 :     from MLRISC there, and I had to specialize the mltree type (on the
3366 :     x86) to include the necessary extensions. (The extensions themselves
3367 :     were already there and redy to go in MLRISC/x86).
3368 :    
3369 :     Everything should be very happy as soon as someone helps me with
3370 :     mlriscGen.sml...
3371 :    
3372 :     In any case, nothing of this should matter to anyone as long as the
3373 :     new primop is not being used (which is going to be the case unless you
3374 :     find it where I hid it :). The rest of the compiler is completely
3375 :     unaffected.
3376 :    
3377 :     ----------------------------------------------------------------------
3378 :     Name: Matthias Blume
3379 : blume 772 Date: 2001/01/05 00:30:00 JST
3380 :     Tag: blume-20010105-primops
3381 :     Description:
3382 :    
3383 :     Added some experimental support for work that I am doing right now.
3384 :     These changes mostly concern added primops, but there is also a new
3385 :     experimental C library in the runtime system (but currently not enabled
3386 :     anywhere except on Linux/X86).
3387 :    
3388 :     In the course of adding primops (and playing with them), I discovered that
3389 :     Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
3390 :     broken. (Zhong was very right he labeled this stuff as "major gross hack".)
3391 :     To recover, I made type information in INL_PRIM mandatory and changed
3392 :     prim.sml as well as built-in.sml accordingly. The InLine structure now
3393 :     has complete, correct type information (i.e., no bottom types).
3394 :    
3395 :     Since all these changes mean that we need new binfiles, I also bumped the
3396 :     version number to 110.32.1.
3397 :    
3398 :     ----------------------------------------------------------------------
3399 :     Name: Matthias Blume
3400 : blume 771 Date: 2000/12/30 22:10:00 JST
3401 :     Tag: blume-20001230-various
3402 :     Description:
3403 :    
3404 :     Added proxy libraries for MLRISC and let MLRISC libraries refer
3405 :     to each other using path anchors. (See CM manual for explanation.)
3406 :    
3407 :     Updated CM documentation.
3408 :    
3409 :     Fixed some bugs in CM.
3410 :    
3411 :     Implemented "proxy" libraries (= syntactic sugar for CM).
3412 :    
3413 :     Added "-quiet" option to makeml and changed runtime system accordingly.
3414 :    
3415 :     Added cleanup handler for exportML to reset timers and compiler stats.
3416 :    
3417 :     ----------------------------------------------------------------------
3418 : george 761 Name: Lal George
3419 :     Date: 2000/12/22 22:22:58 EST 2000
3420 :     Tag: Release_110_32
3421 :     Description:
3422 :    
3423 :     Infinite precision used throughout MLRISC.
3424 :     see MLRISC/mltree/machine-int.sig
3425 :    
3426 :     ----------------------------------------------------------------------
3427 : blume 756 Name: Matthias Blume
3428 : blume 760 Date: 2000/12/22 23:16:00 JST
3429 :     Tag: blume-20001222-warn
3430 :     Description:
3431 : blume 759
3432 : blume 760 Corrected wording and formatting of some CM warning message which I
3433 :     broke in my previous patch.
3434 : blume 759
3435 :     ----------------------------------------------------------------------
3436 :     Name: Matthias Blume
3437 : blume 763 Date: 2000/12/22 21:20:00 JST
3438 :     Tag: blume-20001222-anchorenv
3439 :     Description:
3440 :    
3441 :     Fixed CM's handling of anchor environments in connection with CMB.make.
3442 :    
3443 :     ----------------------------------------------------------------------
3444 :     Name: Matthias Blume
3445 : blume 758 Date: 2000/12/22 13:15:00 JST
3446 :     Tag: blume-20001222-cleanup
3447 :     Description:
3448 :    
3449 :     Removed src/cm/ffi which does not (and did not) belong here.
3450 :    
3451 :     ----------------------------------------------------------------------
3452 :     Name: Matthias Blume
3453 : blume 757 Date: 2000/12/21 23:55:00 JST
3454 :     Tag: blume-20001221-exn
3455 :     Description:
3456 :    
3457 :     Probably most important: CM no longer silently swallows all exceptions
3458 :     in the compiler.
3459 :     Plus: some other minor CM changes. For example, CM now reports some
3460 :     sizes for generated binfiles (code, data, envpickle, lambdapickle).
3461 :    
3462 :     ----------------------------------------------------------------------
3463 :     Name: Matthias Blume
3464 : blume 756 Date: 2000/12/15 00:01:05 JST
3465 :     Tag: blume-20001215-dirtool
3466 :     Description:
3467 :    
3468 :     - "dir" tool added.
3469 :     - improvements and cleanup to Tools structure
3470 :     - documentation updates
3471 :    
3472 :     ----------------------------------------------------------------------
3473 : leunga 752 Name: Allen Leung
3474 : leunga 755 Date: Thu Dec 14 03:45:24 EST 2000
3475 :     Description:
3476 :     Tag: leunga-20001214-int-inf
3477 :     Description:
3478 :    
3479 :     In IntInf, added these standard functions, which are missing from our
3480 :     implementation:
3481 :    
3482 :     andb : int * int -> int
3483 :     xorb : int * int -> int
3484 :     orb : int * int -> int
3485 :     notb : int -> int
3486 :     << : int * word -> int
3487 :     ~>> : int * word -> int
3488 :    
3489 :     Not tested, I hope they are correct.
