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

SCM Repository

[smlnj] Diff of /sml/trunk/HISTORY
ViewVC logotype

Diff of /sml/trunk/HISTORY

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 570, Wed Mar 8 17:30:13 2000 UTC revision 716, Sat Nov 4 05:42:35 2000 UTC
# Line 13  Line 13 
13  Description:  Description:
14    
15  ----------------------------------------------------------------------  ----------------------------------------------------------------------
16  Name:  Name: Matthias Blume
17  Date:  Date: 2000/11/04 14:45:00
18  Tag:  Tag: blume-20001104-mlbuild
19  Description:  Description:
20    
21    - Made ml-build faster on startup.
22    - Documentation fixes.
23    
24    ----------------------------------------------------------------------
25    Name: Matthias Blume
26    Date: 2000/11/02 17:00:00 JST
27    Tag: blume-20001102-condcomp
28    Description:
29    
30    - Small tweaks to pickler -- new BOOTFILES!
31    - Version bumped to 110.29.2.
32    - Added conditional compilation facility to init.cmi (see comment there).
33    ----------------------------------------------------------------------
34    Name: Allen Leung
35    Date: 2000/10/23 19:31:00
36    Tag: leunga-20001023-demo-ra
37    
38    1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
39    2. Test programs for the graph library updated
40    3. Some new MLRISC demo programs added
41    
42    ----------------------------------------------------------------------
43    Name: Matthias Blume
44    Date: 2000/08/31 22:15:00 JST
45    Tag: blume-20001017-errmsg
46    Description:
47    
48    More error message grief: Where there used to be no messages, there
49    now were some that had bogus error regions.  Fixed.
50    
51    ----------------------------------------------------------------------
52    Name: Matthias Blume
53    Date: 2000/08/31 17:30:00 JST
54    Tag: blume-20001017-v110p29p1
55    Description:
56    
57    I made a version 110.29.1 with new bootfiles.
58    
59    Changes:  Modified pickler/unpickler for faster and leaner unpickling.
60              CM documentation changes and a small bugfix in CM's error reporting.
61    
62    ----------------------------------------------------------------------
63    Name: Lal George
64    Date: 2000/09/27 14:42:35 EDT
65    Tag: george-20000927-nodestatus
66    Description:
67    
68    Changed the type of the nodestatus, so that:
69    
70            SPILLED(~1)             is now SPILLED
71            SPILLED(m) where m>=0   is now MEMREG(m)
72            SPILLED(s) where s<~1   is now SPILL_LOC(~s)
73    
74    ----------------------------------------------------------------------
75    Name: Matthias Blume
76    Date: 2000/09/07 14:45:00 JST
77    Tag: blume-20000907-cmerrmsg
78    Description:
79    
80    Small tweak to CM to avoid getting ML syntax error messages twice.
81    
82    ----------------------------------------------------------------------
83    Name: Matthias Blume
84    Date: 2000/08/31 18:00:00 JST
85    Tag: blume-20000831-cvsbootfiles
86    Description:
87    
88    New URL for boot files (because the 110.29 files on the BL server do
89    now work correctly with my updated install scripts for yacc and lex).
90    
91    ----------------------------------------------------------------------
92    Name: Matthias Blume
93    Date: 2000/08/08 12:33:00 JST
94    Tag: blume-20000808-manual
95    Description:
96    
97    Tiny update to CM manual.
98    
99    ----------------------------------------------------------------------
100    Name: Allen Leung
101    Date: 2000/08/7 19:31:00
102    Tag: leunga-20000807-a-whole-bunch-of-stuff
103    
104      Moby, C--, SSA, x86, machine descriptions etc.  Should only affect C--
105    and Mobdy.
106    
107    1.  x86
108    
109       a.  Fixes to peephole module by John and Dan.
110       b.  Assembly fix to SETcc by Allen.
111       c.  Fix to c-call by John.
112       d.  Fix to spilling by John.  (This one deals with the missing FSTPT case)
113       e.  Instruction selection optimization to SETcc as suggested by John.
114    
115           For example,
116    
117            MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
118    
119           should generate:
120    
121            MOVL a, x
122            SUBL b, x
123            SHRL 31, x
124    
125    2.  IR stuff
126    
127         A bunch of new DJ-graph related algorithms added.  These
128         speed up SSA construction.
129    
130    3.  SSA + Scheduling
131    
132         Added code for SSA and scheduling to the repository
133    
134    ----------------------------------------------------------------------
135    Name: Lal George
136    Date: 2000/07/27 11:53:14 EDT
137    
138    Tag: lal-20000727-linux-ppc
139    Description:
140    
141     Made changes to support Linux PPC.
142     p.s. I have confirmation that the 110.29 boot files work fine.
143    
144    ----------------------------------------------------------------------
145    Name: Matthias Blume
146    Date: 2000/07/27 17:40:00 JST
147    Tag: blume-20000727-scripts
148    Description:
149    
150    !!!! WARNING !!!!
151    You must recompile the runtime system!
152    !!!! WARNING !!!!
153    
154    This is basically another round of script-enhancements:
155    
156    1. sml, ml-build, and ml-makedepend accept options -D and -U to define
157       and undefine CM preprocessor symbols.
158    
159    2. ml-build avoids generating a new heap image if it finds that the
160       existing one is still ok.  (The condition is that no ML file had to
161       be recompiled and all ML files are found to be older that the heap
162       file.)
163    
164       To make this work smoothly, I also hacked the runtime system as
165       well as SMLofNJ.SysInfo to get access to the heap image suffix
166       (.sparc-solaris, ...) that is currently being used.
167    
168       Moreover, the signature of CM.mk_standalone has changed.  See the
169       CM manual.
170    
171    3. ml-makedepend accepts additional options -n, -a, and -o.  (See the
172       CM manual for details.)
173    
174    4. More CM manual updates:
175        - all of the above has been documented.
176        - there is now a section describing the (CM-related) command line
177          arguments that are accepted by the "sml" command
178    
179    ----------------------------------------------------------------------
180    Name: Matthias Blume
181    Date: 2000/07/25 16:20:00 JST
182    Tag: blume-20000725-makedepend
183    Description:
184    
185    Added a script called ml-makedepend.  This can be used in makefiles
186    for Unix' make in a way very similar to the "makedepend" command for
187    C.
188    
189    The script internally uses function CM.sources.
190    
191    Synopsis:
192    
193        ml-makedepend [-f makefile] cmfile targetname
194    
195    The default for the makefile is "makefile" (or "Makefile" should
196    "makefile" not exist).
197    
198    ml-makedepend adds a cmfile/targetname-specific section to this
199    makefile (after removing the previous version of this section).  The
200    section contains a single dependency specification with targetname on
201    the LHS (targetname is an arbitrary name), and a list of files derived
202    from the cmfile on the RHS.  Some of the files on the RHS are
203    ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
204    "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
205    path names.  The makefile writer is responsible for making sure that
206    these variables have correct at the time "make" is invoked.
207    
208    ----------------------------------------------------------------------
209    Name: Matthias Blume
210    Date: 2000/07/22 23:30:00 JST
211    Tag: blume-20000722-urlupdate
212    Description:
213    
214    Changed BOOT and config/srcarchiveurl to point to BL server:
215    
216        ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
217    
218    ----------------------------------------------------------------------
219    Name: Matthias Blume
220    Date: 2000/07/18 18:00:00 JST
221    Tag: blume-20000718-Version_110_29
222    Description:
223    
224    1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
225    
226    2. Updated config/version to 110.29
227    
228    3. Updated config/srcarchiveurl
229    
230    3. New boot files!
231       ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
232    
233    ----------------------------------------------------------------------
234    Name: Matthias Blume
235    Date: 2000/07/11 13:58:00 JST
236    Tag: blume-20000711-doctypo
237    Description:
238    
239    Fixed a few typos in CM manual.
240    
241    ----------------------------------------------------------------------
242    Name: Allen Leung
243    Date: 2000/06/15 00:38:00
244    Tag: leunga-20000704-sparc-x86
245    
246    1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
247    2. fix to x86 RET bug [found by Dan Grossman]
248    3. sparc assembly bug fix for ticc instructions [found by Fermin]
249    
250       Affects c-- and moby only
251    
252    ----------------------------------------------------------------------
253    Name: Matthias Blume
254    Date: 2000/07/04 15:26:00
255    Tag: blume-20000704-trigger
256    Description:
257    
258    1. Improvements to CM manual.
259    2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
260       of getting a back-trace.  The function, when called, raises an
261       internal exception which explicitly carries the full back-trace history,
262       so it is unaffected by any intervening handle-raise pairs ("trivial"
263       or not).  The interactive loop will print that history once it arrives
264       at top level.
265       Short of having all exceptions implicitly carry the full history, the
266       recommended way of using this facility is:
267         - compile your program with instrumentation "on"
268         - run it, when it raises an exception, look at the history
269         - if the history is "cut off" because of some handler, go and modify
270           your program so that it explicitly calls BTrace.trigger
271         - recompile (still instrumented), and rerun; look at the full history
272    
273    ----------------------------------------------------------------------
274    Name: Matthias Blume
275    Date: 2000/07/03 15:36:00 JST
276    Tag: blume-20000702-manual
277    Description:
278    
279    Small corrections and updates to CM manual.
280    
281    ----------------------------------------------------------------------
282    Name: Matthias Blume
283    Date: 2000/06/29 16:04:00 JST
284    Tag: blume-20000629-yacctool
285    Description:
286    
287    Changes:
288    
289    1. Class "mlyacc" now takes separate arguments to pass options to
290       generated .sml- and .sig-files independently.
291    2. Corresponding CM manual updates.
292    3. BTrace module now also reports call sites.  (However, for loop clusters
293       it only shows from where the cluster was entered.)  There are associated
294       modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
295    
296    ----------------------------------------------------------------------
297    Name: Matthias Blume
298    Date: 2000/06/27 16:51:00 JST
299    Tag: blume-20000627-noweb
300    Description:
301    
302    Changes:
303    
304     1. Implemented "subdir" and "witness" options for noweb tool.
305        This caused some slight internal changes in CM's tool implementation.
306     2. Fixed bug in "tool plugin" mechanism.  This is essentially cleaning
307        some remaining issues from earlier path anchor changes.
308     3. Updated CM manual accordingly.
309    
310     4. Changed implementation of back-tracing so that I now consider it
311        ready for prime-time.
312    
313        In particular, you don't have to explicitly trigger the back-trace
314        anymore.  Instead, if you are running BTrace-instrumented code and
315        there is an uncaught exception (regardless of whether or not it was
316        raised in instrumented code), the top-level evalloop will print
317        the back-trace.
318    
319        Features:
320    
321          - Instrumented and uninstrumented code work together seemlessly.
322            (Of course, uninstrumented code is never mentioned in actual
323             back-traces.)
324    
325          - Asymptotic time- and space-complexity of instrumented code is
326            equal to that of uninstrumented code.  (This means that
327            tail-recursion is preserved by the instrumentation phase.)
328    
329          - Modules whose code has been instrumented in different sessions
330            work together without problem.
331    
332          - There is no penalty whatsoever on uninstrumented code.
333    
334          - There is no penalty on "raise" expressions, even in
335            instrumented code.
336    
337        A potential bug (or perhaps it is a feature, too):
338    
339          A back-trace reaches no further than the outermost instrumented
340          non-trivial "raise".  Here, a "trivial" raise is one that is the
341          sole RHS of a "handle" rule.  Thus, back-traces reach trough
342    
343               <exp> handle e => raise e
344    
345          and even
346    
347               <exp> handle Foo => raise Bar
348    
349          and, of course, through
350    
351               <exp> handle Foo => ...
352    
353         if the exception was not Foo.
354    
355         Back-traces always reach right through any un-instrumented code
356         including any of its "handle" expressions, trivial or not.
357    
358       To try this out, do the following:
359    
360         - Erase all existing binfiles for your program.
361           (You may keep binfiles for those modules where you think you
362            definitely don't need back-tracing.)
363         - Turn on back-trace instrumentation:
364              SMLofNJ.Internals.BTrace.mode (SOME true);
365         - Recompile your program.  (I.e., run "CM.make" or "use".)
366         - You may now turn instrumentation off again (if you want):
367              SMLofNJ.Internals.BTrace.mode (SOME false);
368         - Run your program as usual.  If it raises an exception that
369           reaches the interactive toplevel, then a back-trace will
370           automatically be printed.  After that, the toplevel loop
371           will print the exception history as usual.
372    
373    ----------------------------------------------------------------------
374    Name: Matthias Blume
375    Date: 2000/06/26 09:56:46 JST
376    Tag: blume-20000626-setup
377    Description:
378    
379    CM: - setup-parameter to "sml" added; this can be used to run arbitrary
380          ML code before and after compiling a file (e.g., to set compiler
381          flags)
382    
383    Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
384              - associated changes to btrace.sml (BTrace instrumentation pass)
385              - cleaner implementation of btimp.sml (BTrace tracing and report
386                module)
387    
388    CM manual: * new path encoding documented
389               * description of setup-parameter to "sml" added
390    
391    The biggest user-visible change to back-tracing is that it is no
392    longer necessary to compile all traced modules within the same
393    session.  (This was a real limitation.)
394    
395    ----------------------------------------------------------------------
396    Name: Matthias Blume
397    Date: 2000/06/24 12:40:00 JST
398    Tag: blume-20000624-startup
399    Description:
400    
401    Fixes startup slowdown problem.  (I was calling SrcPath.sync a _tad_
402    bit too often -- to put it mildly. :)
403    
404    ----------------------------------------------------------------------
405    Name: Matthias Blume
406    Date: 2000/06/23 18:20:00 JST
407    Tag: blume-20000623-btrace
408    Description:
409    
410    This updates adds a backtrace facility to aid programmers in debugging
411    their programs.  This involves the following changes:
412    
413    1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
414       keeping track of the current call stack.  When programs are compiled
415       in a special mode, the compiler will insert calls to these hooks
416       into the user program.
417       "Hook" means that it is possible for different implementations of
418       back-tracing to register themselves (at different times).
419    
420    2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
421       as an Absyn.dec->Absyn.dec rewrite.  Normally this phase is turned off.
422       It can be turned on using this call:
423         SMLofNJ.Internals.BTrace.mode (SOME true);
424       Turning it off again:
425         SMLofNJ.Internals.BTrace.mode (SOME false);
426       Querying the current status:
427         SMLofNJ.Internals.BTrace.mode NONE;
428       Annotated programs are about twice as big as normal ones, and they
429       run a factor of 2 to 4 slower with a dummy back-trace plugin (one
430       where all hooks do nothing).  The slowdown with a plugin that is
431       actually useful (such as the one supplied by default) is even greater,
432       but in the case of the default plugin it is still only an constant
433       factor (amortized).
434    
435    3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
436       with a sub-structure BTrace for controlling back-tracing.  In particular,
437       the above-mentioned function "mode" controls whether the annotation
438       phase is invoked by the compiler.  Another important function is
439       "trigger": when called it aborts the current execution and causes
440       the top-level loop to print a full back-trace.
441    
442    4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
443       for back-tracing.  It keeps track of the dynamic call stack and in
444       addition to that it keeps a partial history at each "level" of that
445       stack.  For example, if a tail-calls b, b tail-calls c, and c tail-calls
446       d and b (at separate times, dynamically), then the report will show:
447    
448       GOTO   d
449             /c
450       GOTO  \b
451       CALL   a
452    
453       This shows that there was an initial non-tail call of a, then a
454       tail-call to b or c, looping behavior in a cluster of functions that
455       consist of b and c, and then a goto from that cluster (i.e., either from
456       b or from c) to d.
457    
458       Note that (depending on the user program) the amount of information
459       that the back-trace module has to keep track of at each level is bounded
460       by a constant.  Thus, the whole implementation has the same asymptotical
461       complexity as the original program (both in space and in time).