3490 :    
3491 :     ----------------------------------------------------------------------
3492 :     Name: Allen Leung
3493 : leunga 752 Date: Fri Dec 8 19:23:26 EST 2000
3494 :     Description:
3495 :     Tag: leunga-20001208-nowhere
3496 :     Description:
3497 :    
3498 :     Slight improvements to the 'nowhere' tool to handle OR-patterns,
3499 :     to generate better error messages etc. Plus a brief manual.
3500 :    
3501 :     ----------------------------------------------------------------------
3502 : george 749 Name: Lal George
3503 :     Date: 2000/12/08 09:54:02 EST 2000
3504 :     Tag: Release_110_31
3505 :     Description:
3506 : george 705
3507 : george 749 - Version 110.31
3508 : leunga 591 ----------------------------------------------------------------------
3509 : leunga 744 Name: Allen Leung
3510 :     Date: Thu Dec 7 22:01:04 EST 2000
3511 :     Tag: leunga-20001207-cell-monster-hack
3512 :     Description:
3513 :    
3514 :     Major MLRISC internal changes. Affect all clients.
3515 :     Summary:
3516 :    
3517 :     1. Type CELLS.cell = int is now replaced by a datatype.
3518 :     As a result, the old regmap is now gone. Almost all interfaces
3519 :     in MLRISC change as a consequence.
3520 :    
3521 :     2. A new brand version of machine description tool (v3.0) that generates
3522 :     modules expecting the new interface. The old version is removed.
3523 :    
3524 :     3. The RA interface has been further abstracted into two new functors.
3525 :     RISC_RA and X86RA. These functors have much simpler interfaces.
3526 :     [See also directory MLRISC/demo.]
3527 :    
3528 :     4. Some other new source->source code generation tools are available:
3529 :    
3530 :     a. MLRISC/Tools/RewriteGen -- generate rewriters from rules.
3531 :     b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules.
3532 :     I use this tool to generate the peephole optimizers---with the new
3533 :     cell type changes, peephole rules are becoming difficult to write
3534 :     without conditional pattern matching.
3535 :    
3536 :     5. More Intmap -> IntHashTable change. Previous changes by Matthias didn't
3537 :     cover the entire MLRISC source tree so many things broke.
3538 :    
3539 :     6. CM files have been moved to the subdirectory MLRISC/cm.
3540 :     They are moved because there are a lot of them and they clutter up the
3541 :     root dir.
3542 :    
3543 :     7. More detailed documentation to come...
3544 :    
3545 :     NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do
3546 :     a makeml -rebuild first. This is because of other other
3547 :     changes that Matthias has made (see below).
3548 :    
3549 :    
3550 :     ----------------------------------------------------------------------
3551 : blume 733 Name: Matthias Blume
3552 : blume 742 Date: 2000/11/30 23:12:00 JST
3553 :     Tag: blume-20001130-filereorg
3554 :     Description:
3555 :    
3556 :     Some manual updates and some file reorganizations in CM.
3557 :    
3558 :     ----------------------------------------------------------------------
3559 :     Name: Matthias Blume
3560 : blume 737 Date: 2000/11/24 17:45:00 JST
3561 :     Tag: blume-20001124-link
3562 :     Description:
3563 :    
3564 :     Drastically improved link traversal code for the case that the dynamic
3565 :     value was already loaded at bootstrap time. As a result, CM and CMB
3566 :     now both load blazingly fast -- even on a very slow machine. Also,
3567 :     memory consumption has been further reduced by this.
3568 :    
3569 :     Warning: The format of the PIDMAP file has changed. THerefore, to
3570 :     bootstrap you have to do this:
3571 :    
3572 :     1. Run CMB.make
3573 :     2. Make a symbolic link for the boot directory:
3574 :     ln -s sml.boot.ARCH-OS xxx
3575 :     3. "Rebuild" the boot directory:
3576 :     ./makeml -boot xxx -rebuild sml ; rm xxx
3577 :     4. Boot normally:
3578 :     ./makeml
3579 :    
3580 :     ----------------------------------------------------------------------
3581 :     Name: Matthias Blume
3582 : blume 735 Date: 2000/11/21 21:20:00 JST
3583 :     Tag: blume-20001121-tools
3584 :     Description:
3585 :    
3586 :     Continued hacking on autoloading problem -- with success this time.
3587 :     Also changed tool-plugin mechanism. See new CM manual.
3588 :    
3589 :     ----------------------------------------------------------------------
3590 :     Name: Matthias Blume
3591 : blume 734 Date: 2000/11/19 14:30:00 JST
3592 :     Tag: blume-20001119-autoload
3593 :     Description:
3594 :    
3595 :     Some hacking to make autoloading faster. Success for CMB, no success
3596 :     so far for CM. There is a reduced structure CM' that autoloads faster.
3597 :     (This is a temporary, non-documented hack to be eliminated again when
3598 :     the general problem is solved.)
3599 :    
3600 :     ----------------------------------------------------------------------
3601 :     Name: Matthias Blume
3602 : blume 733 Date: 2000/11/17 14:10:00 JST
3603 :     Tag: blume-20001117-pickle-lib
3604 :     Description:
3605 :    
3606 :     1. Eliminated comp-lib.cm
3607 :     2. Made pickle-lib.cm
3608 :     3. Eliminated all uses of intset.sml (from comp-lib.cm)
3609 :     4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
3610 :     equivalent constructs from smlnj-lib.cm (INtHashTable).
3611 :     5. Point 4. also goes for those uses of intmap.* in MLRISC.
3612 :     Duplicated intmap modules thrown out.
3613 :     6. Hunted down all duplicated SCC code and replaced it with
3614 :     equivalent stuff (GraphSCCFn from smlnj-lib.cm).
3615 :     7. Rewrote Feedback module.
3616 :     8. Moved sortedlist.sml into viscomp-lib.cm. Eventually it
3617 :     should be thrown out and equivalent modules from smlnj-lib.cm
3618 :     should be used (IntRedBlackSet, IntListSet, ...).
3619 :    
3620 :     Confirmed that compiler compiles to fixpoint.
3621 :    
3622 :     ----------------------------------------------------------------------
3623 : leunga 731 Name: Allen Leung
3624 :     Date: 2000/11/10 18:00:00
3625 :     Tag: leunga-20001110-new-x86-fp
3626 :    
3627 :     A new x86 floating point code generator has been added.
3628 :     By default this is turned off. To turn this on, do:
3629 :    
3630 :     CM.autoload "$smlnj/compiler.cm";
3631 :     Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
3632 :    
3633 :     Changes:
3634 :    
3635 :     1. Changed FTAN to FPTAN so that the assembly output is correct.
3636 :     2. Changed the extension callback for FTANGENT to generate:
3637 :    
3638 :     fptan
3639 :     fstp %st(0)
3640 :     instead of
3641 :     fptan
3642 :     fstpl ftempmem
3643 :    
3644 :     3. Numerous assembly fixes for x86.
3645 :    
3646 :     5. Cleaned up the machine code output module x86/x86MC.sml and added
3647 :     support for a whole bunch of instructions and addressing modes:
3648 :    
3649 :     fadd/fsub/fsubr/fmul/fdiv/fdivr %st, %st(n)
3650 :     faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp %st, %st(n)
3651 :     fadd/fsub/fsubr/fmul/fdiv/fdivr %st(n), %st
3652 :     fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
3653 :     fxch %st(n)
3654 :     fld %st(n)
3655 :     fst %st(n)
3656 :     fst mem
3657 :     fstp %st(n)
3658 :     fucom %st(n)
3659 :     fucomp %st(n)
3660 :    
3661 :     All these are now generated when the fast fp mode is turned on.
3662 :    
3663 :     6. Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
3664 :    
3665 :     ----------------------------------------------------------------------
3666 : blume 729 Name: Matthias Blume
3667 :     Date: 2000/11/09 11:20:00 JST
3668 :     Tag: blume-20001109-scc
3669 :     Description:
3670 :    
3671 :     Eliminated some code duplication:
3672 :    
3673 :     1. Added "where" clause to GraphSCCFn in SML/NJ Library.
3674 :     (Otherwise the functor is useless.)
3675 :     2. Used GraphSCCFn where SCCUtilFun was used previously.
3676 :     3. Got rid of SCCUtilFun (in comp-lib.cm).
3677 :    
3678 :     ----------------------------------------------------------------------
3679 : george 721 Name: Lal George
3680 :     Date: 2000/11/06 09:02:21 EST 2000
3681 :     Tag: Release_110_30
3682 :     Description:
3683 :    
3684 :     - Version 110.30
3685 :     ----------------------------------------------------------------------
3686 : blume 715 Name: Matthias Blume
3687 : blume 716 Date: 2000/11/04 14:45:00
3688 :     Tag: blume-20001104-mlbuild
3689 :     Description:
3690 :    
3691 :     - Made ml-build faster on startup.
3692 :     - Documentation fixes.
3693 :    
3694 :     ----------------------------------------------------------------------
3695 :     Name: Matthias Blume
3696 : blume 715 Date: 2000/11/02 17:00:00 JST
3697 :     Tag: blume-20001102-condcomp
3698 :     Description:
3699 :    
3700 :     - Small tweaks to pickler -- new BOOTFILES!
3701 :     - Version bumped to 110.29.2.
3702 :     - Added conditional compilation facility to init.cmi (see comment there).
3703 :     ----------------------------------------------------------------------
3704 : leunga 713 Name: Allen Leung
3705 :     Date: 2000/10/23 19:31:00
3706 :     Tag: leunga-20001023-demo-ra
3707 :    
3708 :     1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
3709 :     2. Test programs for the graph library updated
3710 :     3. Some new MLRISC demo programs added
3711 :    
3712 :     ----------------------------------------------------------------------
3713 : blume 710 Name: Matthias Blume
3714 : blume 711 Date: 2000/08/31 22:15:00 JST
3715 :     Tag: blume-20001017-errmsg
3716 :     Description:
3717 :    
3718 :     More error message grief: Where there used to be no messages, there
3719 :     now were some that had bogus error regions. Fixed.
3720 :    
3721 :     ----------------------------------------------------------------------
3722 :     Name: Matthias Blume
3723 : blume 710 Date: 2000/08/31 17:30:00 JST
3724 :     Tag: blume-20001017-v110p29p1
3725 :     Description:
3726 :    
3727 :     I made a version 110.29.1 with new bootfiles.
3728 :    
3729 :     Changes: Modified pickler/unpickler for faster and leaner unpickling.
3730 :     CM documentation changes and a small bugfix in CM's error reporting.