462    
463    5. compiler/TopLevel/interact/evalloop.sml has been modified to
464       handle the special exception SMLofNJ.Internals.BTrace.BTrace
465       which is raised by the "trigger" function mentioned above.
466    
467    Notes on usage:
468    
469    - Annotated code works well together with unannotated code:
470    Unannotated calls simply do not show up at all in the backtrace.
471    
472    - It is not a good idea to let modules that were annotated during
473    different sessions run at the same time.  This is because the compiler
474    chooses small integers to identify individual functions, and there
475    will be clashes if different modules were compiled in separate sessions.
476    (Nothing will crash, and you will even be told about the clashes, but
477    back-trace information will in general not be useful.)
478    
479    - Back-tracing can be confused by callcc and capture.
480    
481    - The only way of getting a back-trace right now is to explicitly
482    invoke the "trigger" function from your user program.  Eventually, we
483    should make every exception carry back-trace information (if
484    available).  But since this creates more overhead at "raise"-time
485    (similar to the current exnHistory overhead), I have not yet
486    implemented this.  (The implementation will be rather easy.)  With
487    exceptions carrying back-trace information, this facility will be even
488    more useful because users don't need to modify their programs...
489    
490    - While it is possible to compile the compiler with back-trace
491    annotations turned on (I did it to get some confidence in
492    correctness), you must make absolutely sure that core.sml and
493    btimp.sml are compiled WITHOUT annotation!  (core.sml cannot actually
494    be compiled with annotation because there is no core access yet, but
495    if you compile btimp.sml with annotation, then the system will go into
496    an infinite recursion and crash.)
497    Since CM currently does not know about BTrace, the only way to turn
498    annotations on and off for different modules of the compiler is to
499    interrupt CMB.make, change the settings, and re-invoke it.  Of course,
500    this is awkward and clumsy.
501    
502    Sample sessions:
503    
504    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
505    - SMLofNJ.Internals.BTrace.mode (SOME true);
506    [autoloading]
507    [autoloading done]
508    val it = false : bool
509    - structure X = struct
510    -     fun main n = let
511    -         fun a (x, 0) = d x
512    -           | a (x, n) = b (x, n - 1)
513    -         and b (x, n) = c (x, n)
514    -         and c (x, n) = a (x, n)
515    -         and d x = e (x, 3)
516    -         and e (x, 0) = f x
517    -           | e (x, n) = e (x, n - 1)
518    -         and f 0 = SMLofNJ.Internals.BTrace.trigger ()
519    -           | f n = n * g (n - 1)
520    -         and g n = a (n, 3)
521    -     in
522    -         f n
523    -     end
524    - end;
525    structure X : sig val main : int -> int end
526    - X.main 3;
527    *** BACK-TRACE ***
528    GOTO   stdIn:4.2-13.20: X.main[2].f
529    GOTO-( stdIn:4.2-13.20: X.main[2].e
530    GOTO   stdIn:4.2-13.20: X.main[2].d
531         / stdIn:4.2-13.20: X.main[2].a
532         | stdIn:4.2-13.20: X.main[2].b
533    GOTO-\ stdIn:4.2-13.20: X.main[2].c
534    CALL   stdIn:4.2-13.20: X.main[2].g
535    GOTO   stdIn:4.2-13.20: X.main[2].f
536    GOTO-( stdIn:4.2-13.20: X.main[2].e
537    GOTO   stdIn:4.2-13.20: X.main[2].d
538         / stdIn:4.2-13.20: X.main[2].a
539         | stdIn:4.2-13.20: X.main[2].b
540    GOTO-\ stdIn:4.2-13.20: X.main[2].c
541    CALL   stdIn:4.2-13.20: X.main[2].g
542    GOTO   stdIn:4.2-13.20: X.main[2].f
543    GOTO-( stdIn:4.2-13.20: X.main[2].e
544    GOTO   stdIn:4.2-13.20: X.main[2].d
545         / stdIn:4.2-13.20: X.main[2].a
546         | stdIn:4.2-13.20: X.main[2].b
547    GOTO-\ stdIn:4.2-13.20: X.main[2].c
548    CALL   stdIn:4.2-13.20: X.main[2].g
549    GOTO   stdIn:4.2-13.20: X.main[2].f
550    CALL   stdIn:2.15-17.4: X.main[2]
551    -
552    
553    (Note that because of a FLINt bug the above code currently does not
554    compile without BTrace turned on.)
555    
556    Here is another example, using my modified Tiger compiler:
557    
558    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
559    - SMLofNJ.Internals.BTrace.mode (SOME true);
560    [autoloading]
561    [autoloading done]
562    val it = false : bool
563    - CM.make "sources.cm";
564    [autoloading]
565    ...
566    [autoloading done]
567    [scanning sources.cm]
568    [parsing (sources.cm):parse.sml]
569    [creating directory CM/SKEL ...]
570    [parsing (sources.cm):tiger.lex.sml]
571    ...
572    [wrote CM/sparc-unix/semant.sml]
573    [compiling (sources.cm):main.sml]
574    [wrote CM/sparc-unix/main.sml]
575    [New bindings added.]
576    val it = true : bool
577    - Main.compile ("../testcases/merge.tig", "foo.out");
578    *** BACK-TRACE ***
579    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
580    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
581    CALL   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
582    GOTO   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
583    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
584    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
585    CALL   lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
586         / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
587    CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
588    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
589    CALL   lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
590    CALL   main.sml:1.18-118.4: Main.compile[2]
591    -
592    
593    ----------------------------------------------------------------------
594    Name: Matthias Blumen
595    Date: 2000/06/21 18:00:00 JST
596    Tag: blume-20000621-manual
597    Description:
598    
599    CM manual update: Path environments documented.
600    
601    ----------------------------------------------------------------------
602    Name: Matthias Blume
603    Date: 2000/06/19 13:40:00
604    Tag: blume-20000619-manual
605    Description:
606    
607    CM manual and system/README update.  This only covers the fact that
608    there are no more implicit anchors.  (Path environments and the "bind"
609    option to "cm" have yet to be documented.)
610    
611    ----------------------------------------------------------------------
612    Name: Matthias Blume
613    Date: 2000/06/19 11:05:00 JST
614    Tag: blume-20000619-chdir-bugfix
615    Description:
616    
617    Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
618    
619    ----------------------------------------------------------------------
620    Name: Matthias Blume
621    Date: 2000/06/18 22:00:10 JST
622    Tag: blume-20000618-implicit-anchors-really-gone
623    Description:
624    
625    I updates the previous HISTORY entry where I forgot to mention that
626    implicit anchors are no longer with us.
627    
628    The current update also gets rid of the (now useless) controller
629    CM.Control.implicit_anchors.
630    
631    ----------------------------------------------------------------------
632    Name: Matthias Blume
633    Date: 2000/06/16 17:30:00 JST
634    Tag: blume-20000616-anchorenv
635    Description:
636    
637    This patch implements the long anticipated (just kidding :) "anchor
638    environment" mechanism.  In the course of doing this, I also
639    re-implemented CM's internal "SrcPath" module from scratch.  The new
640    one should be more robust in certain boundary cases.  In any case, it
641    is a lot cleaner than its predecessor (IMHO).
642    
643    This time, although there is yet another boot file format change, I
644    kept the unpickler backward-compatible.  As a result, no new bootfiles
645    are necessary and bootstrapping is straightforward.  (You cannot read
646    new bootfiles into an old system, but the other way around is no
647    problem.)
648    
649    Visible changes:
650    
651    ** 0. Implicit path anchors (without the leading $-symbol) are no
652    longer recognized at all. This means that such path names are not
653    illegal either.  For example, the name basis.cm simply refers to a
654    local file called "basis.cm" (i.e, the name is an ordinary path
655    relative to .cm-files directory).  Or, to put it differently, only
656    names that start with $ are anchored paths.
657    
658    ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
659    vanished.