3731 :    
3732 :     ----------------------------------------------------------------------
3733 : george 705 Name: Lal George
3734 :     Date: 2000/09/27 14:42:35 EDT
3735 :     Tag: george-20000927-nodestatus
3736 :     Description:
3737 :    
3738 :     Changed the type of the nodestatus, so that:
3739 :    
3740 :     SPILLED(~1) is now SPILLED
3741 :     SPILLED(m) where m>=0 is now MEMREG(m)
3742 :     SPILLED(s) where s<~1 is now SPILL_LOC(~s)
3743 :    
3744 :     ----------------------------------------------------------------------
3745 : blume 697 Name: Matthias Blume
3746 : blume 703 Date: 2000/09/07 14:45:00 JST
3747 :     Tag: blume-20000907-cmerrmsg
3748 :     Description:
3749 :    
3750 :     Small tweak to CM to avoid getting ML syntax error messages twice.
3751 :    
3752 :     ----------------------------------------------------------------------
3753 :     Name: Matthias Blume
3754 : blume 701 Date: 2000/08/31 18:00:00 JST
3755 :     Tag: blume-20000831-cvsbootfiles
3756 :     Description:
3757 :    
3758 :     New URL for boot files (because the 110.29 files on the BL server do
3759 :     now work correctly with my updated install scripts for yacc and lex).
3760 :    
3761 :     ----------------------------------------------------------------------
3762 :     Name: Matthias Blume
3763 : blume 697 Date: 2000/08/08 12:33:00 JST
3764 :     Tag: blume-20000808-manual
3765 :     Description:
3766 :    
3767 :     Tiny update to CM manual.
3768 :    
3769 :     ----------------------------------------------------------------------
3770 : leunga 695 Name: Allen Leung
3771 :     Date: 2000/08/7 19:31:00
3772 :     Tag: leunga-20000807-a-whole-bunch-of-stuff
3773 :    
3774 :     Moby, C--, SSA, x86, machine descriptions etc. Should only affect C--
3775 :     and Mobdy.
3776 :    
3777 :     1. x86
3778 :    
3779 :     a. Fixes to peephole module by John and Dan.
3780 :     b. Assembly fix to SETcc by Allen.
3781 :     c. Fix to c-call by John.
3782 :     d. Fix to spilling by John. (This one deals with the missing FSTPT case)
3783 :     e. Instruction selection optimization to SETcc as suggested by John.
3784 :    
3785 :     For example,
3786 :    
3787 :     MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
3788 :    
3789 :     should generate:
3790 :    
3791 :     MOVL a, x
3792 :     SUBL b, x
3793 :     SHRL 31, x
3794 :    
3795 :     2. IR stuff
3796 :    
3797 :     A bunch of new DJ-graph related algorithms added. These
3798 :     speed up SSA construction.
3799 :    
3800 :     3. SSA + Scheduling
3801 :    
3802 :     Added code for SSA and scheduling to the repository
3803 :    
3804 :     ----------------------------------------------------------------------
3805 : george 694 Name: Lal George
3806 :     Date: 2000/07/27 11:53:14 EDT
3807 :    
3808 :     Tag: lal-20000727-linux-ppc
3809 :     Description:
3810 :    
3811 :     Made changes to support Linux PPC.
3812 :     p.s. I have confirmation that the 110.29 boot files work fine.
3813 :    
3814 :     ----------------------------------------------------------------------
3815 : blume 684 Name: Matthias Blume
3816 : blume 692 Date: 2000/07/27 17:40:00 JST
3817 :     Tag: blume-20000727-scripts
3818 :     Description:
3819 :    
3820 :     !!!! WARNING !!!!
3821 :     You must recompile the runtime system!
3822 :     !!!! WARNING !!!!
3823 :    
3824 :     This is basically another round of script-enhancements:
3825 :    
3826 :     1. sml, ml-build, and ml-makedepend accept options -D and -U to define
3827 :     and undefine CM preprocessor symbols.
3828 :    
3829 :     2. ml-build avoids generating a new heap image if it finds that the
3830 :     existing one is still ok. (The condition is that no ML file had to
3831 :     be recompiled and all ML files are found to be older that the heap
3832 :     file.)
3833 :    
3834 :     To make this work smoothly, I also hacked the runtime system as
3835 :     well as SMLofNJ.SysInfo to get access to the heap image suffix
3836 :     (.sparc-solaris, ...) that is currently being used.
3837 :    
3838 :     Moreover, the signature of CM.mk_standalone has changed. See the
3839 :     CM manual.
3840 :    
3841 :     3. ml-makedepend accepts additional options -n, -a, and -o. (See the
3842 :     CM manual for details.)
3843 :    
3844 :     4. More CM manual updates:
3845 :     - all of the above has been documented.
3846 :     - there is now a section describing the (CM-related) command line
3847 :     arguments that are accepted by the "sml" command
3848 :    
3849 :     ----------------------------------------------------------------------
3850 :     Name: Matthias Blume
3851 : blume 691 Date: 2000/07/25 16:20:00 JST
3852 :     Tag: blume-20000725-makedepend
3853 :     Description:
3854 :    
3855 :     Added a script called ml-makedepend. This can be used in makefiles
3856 :     for Unix' make in a way very similar to the "makedepend" command for
3857 :     C.
3858 :    
3859 :     The script internally uses function CM.sources.
3860 :    
3861 :     Synopsis:
3862 :    
3863 :     ml-makedepend [-f makefile] cmfile targetname
3864 :    
3865 :     The default for the makefile is "makefile" (or "Makefile" should
3866 :     "makefile" not exist).
3867 :    
3868 :     ml-makedepend adds a cmfile/targetname-specific section to this
3869 :     makefile (after removing the previous version of this section). The
3870 :     section contains a single dependency specification with targetname on
3871 :     the LHS (targetname is an arbitrary name), and a list of files derived
3872 :     from the cmfile on the RHS. Some of the files on the RHS are
3873 :     ARCH/OPSYS-specific. Therefore, ml-makedepend inserts references to
3874 :     "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
3875 :     path names. The makefile writer is responsible for making sure that
3876 :     these variables have correct at the time "make" is invoked.
3877 :    
3878 :     ----------------------------------------------------------------------
3879 :     Name: Matthias Blume
3880 : blume 690 Date: 2000/07/22 23:30:00 JST
3881 :     Tag: blume-20000722-urlupdate
3882 :     Description:
3883 :    
3884 :     Changed BOOT and config/srcarchiveurl to point to BL server:
3885 :    
3886 :     ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
3887 :    
3888 :     ----------------------------------------------------------------------
3889 :     Name: Matthias Blume
3890 : blume 686 Date: 2000/07/18 18:00:00 JST
3891 :     Tag: blume-20000718-Version_110_29
3892 :     Description:
3893 :    
3894 :     1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
3895 :    
3896 :     2. Updated config/version to 110.29
3897 :    
3898 :     3. Updated config/srcarchiveurl
3899 :    
3900 :     3. New boot files!
3901 :     ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
3902 :    
3903 :     ----------------------------------------------------------------------
3904 :     Name: Matthias Blume
3905 : blume 684 Date: 2000/07/11 13:58:00 JST
3906 :     Tag: blume-20000711-doctypo
3907 :     Description:
3908 :    
3909 :     Fixed a few typos in CM manual.
3910 :    
3911 :     ----------------------------------------------------------------------
3912 : leunga 683 Name: Allen Leung
3913 :     Date: 2000/06/15 00:38:00
3914 :     Tag: leunga-20000704-sparc-x86
3915 :    
3916 :     1. x86 peephole improvement sp += k; sp -= k => nop [from John]
3917 :     2. fix to x86 RET bug [found by Dan Grossman]
3918 :     3. sparc assembly bug fix for ticc instructions [found by Fermin]
3919 :    
3920 :     Affects c-- and moby only
3921 :    
3922 :     ----------------------------------------------------------------------
3923 : blume 680 Name: Matthias Blume
3924 : blume 682 Date: 2000/07/04 15:26:00
3925 :     Tag: blume-20000704-trigger
3926 :     Description:
3927 :    
3928 :     1. Improvements to CM manual.
3929 :     2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
3930 :     of getting a back-trace. The function, when called, raises an
3931 :     internal exception which explicitly carries the full back-trace history,
3932 :     so it is unaffected by any intervening handle-raise pairs ("trivial"
3933 :     or not). The interactive loop will print that history once it arrives
3934 :     at top level.
3935 :     Short of having all exceptions implicitly carry the full history, the
3936 :     recommended way of using this facility is:
3937 :     - compile your program with instrumentation "on"
3938 :     - run it, when it raises an exception, look at the history
3939 :     - if the history is "cut off" because of some handler, go and modify
3940 :     your program so that it explicitly calls BTrace.trigger
3941 :     - recompile (still instrumented), and rerun; look at the full history
3942 :    
3943 :     ----------------------------------------------------------------------
3944 :     Name: Matthias Blume
3945 :     Date: 2000/07/03 15:36:00 JST
3946 : blume 680 Tag: blume-20000702-manual
3947 :     Description:
3948 :    
3949 :     Small corrections and updates to CM manual.
3950 :    
3951 :     ----------------------------------------------------------------------
3952 : blume 682 Name: Matthias Blume
3953 : blume 679 Date: 2000/06/29 16:04:00 JST
3954 :     Tag: blume-20000629-yacctool
3955 :     Description:
3956 :    
3957 :     Changes:
3958 :    
3959 :     1. Class "mlyacc" now takes separate arguments to pass options to
3960 :     generated .sml- and .sig-files independently.
3961 :     2. Corresponding CM manual updates.
3962 :     3. BTrace module now also reports call sites. (However, for loop clusters
3963 :     it only shows from where the cluster was entered.) There are associated
3964 :     modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
3965 :    
3966 :     ----------------------------------------------------------------------
3967 : blume 675 Name: Matthias Blume
3968 : blume 678 Date: 2000/06/27 16:51:00 JST
3969 :     Tag: blume-20000627-noweb
3970 :     Description:
3971 :    
3972 :     Changes:
3973 :    
3974 :     1. Implemented "subdir" and "witness" options for noweb tool.
3975 :     This caused some slight internal changes in CM's tool implementation.
3976 :     2. Fixed bug in "tool plugin" mechanism. This is essentially cleaning
3977 :     some remaining issues from earlier path anchor changes.
3978 :     3. Updated CM manual accordingly.
3979 :    
3980 :     4. Changed implementation of back-tracing so that I now consider it
3981 :     ready for prime-time.
3982 :    
3983 :     In particular, you don't have to explicitly trigger the back-trace
3984 :     anymore. Instead, if you are running BTrace-instrumented code and
3985 :     there is an uncaught exception (regardless of whether or not it was
3986 :     raised in instrumented code), the top-level evalloop will print
3987 :     the back-trace.
3988 :    
3989 :     Features:
3990 :    
3991 :     - Instrumented and uninstrumented code work together seemlessly.
3992 :     (Of course, uninstrumented code is never mentioned in actual
3993 :     back-traces.)
3994 :    
3995 :     - Asymptotic time- and space-complexity of instrumented code is
3996 :     equal to that of uninstrumented code. (This means that
3997 :     tail-recursion is preserved by the instrumentation phase.)
3998 :    
3999 :     - Modules whose code has been instrumented in different sessions
4000 :     work together without problem.