660    
661    John (Reppy) had critizised this as soon as I originally proposed and
662    implemented it, but at that time I did not really deeply believe
663    him. :) Now I came full-circle because I need the $<singlearc> syntax
664    in another place where it cannot be seen as an abbreviation for
665    $/<singlearc>.  To avoid the confusion, $<singlearc> now means what it
666    seems to mean (i.e., it "expands" into the corresponding anchor
667    value).
668    
669    However, when paths are used as members in CM description files, it
670    continues to be true that there must be at least another arc after the
671    anchor.  This is now enforced separately during semantic analysis
672    (i.e., from a lexical/syntactical point of view, the notation is ok.)
673    
674    ** 2. The "cm" class now accepts an option "bind".  The option's value
675    is a sub-option list of precisely two items -- one labeled "anchor"
676    and the other one labeled "value".  As you might expect, "anchor" is
677    used to specify an anchor name to be bound, and "value" specifies what
678    the anchor is being bound to.
679    
680    The value must be a directory name and can be given in either standard
681    syntax (including the possibility that it is itself an anchored path)
682    or native syntax.
683    
684    Examples:
685    
686       foo.cm (bind:(anchor:bar value:$mystuff/bar))
687       lib.cm (bind:(anchor:a value:"H:\\x\\y\\z"))  (* only works under windows *)
688    
689    and so on.
690    
691    The meaning of this is that the .cm-file will be processed with an
692    augmented anchor environment where the given anchor(s) is/are bound to
693    the given values(s).
694    
695    The rationale for having this feature is this: Suppose you are trying
696    to use two different (already stable) libraries a.cm and b.cm (that
697    you perhaps didn't write yourself).  Further, suppose each of these
698    two libraries internally uses its own auxiliary library $aux/lib.cm.
699    Normally you would now have a problem because the anchor "lib" can not
700    be bound to more than one value globally.  Therefore, the project that
701    uses both a.cm and b.cm must locally redirect the anchor to some other
702    place:
703    
704       a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
705       b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
706    
707    This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
708    /usr/lib/smlnj/b-stuff/aux.cm, respectively.
709    
710    Hard-wiring path names is a bit inflexible (and CM will verbosely warn
711    you when you do so at the time of CM.stabilize).  Therefore, you can
712    also use an anchored path as the value:
713    
714      a.cm (bind:(anchor:lib value:$a-lib))
715      b.cm (bind:(anchor:lib value:$b-lib))
716    
717    Now you can globally configure (using the usual CM.Anchor.anchor or
718    pathconfig machinery) bindings for "a-lib" and "b-lib".  Since "lib"
719    itself is always locally bound, setting it globally is no longer
720    meaningful or necessary (but it does not hurt either).  In fact, "lib"
721    can still be used as a global anchor for separate purposes.  As a
722    matter of fact, one can locally define "lib" in terms of a global
723    "lib":
724    
725      a.cm (bind:(anchor:lib value:$lib/a))
726      b.cm (bind:(anchor:lib value:$lib/b))
727    
728    ** 3: The encoding of path names has changed.  This affects the way
729    path names are shown in CM's progress report and also the internal
730    protocol encoding used for parallel make.
731    
732    The encoding now uses one or more ':'-separated segments.  Each
733    segments corresponds to a file that has been specified relative to the
734    file given by its preceding segment.  The first segment is either
735    relative to the CWD, absolute, or anchored.  Each segment itself is
736    basically a Unix pathname; all segments but the first are relative.
737    
738    Example:
739    
740       $foo/bar/baz.cm:a/b/c.sml
741    
742    This path denotes the file bar/a/b/c.sml relative to the directory
743    denoted by anchor "foo".  Notice that the encoding also includes
744    baz.cm which is the .cm-file that listed a/b/c.sml.  As usual, such
745    paths are resolved relative to the .cm-files directory, so baz.cm must
746    be ignored to get the "real" pathname.
747    
748    To make this fact more obvious, CM puts the names of such "virtual
749    arcs" into parentheses when they appear in progress reports. (No
750    parentheses will appear in the internal protocol encoding.)  Thus,
751    what you really see is:
752    
753      $foo/bar/(baz.cm):a/b/c.sml
754    
755    I find this notation to be much more informative than before.
756    
757    Another new feature of the encoding is that special characters
758    including parentheses, colons, (back)slashes, and white space are
759    written as \ddd (where ddd is the decimal encoding of the character).
760    
761    *** The CM manual still needs to be updated.
762    
763    ----------------------------------------------------------------------
764    Name: Allen Leung
765    Date: 2000/06/15 00:38:00
766    Tag: leunga-20000615-x86-peephole
767    
768    x86 Peephole fix by Fermin.  Affects c-- and moby only.
769    
770    ----------------------------------------------------------------------
771    Name: Matthias Blume
772    Date: 2000/06/12 11:40:00
773    Tag: blume-20000612-parmakefix
774    Description:
775    
776    More cleanup after changing the file naming scheme: This time I
777    repaired the parallel make mechanism for CMB.make which I broke earlier.
778    
779    ----------------------------------------------------------------------
780    Name: Allen Leung
781    Date: 2000/06/09 01:25:00
782    Tag: leunga-20000609-various
783    
784    None of these things should affect normal SML/NJ operations
785    
786    1. Peephole improvements provided by Fermin (c--)
787    2. New annotation DEFUSE for adding extra dependence (moby)
788    3. New X86 LOCK instructions (moby)
789    4. New machine description language for reservation tables (scheduling)
790    5. Fixes to various optimization/analysis modules (branch chaining, dominator
791       trees etc.)
792    6. I've changed the CM files so that they can work with versions
793       110.0.6, 110.25 and 110.28
794    
795    ----------------------------------------------------------------------
796    Name: Matthias Blume
797    Date: 2000/06/09 12:40:00
798    Tag: blume-20000609-log
799    Description:
800    
801    - Removed all(?) remaining RCS Log entries from sources.
802    
803    - Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
804      anchored paths).
805    
806    ----------------------------------------------------------------------
807    Name: Matthias Blume
808    Date: 2000/06/07 17:00:00 JST
809    Tag: blume-20000607-no-implicit-anchors
810    Description:
811    
812    1. This update changes the default setting for
813    CM.Control.implicit_anchors from true to false.  This means that
814    implicit anchors are no longer permitted by default.  I also tried to
815    make sure that nothing else still relies on implicit anchors.
816    (This is the next step on the schedule towards a CM that does not even
817    have the notion of implicit anchors anymore.)
818    
819    2. More CM manual updates.
820    
821    3. I managed to track down and fix the pickling bug I mentioned last
822    time.  Because of the previously existing workaround, this entails no
823    immediate practical changes.
824    
825    ----------------------------------------------------------------------
826    Name: Matthias Blume
827    Date: 2000/06/06 11:15:00 JST
828    Tag: blume-20000606-lazierpickle
829    Description:
830    
831    !!!! NEW BOOT FILES !!!!
832    
833    * The main purpose of this update is to make library pickles lazier in
834    order to reduce the initial space penalty for autoloading a library.
835    As a result, it is now possible to have $smlnj/compiler.cm
836    pre-registered.  This should take care of the many complaints or
837    inquiries about missing structure Compiler.  This required changes to
838    CM's internal data structures and small tweaks to some algorithms.
839    
840    As a neat additional effect, it is no longer necessary (for the sake
841    of lean heap image files) to distinguish between a "minimal" CM and a
842    "full" CM.  Now, there is only one CM (i.e., the "full" version:
843    $smlnj/cm.cm aka $smlnj/cm/full.cm), and it is always available at the
844    interactive top level. ($smlnj/cm/minimal.cm is gone.)
845    
846    To make the life of compiler-hackers easier, "makeml" now also
847    pre-registers $smlnj/cmb.cm (aka $smlnj/cmb/current.cm).  In other
848    words, after you bootstrap a new sml for the first time, you will not
849    have to autoload $smlnj/cmb.cm again afterwards.  (The first time
850    around you will still have to do it, though.)
851    
852    * A second change consists of major updates to the CM manual.  There
853    are now several appendices with summary information and also a full
854    specification of the CM description file syntax.
855    
856    * In directory src/system I added the script "allcross".  This script
857    invokes sml and cross-compiles the compiler for all supported
858    architectures.  (Useful when providing a new set of boot files.)