4001 :    
4002 :     - There is no penalty whatsoever on uninstrumented code.
4003 :    
4004 :     - There is no penalty on "raise" expressions, even in
4005 :     instrumented code.
4006 :    
4007 :     A potential bug (or perhaps it is a feature, too):
4008 :    
4009 :     A back-trace reaches no further than the outermost instrumented
4010 :     non-trivial "raise". Here, a "trivial" raise is one that is the
4011 :     sole RHS of a "handle" rule. Thus, back-traces reach trough
4012 :    
4013 :     <exp> handle e => raise e
4014 :    
4015 :     and even
4016 :    
4017 :     <exp> handle Foo => raise Bar
4018 :    
4019 :     and, of course, through
4020 :    
4021 :     <exp> handle Foo => ...
4022 :    
4023 :     if the exception was not Foo.
4024 :    
4025 :     Back-traces always reach right through any un-instrumented code
4026 :     including any of its "handle" expressions, trivial or not.
4027 :    
4028 :     To try this out, do the following:
4029 :    
4030 :     - Erase all existing binfiles for your program.
4031 :     (You may keep binfiles for those modules where you think you
4032 :     definitely don't need back-tracing.)
4033 :     - Turn on back-trace instrumentation:
4034 :     SMLofNJ.Internals.BTrace.mode (SOME true);
4035 :     - Recompile your program. (I.e., run "CM.make" or "use".)
4036 :     - You may now turn instrumentation off again (if you want):
4037 :     SMLofNJ.Internals.BTrace.mode (SOME false);
4038 :     - Run your program as usual. If it raises an exception that
4039 :     reaches the interactive toplevel, then a back-trace will
4040 :     automatically be printed. After that, the toplevel loop
4041 :     will print the exception history as usual.
4042 :    
4043 :     ----------------------------------------------------------------------
4044 :     Name: Matthias Blume
4045 : blume 677 Date: 2000/06/26 09:56:46 JST
4046 :     Tag: blume-20000626-setup
4047 :     Description:
4048 :    
4049 :     CM: - setup-parameter to "sml" added; this can be used to run arbitrary
4050 :     ML code before and after compiling a file (e.g., to set compiler
4051 :     flags)
4052 :    
4053 :     Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
4054 :     - associated changes to btrace.sml (BTrace instrumentation pass)
4055 :     - cleaner implementation of btimp.sml (BTrace tracing and report
4056 :     module)
4057 :    
4058 :     CM manual: * new path encoding documented
4059 :     * description of setup-parameter to "sml" added
4060 :    
4061 :     The biggest user-visible change to back-tracing is that it is no
4062 :     longer necessary to compile all traced modules within the same
4063 :     session. (This was a real limitation.)
4064 :    
4065 :     ----------------------------------------------------------------------
4066 :     Name: Matthias Blume
4067 : blume 676 Date: 2000/06/24 12:40:00 JST
4068 :     Tag: blume-20000624-startup
4069 :     Description:
4070 :    
4071 :     Fixes startup slowdown problem. (I was calling SrcPath.sync a _tad_
4072 :     bit too often -- to put it mildly. :)
4073 :    
4074 :     ----------------------------------------------------------------------
4075 :     Name: Matthias Blume
4076 : blume 675 Date: 2000/06/23 18:20:00 JST
4077 :     Tag: blume-20000623-btrace
4078 :     Description:
4079 :    
4080 :     This updates adds a backtrace facility to aid programmers in debugging
4081 :     their programs. This involves the following changes:
4082 :    
4083 :     1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
4084 :     keeping track of the current call stack. When programs are compiled
4085 :     in a special mode, the compiler will insert calls to these hooks
4086 :     into the user program.
4087 :     "Hook" means that it is possible for different implementations of
4088 :     back-tracing to register themselves (at different times).
4089 :    
4090 :     2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
4091 :     as an Absyn.dec->Absyn.dec rewrite. Normally this phase is turned off.
4092 :     It can be turned on using this call:
4093 :     SMLofNJ.Internals.BTrace.mode (SOME true);
4094 :     Turning it off again:
4095 :     SMLofNJ.Internals.BTrace.mode (SOME false);
4096 :     Querying the current status:
4097 :     SMLofNJ.Internals.BTrace.mode NONE;
4098 :     Annotated programs are about twice as big as normal ones, and they
4099 :     run a factor of 2 to 4 slower with a dummy back-trace plugin (one
4100 :     where all hooks do nothing). The slowdown with a plugin that is
4101 :     actually useful (such as the one supplied by default) is even greater,
4102 :     but in the case of the default plugin it is still only an constant
4103 :     factor (amortized).
4104 :    
4105 :     3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
4106 :     with a sub-structure BTrace for controlling back-tracing. In particular,
4107 :     the above-mentioned function "mode" controls whether the annotation
4108 :     phase is invoked by the compiler. Another important function is
4109 :     "trigger": when called it aborts the current execution and causes
4110 :     the top-level loop to print a full back-trace.
4111 :    
4112 :     4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
4113 :     for back-tracing. It keeps track of the dynamic call stack and in
4114 :     addition to that it keeps a partial history at each "level" of that
4115 :     stack. For example, if a tail-calls b, b tail-calls c, and c tail-calls
4116 :     d and b (at separate times, dynamically), then the report will show:
4117 :    
4118 :     GOTO d
4119 :     /c
4120 :     GOTO \b
4121 :     CALL a
4122 :    
4123 :     This shows that there was an initial non-tail call of a, then a
4124 :     tail-call to b or c, looping behavior in a cluster of functions that
4125 :     consist of b and c, and then a goto from that cluster (i.e., either from
4126 :     b or from c) to d.