859    
860    * There seems to be a latent bug in my "lazy pickles" mechanism.  I
861    added a small tweak to pickle-util.sml to work around this problem,
862    but it is not a proper fix yet.  I will investigate further.  (The
863    effect of the bug was an inflation of library pickle size.)
864    
865    * Version number increased to 110.28.1 (to avoid compatibility problems).
866    
867    ----------------------------------------------------------------------
868    Name: Allen Leung
869    Date: 2000/05/25 17:28 EDT
870    Tag: leunga-20000525-ra
871    Description:
872    
873      Fixed a bug in freezing phase of the register allocator.
874    
875    ----------------------------------------------------------------------
876    Name: Allen Leung
877    Date: 2000/05/15 22:53 EDT
878    Tag: leunga-20000515-alpha-x86-ra
879    Description:
880    
881      1. Alpha
882    
883          Slight cleanup.  Removed the instruction SGNXL
884    
885      2. X86
886    
887          Added the following instructions to the instruction set:
888    
889            ROLx, RORx,
890            BTx, BTSx, BTLx, BTRx,
891            XCHGx, and variants with the LOCK prefix
892    
893      3. Register Allocation
894    
895          The module ra-rewrite-with-renaming has been improved.
896    
897      These have no effect on SML/NJ.
898    
899    ----------------------------------------------------------------------
900    Name: Matthias Blume
901    Date: 2000/05/15 16:20:00 JST
902    Tag: blume-20000515-lightrebuild
903    Description:
904    
905    1. I added an alternative to "-rebuild" to "makeml".  The difference is
906       that prior to calling CMB.make' the CM-variable "LIGHT" will be
907       defined.  In effect, the command will not build any cross-compiler
908       backends and therefore finish more quickly.
909    
910       The "fixpt" script also takes a "-light" switch to be able to use
911       this new facility while compiling for a fixpoint.
912    
913    2. I replaced all mentions of anchored paths in group owner specifications
914       with simple relative paths (usually starting with "..").
915       The rationale is that a library's internal workings should not be
916       compromised by the lack of some anchor.  (An anchor is necessary
917       for someone who wants to refer to the library by an anchored path,
918       but it should not be necessary to build the same library in the first
919       place.)
920    
921    3. I changed the way CM's tool mechanism determines the shell command
922       string used for things like ml-yacc etc. so that it does not break
923       when CM.Control.implicit_anchors is turned off.
924    
925    ----------------------------------------------------------------------
926    Name: Matthias Blume
927    Date: 2000/05/12 18:20:00 JST
928    Tag: blume-20000512-ml-build
929    Description:
930    
931    Fixed a bug in config/_ml-build that prevented ml-yacc and ml-lex from
932    getting installed properly (by config/install.sh).
933    
934    ----------------------------------------------------------------------
935    Name: Matthias Blume
936    Date: 2000/05/12 17:30:00 JST
937    Tag: blume-20000512-anchors
938    Description:
939    
940    !!! NEW BOOT FILES !!!
941    
942    This change is in preparation of fading out support for "implicitly
943    anchored path names".  I went through all sources and used the
944    explicit (and relatively new) $-notation.  See system/README and the
945    CM manual for more info on this.
946    
947    I also modified the anchoring scheme for some things such as "smlnj",
948    "MLRISC", "cm", etc. to take advantage of the fact that explicit
949    anchors are more expressive: anchor name and first arc do not have to
950    coincide.  This entails the following user-visible change:
951    
952    You have to write $smlnj/foo/bar instead of smlnj/foo/bar.  In
953    particular, when you fire up sml with a command-line argument, say,
954    e.g.:
955    
956       sml '$smlnj/cmb.cm'
957    
958    At the ML toplevel prompt:
959    
960       CM.autoload "$smlnj/cmb.cm";
961    
962    There is also a new controller in CM.Control that can be used to turn
963    off all remaining support for implicit anchors by saying:
964    
965        CM.autoload "$smlnj/
966        #set CM.Control.implicit_anchors false;
967    
968    This causes CM to reject implicitly anchored paths.  This is (for the
969    time being) less permissive than the "final" version where there will
970    be no more such implicit anchors and relative paths will be just that:
971    relative.
972    
973    The next step (version after next version?) will be to make the
974    default for CM.Control.implicit_anchors false.  After the dust has
975    settled, I can then produce the "final" version of this...
976    
977    Note: Since bootstrapping is a bit tricky, I provided new boot files.
978    
979    ----------------------------------------------------------------------
980    Name: Matthias Blume
981    Date: 2000/05/11 16:30:00 JST
982    Tag: blume-20000511-sources
983    Description:
984    
985    The main change is that I added function CM.sources as a generalized
986    version of the earlier CM.makedepend.  This entails the following
987    additional changes:
988    
989      - CM.makedepend has been dropped.
990    
991      - CM manual has been updated.
992    
993      - TOOLS signature and API have been changed.
994    
995    ----------------------------------------------------------------------
996    Name: Allen Leung
997    Date: 2000/05/10 21:17 EDT
998    Tag: leunga-20000510-moby-c--ssa
999    Description:
1000    
1001      Various bug fixes and new features for C--, Moby and MLRISC optimizations.
1002    None of these affect SML/NJ.
1003    
1004    1. Register Allocation
1005    
1006        a. A new ra spilling module (ra/ra-spill-with-renaming) is implemented.
1007           This module tries to remove local (i.e. basic block level) redundancies
1008           during spilling.
1009    
1010        b. A new framework for performing region based register allocation.
1011           Not yet entirely functional.
1012    
1013    2. X86
1014    
1015       a. DefUse for POP was missing the stack pointer [found by Lal]
1016       b. Reload for CALL was incorrect in X86Spill [found by John]
1017       c. Various fixes in X86Spill so that it can be used correctly for
1018          the new spilling module.
1019    
1020    3. SSA/IR
1021    
1022       a. New module ir/dj-dataflow.sml implements elimination based
1023          data flow analysis.
1024    
1025    4. MLRiscGen
1026    
1027       a. Fix for gc type annotation
1028    
1029    5. MDGen
1030    
1031       Various fixes for machine description -> ml code translation.  For ssa
1032       only.
1033    
1034    ----------------------------------------------------------------------
1035    Name: Allen Leung
1036    Date: 2000/05/08 22:17 EDT
1037    Tag: leunga-20000508-labexp
1038    Description:
1039    
1040      Fermin has found a few assembly problems with constant expressions
1041      generated in LabelExp.  Mostly, the problems involve extra parentheses,
1042      which choke on dumb assemblers.  This is his fix.
1043    
1044    ----------------------------------------------------------------------
1045    Name: Dave MacQueen
1046    Date: 2000/04/09 14:00 EDT
1047    Tag: dbm-20000502-Version_110_28
1048    Description:
1049    
1050    1. Updated src/compiler/TopLevel/main/version.sml to version 110.28
1051    
1052    2. Updated config/version to 110.28
1053    
1054    3. Updated config/srcarchiveurl
1055    
1056    3. New boot files!
1057       ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.28/
1058    
1059    ----------------------------------------------------------------------
1060    Name: Matthias Blume
1061    Date: 2000/05/01 19:05:00 JST
1062    Tag: blume-20000501-noweb
1063    Description:
1064    
1065    A new noweb tool has been added.  The existing system is entirely
1066    unaffected by this, but some CM users have asked for renewed noweb
1067    support.  Everything is documented in the CM manual.
1068    
1069    New (plugin) libraries:
1070    
1071       noweb-tool.cm
1072       nw-ext.cm
1073    
1074    ----------------------------------------------------------------------
1075    Name: Dave MacQueen
1076    Date: 2000/04/30 12:40PM EDT
1077    Tag: dbm-20000430-bug_fixes
1078    Description:
1079    
1080    1. Fix for bug 1498
1081       smlnj/src/system/Basis/Implementation/Unsafe/object.sig
1082       smlnj/src/system/Basis/Implementation/Unsafe/object.sml
1083         added toRealArray function
1084       smlnj/src/compiler/MiscUtil/print/ppobj.sml
1085         added check for tag Obj.RealArray to array printing case in ppObj
1086    
1087    2. Fix for bug 1510
1088       smlnj/src/compiler/Semant/types/typesutil.sml
1089         fixed definition of dummyargs (used by equalTycon) so that
1090         dummy args are distinct types
1091    
1092    ----------------------------------------------------------------------
1093    Name: Matthias Blume
1094    Date: 2000/04/30 01:00:00 JST
1095    Tag: blume-20000430-versions
1096    Description:
1097    
1098    1. CM version numbering added.  This is an implementation of Lal's
1099       proposal for adding version numbers and version checking to .cm
1100       files.  Lal said that his proposal was just that -- a proposal.
1101       For the time being I went ahead and implemented it so that people
1102       can comment on it.  Everything is completely backward-compatible
1103       (except for the stable library format, i.e., new bootfiles!).
1104    
1105       As usual, see the CM manual for details.
1106    
1107    2. An alternative syntax for anchored paths has been implemented.
1108       Dave has recently voiced the same concerns that I had when I did
1109       this, so there should be some support.  My take is that eventually
1110       I will let support for the current syntax (where anchors are
1111       "implicit") fade out in favor of the new, explicit syntax.
1112       In order to be backward-compatible, both old and new syntax are
1113       currently supported.
1114    
1115       Again, see the CM manual for details.
1116    
1117    3. Parallel make is trying to be slightly smarter:  When the master
1118       process finds a "bottleneck", i.e., when there is only one
1119       compilation unit that can be compiled and everybody else is
1120       waiting on it, then it will simply compile it directly instead
1121       of clumsily telling one of the slaves to do it.
1122    
1123    4. Support for "unsharing" added.  This is necessary in order to be
1124       able to have two different versions of the same library running
1125       at the same time (e.g., for trying out a new MLRISC while still
1126       having the old MLRISC linked into the current compiler, etc.)
1127       See the CM manual.
1128    
1129    5. Simple "makedepend" functionality added for generating Makefile
1130       dependency information.  (This is rather crude at the moment.
1131       Expect some changes here in the future.)
1132    
1133    6. ".fun" added as a recognized suffix for ML files. Also documented
1134       explicitly in the manual that the fallback behavior (unknown suffix
1135       -> ML file) is not an official feature!
1136    
1137    7. Small changes to the pickler for stable libraries.
1138    
1139    8. Several internal changes to CM (for cleanup/improvement).
1140    
1141    
1142    !!!! NEW BINFILES !!!!
1143    
1144    ----------------------------------------------------------------------
1145    Name: Matthias Blume
1146    Date: 2000/04/28 17:30:00 JST
1147    Tag: blume-20000428-pathconfig
1148    Description:
1149    
1150    1. I changed config/install.sh to remove duplicate entries from the
1151       lib/pathconfig file at the end.  Moreover, the final version of
1152       lib/pathconfig is sorted alphabetically.  The same (sorting) is done
1153       in src/system/installml.
1154    
1155    2. The config/install.sh script now consistently uses relative
1156       pathnames in lib/pathconfig whenever the anchor is in the lib
1157       directory.  (So far this was true for the libraries that come
1158       pre-compiled and bundled as part of the bootfiles but not for
1159       libraries that are compiled by the script itself.)
1160    
1161    ----------------------------------------------------------------------
1162    Name: Matthias Blume
1163    Date: 2000/04/26 13:10:00 JST
1164    Tag: blume-20000426-fun_suffix
1165    Description:
1166    
1167    Added ".fun" as a recognized file name suffix (for ML code).
1168    
1169    ----------------------------------------------------------------------
1170    Name: Allen Leung
1171    Date: 2000/04/25 17:00:00 EST
1172    Tag: leunga-20000425-alpha-ra
1173    Description:
1174    
1175    1. Alpha
1176    
1177        PSEUDOARITH was missing in AlphaRewrite.  This causes an endless loop
1178    in C--.
1179    
1180    2. RA
1181    
1182       Added a flag "ra-dump-size" to print out the size of the flowgraph
1183       and the interference graph.
1184    
1185    ----------------------------------------------------------------------
1186    Name: Dave MacQueen
1187    Date: 2000/04/25/
1188    Tag: dbm-20000425-mlyacc_doc_examples
1189    Description:
1190      Updated mlyacc.tex sections 5 and 7 for SML '97 and CM.
1191      Updated all three examples in src/ml-yacc/examples to run
1192      under 110.* using CM.make.
1193    
1194    ----------------------------------------------------------------------
1195    Name: Allen Leung
1196    Date: 2000/04/20 23:04:00 EST
1197    Tag: leunga-20000420-ssa-c---stuff
1198    Description:
1199    
1200      This update synchronizes my repository with Yale's.  Most of these
1201    changes, however, do not affect SML/NJ at all (the RA is an exception).
1202    
1203    1. Register Allocator
1204    
1205       a. An improvement in the interference graph construction:
1206          Given a copy
1207    
1208                s <- t
1209    
1210          no interference edge between s and t is added for this definition of s.
1211    
1212       b. I've added two new spill heuristic modules that Fermin and I developed
1213          (in the new library RA.cm). These are unused in SML/NJ but maybe
1214          useful for others (Moby?)
1215    
1216    2. X86
1217    
1218       a. Various fixes in the backend provided by Fermin [C--] and Lal.
1219    
1220    3. Alpha
1221    
1222       a. Added the BSR instruction and code generation that goes with it [C--]
1223       b. Other fixes too numerous to recount provided by Fermin [C--]
1224    
1225    4. Regmaps
1226    
1227       a. The regmaps are not initialized with the identity physical bindings
1228          at creation time.  This is unneeded.
1229    
1230    5. MLRISC Optimizations
1231    
1232       a. The DJ-Graph module can now compute the iterated dominance frontiers
1233          intersects with liveness incrementally in linear time! Woohoo!
1234          This is now used in my new SSA construction algorithm.
1235    
1236       b. THe branch reorganization module is now smarter about linear chains of
1237          basic blocks.
1238    
1239    
1240    ----------------------------------------------------------------------
1241    Name: Matthias Blume
1242    Date: 2000/04/12 13:52:00 JST
1243    Tag: blume_main_v110p27_1
1244    Description:
1245    
1246    Changed install.sh script to handle archive files without version number
1247    and to use "boot.<arch>-<os>" instead of "sml.boot.<arch>-<os>" for the
1248    name of the boot file archive.
1249    
1250    ----------------------------------------------------------------------
1251    Name: Dave MacQueen
1252    Date: 2000/04/09 14:00 EDT
1253    Tag: dbm-20000410-Version_110_27
1254    Description:
1255    
1256    1. Updated src/compiler/TopLevel/main/version.sml to version 110.27
1257    
1258    2. Updated src/config/version to 110.27
1259    
1260    3. New boot files!
1261    
1262    ----------------------------------------------------------------------
1263    Name: Allen Leung
1264    Date: 2000/04/09 19:09:00 EST
1265    Tag: leunga-20000409-misc
1266    Description:
1267    
1268    1.  Yet another fix for x86 assembly for idivl, imull, mull and friends.
1269    
1270    2.  Miscellaneous improvements to MLRISC (unused in sml/nj)
1271    
1272    ----------------------------------------------------------------------
1273    Name: Stefan
1274    Date: 2000/04/07 10:00:00 EDT
1275    Tag: monnier-20000406-branch-handling
1276    Description:
1277    
1278    Improved handling of branches (mostly those generated from
1279    polymorphic equality), removed switchoff and changed the
1280    default optimization settings (more cpsopt and less flintopt).