4127 :    
4128 :     Note that (depending on the user program) the amount of information
4129 :     that the back-trace module has to keep track of at each level is bounded
4130 :     by a constant. Thus, the whole implementation has the same asymptotical
4131 :     complexity as the original program (both in space and in time).
4132 :    
4133 :     5. compiler/TopLevel/interact/evalloop.sml has been modified to
4134 :     handle the special exception SMLofNJ.Internals.BTrace.BTrace
4135 :     which is raised by the "trigger" function mentioned above.
4136 :    
4137 :     Notes on usage:
4138 :    
4139 :     - Annotated code works well together with unannotated code:
4140 :     Unannotated calls simply do not show up at all in the backtrace.
4141 :    
4142 :     - It is not a good idea to let modules that were annotated during
4143 :     different sessions run at the same time. This is because the compiler
4144 :     chooses small integers to identify individual functions, and there
4145 :     will be clashes if different modules were compiled in separate sessions.
4146 :     (Nothing will crash, and you will even be told about the clashes, but
4147 :     back-trace information will in general not be useful.)
4148 :    
4149 :     - Back-tracing can be confused by callcc and capture.
4150 :    
4151 :     - The only way of getting a back-trace right now is to explicitly
4152 :     invoke the "trigger" function from your user program. Eventually, we
4153 :     should make every exception carry back-trace information (if
4154 :     available). But since this creates more overhead at "raise"-time
4155 :     (similar to the current exnHistory overhead), I have not yet
4156 :     implemented this. (The implementation will be rather easy.) With
4157 :     exceptions carrying back-trace information, this facility will be even
4158 :     more useful because users don't need to modify their programs...
4159 :    
4160 :     - While it is possible to compile the compiler with back-trace
4161 :     annotations turned on (I did it to get some confidence in
4162 :     correctness), you must make absolutely sure that core.sml and
4163 :     btimp.sml are compiled WITHOUT annotation! (core.sml cannot actually
4164 :     be compiled with annotation because there is no core access yet, but
4165 :     if you compile btimp.sml with annotation, then the system will go into
4166 :     an infinite recursion and crash.)
4167 :     Since CM currently does not know about BTrace, the only way to turn
4168 :     annotations on and off for different modules of the compiler is to
4169 :     interrupt CMB.make, change the settings, and re-invoke it. Of course,
4170 :     this is awkward and clumsy.
4171 :    
4172 :     Sample sessions:
4173 :    
4174 :     Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
4175 :     - SMLofNJ.Internals.BTrace.mode (SOME true);
4176 :     [autoloading]
4177 :     [autoloading done]
4178 :     val it = false : bool
4179 :     - structure X = struct
4180 :     - fun main n = let
4181 :     - fun a (x, 0) = d x
4182 :     - | a (x, n) = b (x, n - 1)
4183 :     - and b (x, n) = c (x, n)
4184 :     - and c (x, n) = a (x, n)
4185 :     - and d x = e (x, 3)
4186 :     - and e (x, 0) = f x
4187 :     - | e (x, n) = e (x, n - 1)
4188 :     - and f 0 = SMLofNJ.Internals.BTrace.trigger ()
4189 :     - | f n = n * g (n - 1)
4190 :     - and g n = a (n, 3)
4191 :     - in
4192 :     - f n
4193 :     - end
4194 :     - end;
4195 :     structure X : sig val main : int -> int end
4196 :     - X.main 3;
4197 :     *** BACK-TRACE ***
4198 :     GOTO stdIn:4.2-13.20: X.main[2].f
4199 :     GOTO-( stdIn:4.2-13.20: X.main[2].e
4200 :     GOTO stdIn:4.2-13.20: X.main[2].d
4201 :     / stdIn:4.2-13.20: X.main[2].a
4202 :     | stdIn:4.2-13.20: X.main[2].b
4203 :     GOTO-\ stdIn:4.2-13.20: X.main[2].c
4204 :     CALL stdIn:4.2-13.20: X.main[2].g
4205 :     GOTO stdIn:4.2-13.20: X.main[2].f
4206 :     GOTO-( stdIn:4.2-13.20: X.main[2].e
4207 :     GOTO stdIn:4.2-13.20: X.main[2].d
4208 :     / stdIn:4.2-13.20: X.main[2].a
4209 :     | stdIn:4.2-13.20: X.main[2].b
4210 :     GOTO-\ stdIn:4.2-13.20: X.main[2].c
4211 :     CALL stdIn:4.2-13.20: X.main[2].g
4212 :     GOTO stdIn:4.2-13.20: X.main[2].f
4213 :     GOTO-( stdIn:4.2-13.20: X.main[2].e
4214 :     GOTO stdIn:4.2-13.20: X.main[2].d
4215 :     / stdIn:4.2-13.20: X.main[2].a
4216 :     | stdIn:4.2-13.20: X.main[2].b
4217 :     GOTO-\ stdIn:4.2-13.20: X.main[2].c
4218 :     CALL stdIn:4.2-13.20: X.main[2].g
4219 :     GOTO stdIn:4.2-13.20: X.main[2].f
4220 :     CALL stdIn:2.15-17.4: X.main[2]
4221 :     -
4222 :    
4223 :     (Note that because of a FLINt bug the above code currently does not
4224 :     compile without BTrace turned on.)
4225 :    
4226 :     Here is another example, using my modified Tiger compiler:
4227 :    
4228 :     Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
4229 :     - SMLofNJ.Internals.BTrace.mode (SOME true);
4230 :     [autoloading]
4231 :     [autoloading done]
4232 :     val it = false : bool
4233 :     - CM.make "sources.cm";
4234 :     [autoloading]
4235 :     ...
4236 :     [autoloading done]
4237 :     [scanning sources.cm]
4238 :     [parsing (sources.cm):parse.sml]
4239 :     [creating directory CM/SKEL ...]
4240 :     [parsing (sources.cm):tiger.lex.sml]
4241 :     ...
4242 :     [wrote CM/sparc-unix/semant.sml]
4243 :     [compiling (sources.cm):main.sml]
4244 :     [wrote CM/sparc-unix/main.sml]
4245 :     [New bindings added.]
4246 :     val it = true : bool
4247 :     - Main.compile ("../testcases/merge.tig", "foo.out");
4248 :     *** BACK-TRACE ***
4249 :     CALL lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
4250 :     CALL lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
4251 :     CALL lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
4252 :     GOTO lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
4253 :     CALL lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
4254 :     CALL lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
4255 :     CALL lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
4256 :     / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
4257 :     CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
4258 :     CALL lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
4259 :     CALL lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
4260 :     CALL main.sml:1.18-118.4: Main.compile[2]
4261 :     -
4262 :    
4263 :     ----------------------------------------------------------------------
4264 : blume 673 Name: Matthias Blumen
4265 :     Date: 2000/06/21 18:00:00 JST
4266 :     Tag: blume-20000621-manual
4267 :     Description:
4268 :    
4269 :     CM manual update: Path environments documented.
4270 :    
4271 :     ----------------------------------------------------------------------
4272 : blume 666 Name: Matthias Blume
4273 : blume 672 Date: 2000/06/19 13:40:00
4274 :     Tag: blume-20000619-manual
4275 :     Description:
4276 :    
4277 :     CM manual and system/README update. This only covers the fact that
4278 :     there are no more implicit anchors. (Path environments and the "bind"
4279 :     option to "cm" have yet to be documented.)
4280 :    
4281 :     ----------------------------------------------------------------------
4282 :     Name: Matthias Blume
4283 : blume 671 Date: 2000/06/19 11:05:00 JST
4284 :     Tag: blume-20000619-chdir-bugfix
4285 :     Description:
4286 :    
4287 :     Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
4288 :    
4289 :     ----------------------------------------------------------------------
4290 :     Name: Matthias Blume
4291 : blume 670 Date: 2000/06/18 22:00:10 JST
4292 :     Tag: blume-20000618-implicit-anchors-really-gone
4293 :     Description:
4294 :    
4295 :     I updates the previous HISTORY entry where I forgot to mention that
4296 :     implicit anchors are no longer with us.
4297 :    
4298 :     The current update also gets rid of the (now useless) controller
4299 :     CM.Control.implicit_anchors.
4300 :    
4301 :     ----------------------------------------------------------------------
4302 :     Name: Matthias Blume
4303 : blume 666 Date: 2000/06/16 17:30:00 JST
4304 :     Tag: blume-20000616-anchorenv
4305 :     Description:
4306 :    
4307 :     This patch implements the long anticipated (just kidding :) "anchor
4308 :     environment" mechanism. In the course of doing this, I also
4309 :     re-implemented CM's internal "SrcPath" module from scratch. The new
4310 :     one should be more robust in certain boundary cases. In any case, it
4311 :     is a lot cleaner than its predecessor (IMHO).
4312 :    
4313 :     This time, although there is yet another boot file format change, I
4314 :     kept the unpickler backward-compatible. As a result, no new bootfiles
4315 :     are necessary and bootstrapping is straightforward. (You cannot read
4316 :     new bootfiles into an old system, but the other way around is no
4317 :     problem.)
4318 :    
4319 :     Visible changes:
4320 :    
4321 : blume 670 ** 0. Implicit path anchors (without the leading $-symbol) are no
4322 :     longer recognized at all. This means that such path names are not
4323 :     illegal either. For example, the name basis.cm simply refers to a
4324 :     local file called "basis.cm" (i.e, the name is an ordinary path
4325 :     relative to .cm-files directory). Or, to put it differently, only
4326 :     names that start with $ are anchored paths.
4327 : blume 666
4328 : blume 670 ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
4329 :     vanished.
4330 :    
4331 :     John (Reppy) had critizised this as soon as I originally proposed and
4332 : blume 666 implemented it, but at that time I did not really deeply believe
4333 :     him. :) Now I came full-circle because I need the $<singlearc> syntax
4334 :     in another place where it cannot be seen as an abbreviation for
4335 :     $/<singlearc>. To avoid the confusion, $<singlearc> now means what it
4336 :     seems to mean (i.e., it "expands" into the corresponding anchor
4337 :     value).
4338 :    
4339 :     However, when paths are used as members in CM description files, it
4340 :     continues to be true that there must be at least another arc after the
4341 :     anchor. This is now enforced separately during semantic analysis
4342 : blume 670 (i.e., from a lexical/syntactical point of view, the notation is ok.)
4343 : blume 666
4344 :     ** 2. The "cm" class now accepts an option "bind". The option's value
4345 :     is a sub-option list of precisely two items -- one labeled "anchor"
4346 :     and the other one labeled "value". As you might expect, "anchor" is
4347 :     used to specify an anchor name to be bound, and "value" specifies what
4348 :     the anchor is being bound to.
4349 :    
4350 :     The value mu