1281    
1282    ----------------------------------------------------------------------
1283    Name: Allen Leung
1284    Date: 2000/04/06 01:30:00 EST
1285    Tag: leunga-20000406-peephole-x86-SSA-2
1286    Description:
1287    
1288       Forgot a few files.
1289    
1290    ----------------------------------------------------------------------
1291    Name: Allen Leung
1292    Date: 2000/04/06 00:36:00 EST
1293    Tag: leunga-20000406-peephole-x86-SSA
1294    Description:
1295    
1296    1.  New Peephole code
1297    
1298    2.  Minor improvement to X86 instruction selection
1299    
1300    3.  Various fixes to SSA and machine description -> code translator
1301    
1302    ----------------------------------------------------------------------
1303    Name: Matthias Blume
1304    Date: 2000/04/05 12:30:00 JST
1305    Tag: blume_main_v110p26p2_3
1306    Description:
1307    
1308    This update just merges three minor cosmetic updates to CM's sources
1309    to get ready for the 110.27 code freeze on Friday.  No functionality
1310    has changed.
1311    
1312    ----------------------------------------------------------------------
1313    Name: Allen Leung
1314    Date: 2000/04/04 19:39:00 EST
1315    Tag: leunga-20000404-x86-asm
1316    Description:
1317    
1318    1.  Fixed a problem in X86 assembly.
1319    
1320        Things like
1321    
1322           jmp %eax
1323           jmp (%eax)
1324    
1325        should be output as
1326    
1327           jmp *%eax
1328           jmp *(%eax)
1329    
1330    2.  Assembly output
1331    
1332          Added a new flag
1333    
1334              "asm-indent-copies" (default to false)
1335    
1336          When this flag is on, parallel copies will be indented an extra level.
1337    
1338    ----------------------------------------------------------------------
1339    Name: Allen Leung
1340    Date: 2000/04/04 03:18:00 EST
1341    Tag: leunga-20000404-C--Moby
1342    Description:
1343    
1344        All of these fixes are related to C--, Moby, and my own optimization
1345        stuff; so they shouldn't affect SML/NJ.
1346    
1347    1.  X86
1348    
1349        Various fixes related floating point, and extensions.
1350    
1351    2.  Alpha
1352    
1353        Some extra patterns related to loads with signed/zero extension
1354        provided by Fermin.
1355    
1356    3.  Assembly
1357    
1358        When generating assembly, resolve the value of client defined constants,
1359        instead of generating symbolic values.  This is controlled by the
1360        new flag "asm-resolve-constants", which is default to true.
1361    
1362    4.  Machine Descriptions
1363    
1364        a. The precedence parser was slightly broken when parsing infixr symbols.
1365        b. The type generalizing code had the bound variables reversed, resulting
1366           in a problem during arity raising.
1367        c. Various fixes in machine descriptions.
1368    
1369    ----------------------------------------------------------------------
1370    Name: Matthias Blume
1371    Date: 2000/04/03 16:05:00 JST
1372    Tag: blume_main_v110p26p2_2
1373    Description:
1374    
1375    I eliminated coreEnv from compInfo.  Access to the "Core" structure is
1376    now done via the ordinary static environment that is context to each
1377    compilation unit.
1378    
1379    To this end, I arranged that instead of "structure Core" as "structure
1380    _Core" is bound in the pervasive environment.  Core access is done via
1381    _Core (which can never be accidentally rebound because _Core is not a
1382    legal surface-syntax symbol).
1383    
1384    The current solution is much cleaner because the core environment is
1385    now simply part of the pervasive environment which is part of every
1386    compilation unit's context anyway.  In particular, this eliminates all
1387    special-case handling that was necessary until now in order to deal
1388    with dynamic and symbolic parts of the core environment.
1389    
1390    Remaining hackery (to bind the "magic" symbol _Core) is localized in the
1391    compilation manager's bootstrap compiler (actually: in the "init group"
1392    handling).  See the comments in src/system/smlnj/init/init.cmi for
1393    more details.
1394    
1395    I also tried to track down all mentions of "Core" (as string argument
1396    to Symbol.strSymbol) in the compiler and replaced them with a
1397    reference to the new CoreSym.coreSym.  Seems cleaner since the actual
1398    name appears in one place only.
1399    
1400    Binfile and bootfile format have not changed, but the switchover from
1401    the old "init.cmi" to the new one is a bit tricky, so I supplied new
1402    bootfiles anyway.
1403    
1404    ----------------------------------------------------------------------
1405    Name: Allen Leung
1406    Date: 2000/04/02 21:17:00 EST
1407    Tag: leunga-20000402-mltree
1408    Description:
1409    
1410       1. Renamed the constructor CALL in MLTREE by popular demand.
1411       2. Added a bunch of files from my repository.  These are currently
1412          used by other non-SMLNJ backends.
1413    
1414    ----------------------------------------------------------------------
1415    Name: Allen Leung
1416    Date: 2000/03/31 21:15:00 EST
1417    Tag: leunga-20000331-aliasing
1418    Description:
1419    
1420    This update contains a rewritten (and hopefully more correct) module
1421    for extracting aliasing information from CPS.
1422    
1423       To turn on this feature:
1424    
1425            Compiler.Control.CG.memDisambiguate := true
1426    
1427       To pretty print the region information with assembly
1428    
1429           Compiler.Control.MLRISC.getFlag "asm-show-region" := true;
1430    
1431       To control how many levels of aliasing information are printed, use:
1432    
1433           Compiler.Control.MLRISC.getInt "points-to-show-level" := n
1434    
1435       The default of n is 3.
1436    
1437    ----------------------------------------------------------------------
1438    Name: David MacQueen
1439    Date: 2000/03/31 11:15:00 EST
1440    Tag: dbm-20000331-runtime_fix
1441    Description:
1442    
1443    This update contains:
1444    
1445    1. runtime/c-lib/c-libraries.c
1446       includes added in revision 1.2 caused compilation errors on hppa-hpux
1447    
1448    2. fix for bug 1556
1449       system/Basis/Implementation/NJ/internal-signals.sml
1450    
1451    ----------------------------------------------------------------------
1452    Name: Matthias Blume
1453    Date: 2000/03/31 18:00:00 JST
1454    Tag: blume_main_v110p26p2_1
1455    Description:
1456    
1457    This update contains:
1458    
1459    1. A small change to CM's handling of stable libraries:
1460       CM now maintains one "global" modmap that is used for all stable
1461       libraries.  The use of such a global modmap maximizes sharing and
1462       minimizes the need for re-traversing parts of environments during
1463       modmap construction.  (However, this has minor impact since modmap
1464       construction seems to account for just one percent or less of total
1465       compile time.)
1466    
1467    2. I added a "genmap" phase to the statistics.  This is where I got the
1468       "one percent" number (see above).
1469    
1470    3. CM's new tool parameter mechanism just became _even_ better. :)
1471       - The parser understands named parameters and recursive options.
1472       - The "make" and "shell" tools use these new features.
1473         (This makes it a lot easier to cascade these tools.)
1474       - There is a small syntax change: named parameters use a
1475    
1476           <name> : ( <option> ... )            or
1477           <name> : <string>
1478    
1479         syntax.  Previously, named parameters were implemented in an
1480         ad-hoc fashion by each tool individually (by parsing strings)
1481         and had the form
1482    
1483           <name>=<string>
1484    
1485       See the CM manual for a full description of these issues.
1486    
1487    ----------------------------------------------------------------------
1488    Name: Matthias Blume
1489    Date: 2000/03/30 18:00:00 JST
1490    Tag: blume_main_v110p26p2_0
1491    Description:
1492    
1493    !!!!! WARNING !!!!!!
1494    !!  New binfiles  !!
1495    !!!!!!!!!!!!!!!!!!!!
1496    
1497    This update contains:
1498    
1499    1. Moderate changes to CM:
1500    
1501       - Changes to CM's tools mechanism.  In particular, it is now possible
1502       to have tools that accept additional "command line" parameters
1503       (specified in the .cm file at each instance where the tool's class is
1504       used).
1505    
1506       This was done to accommodate the new "make" and "shell" tools which
1507       facilitate fairly seamless hookup to portions of code managed using
1508       Makefiles or Shell scripts.
1509    
1510       There are no classes "shared" or "private" anymore.  Instead, the
1511       sharing annotation is now a parameter to the "sml" class.
1512    
1513       There is a bit of generic machinery for implementing one's own
1514       tools that accept command-line parameters.  However, I am not yet fully
1515       satisfied with that part, so expect changes here in the future.
1516    
1517       All existing tools are described in the CM manual.
1518    
1519       - Slightly better error handling.  (CM now suppresses many followup
1520       error messages that tended to be more annoying than helpful.)
1521    
1522    2. Major changes to the compiler's static environment data structures.
1523    
1524       - no CMStaticEnv anymore.
1525            - no CMEnv, no "BareEnvironment" (actually, _only_ BareEnvironment,
1526              but it is called Environment), no conversions between different
1527              kinds of static environments
1528    
1529       - There is still a notion of a "modmap", but such modmaps are generated
1530         on demand at the time when they are needed.  This sounds slow, but I
1531         sped up the code that generates modmaps enough for this not to lead to
1532         a slowdown of the compiler (at least I didn't detect any).
1533    
1534       - To facilitate rapid modmap generation, static environments now
1535         contain an (optional) "modtree" structure.  Modtree annotations are
1536         constructed by the unpickler during unpickling.  (This means that
1537         the elaborator does not have to worry about modtrees at all.)
1538         Modtrees have the advantage that they are compositional in the same
1539         way as the environment data structure itself is compositional.
1540         As a result, modtrees never hang on to parts of an environment that
1541         has already been rendered "stale" by filtering or rebinding.
1542    
1543       - I went through many, many trials and errors before arriving at the
1544         current solution.  (The initial idea of "linkpaths" did not work.)
1545         But the result of all this is that I have touched a lot of files that
1546         depend on the "modules" and "types" data structures (most of the
1547         elaborator). There were a lot of changes during my "linkpath" trials
1548         that could have been reverted to their original state but weren't.
1549         Please, don't be too harsh on me for messing with this code a bit more
1550         than what was strictly necessary...  (I _did_ resist the tempation
1551         of doing any "global reformatting" to avoid an untimely death at
1552         Dave's hands. :)
1553    
1554       - One positive aspect of the previous point:  At least I made sure that
1555         all files that I touched now compile without warnings (other than
1556         "polyEqual").
1557    
1558       - compiler now tends to run "leaner" (i.e., ties up less memory in
1559         redundant modmaps)
1560    
1561    ----------------------------------------------------------------------
1562    Name: Allen Leung
1563    Date: 2000/03/29 18:00:00
1564    Tag: leunga-20000327-mlriscGen_hppa_alpha_x86
1565    Boot files (optional): ftp://react-ilp.cs.nyu.edu/leunga/110.26.1-sml.boot.x86-unix-20000330.tar.gz
1566    Description:
1567    
1568       This update contains *MAJOR* changes to the way code is generated from CPS
1569    in the module mlriscGen, and in various backend modules.
1570    
1571    CHANGES
1572    =======
1573    
1574    1. MLRiscGen: forward propagation fix.
1575    
1576       There was a bug in forward propagation introduced at about the same time
1577       as the MLRISC x86 backend, which prohibits coalescing to be
1578       performed effectively in loops.
1579    
1580       Effect: speed up of loops in RISC architectures.
1581               By itself, this actually slowed down certain benchmarks on the x86.
1582    
1583    2. MLRiscGen:  forward propagating addresses from consing.
1584    
1585       I've changed the way consing code is generated.  Basically I separated
1586       out the initialization part:
1587    
1588            store tag,   offset(allocptr)
1589            store elem1, offset+4(allocptr)
1590            store elem2, offset+8(allocptr)
1591            ...
1592            store elemn, offset+4n(allocptr)
1593    
1594       and the address computation part:
1595    
1596            celladdr <- offset+4+alloctpr
1597    
1598       and move the address computation part
1599    
1600       Effect:  register pressure is generally lower as a result.  This
1601                makes compilation of certain expressions much faster, such as
1602                long lists with non-trivial elements.
1603    
1604                 [(0,0), (0,0), .... (0,0)]
1605    
1606    3. MLRiscGen: base pointer elimination.
1607    
1608        As part of the linkage mechanism, we generate the sequence:
1609    
1610         L:  ...  <- start of the code fragment
1611    
1612         L1:
1613             base pointer <- linkreg - L1 + L
1614    
1615         The base pointer was then used for computing relocatable addresses
1616       in the code fragment.  Frequently (such as in lots of continuations)
1617       this is not needed.  We now eliminate this sequence whenever possible.
1618    
1619         For compile time efficiency, I'm using a very stupid local heuristic.
1620       But in general, this should be done as a control flow analysis.
1621    
1622       Effect:  Smaller code size.  Speed up of most programs.
1623    
1624    4. Hppa back end
1625    
1626         Long jumps in span dependence resolution used to depend on the existence
1627      of the base pointer.
1628    
1629         A jump to a long label L was expanded into the following sequence:
1630    
1631          LDIL %hi(L-8192), %r29
1632          LDO  %lo(L-8192)(%r29), %r29
1633          ADD  %r29, baseptr, %r29
1634          BV,n %r0(%r29)
1635    
1636         In the presence of change (3) above, this will not work.  I've changed
1637       it so that the following sequence of instructions are generated, which
1638       doesn't mention the base pointer at all:
1639    
1640             BL,n  L', %r29           /* branch and link, L' + 4 -> %r29 */
1641        L':  ADDIL L-(L'+4), %r29     /* Compute address of L */
1642             BV,n  %r0(%r29)          /* Jump */
1643    
1644    5. Alpha back end
1645    
1646          New alpha instructions LDB/LDW have been added, as per Fermin's
1647       suggestions.   This is unrelated to all other changes.
1648    
1649    6. X86 back end
1650    
1651         I've changed andl to testl in the floating point test sequence
1652         whenever appropriate.  The Intel optimization guide states that
1653         testl is preferable to andl.
1654    
1655    7. RA (x86 only)
1656    
1657         I've improved the spill propagation algorithm, using an approximation
1658       of maximal weighted independent sets.   This seems to be necessary to
1659       alleviate the negative effect in light of the slow down in (1).
1660    
1661         I'll write down the algorithm one of these days.
1662    
1663    8. MLRiscGen: frequencies
1664    
1665         I've added an annotation that states that all call gc blocks have zero
1666       execution frequencies.  This improves register allocation on the x86.
1667    
1668    BENCHMARKS
1669    ==========
1670    
1671       I've only perform the comparison on 110.25.
1672    
1673       The platforms are:
1674    
1675        HPPA  A four processor HP machine (E9000) with 5G of memory.
1676        X86   A 300Hhz Pentium II with 128M of memory, and
1677        SPARC An Ultra sparc 2 with 512M of memory.
1678    
1679       I used the following parameters for the SML benchmarks:
1680    
1681                 @SMLalloc
1682         HPPA    256k
1683         SPARC   512k
1684         X86     256k
1685    
1686    COMPILATION TIME
1687    ----------------
1688       Here are the numbers comparing the compilation times of the compilers.
1689       I've only compared 110.25 compiling the new sources versus
1690       a fixpoint version of the new compiler compiling the same.
1691    
1692                     110.25                                  New
1693               Total  Time in RA  Spill+Reload   Total  Time In RA Spill+Reload
1694         HPPA   627s    116s        2684+3584     599s    95s       1003+1879
1695         SPARC  892s    173s        2891+3870     708s    116s      1004+1880
1696         X86    999s    315s       94006+130691   987s    296s    108877+141957
1697    
1698                   110.25         New
1699                Code Size      Code Size
1700         HPPA   8596736         8561421
1701         SPARC  8974299         8785143
1702         X86    9029180         8716783
1703    
1704       So in summary, things are at least as good as before.   Dramatic
1705       reduction in compilation is obtained on the Sparc; I can't explain it,
1706       but it is reproducible.  Perhaps someone should try to reproduce this
1707       on their own machines.
1708    
1709    SML BENCHMARKS
1710    --------------
1711    
1712        On the average, all benchmarks perform at least as well as before.
1713    
1714          HPPA         Compilation Time     Spill+Reload      Run Time
1715                     110.25  New            110.25    New   110.25  New
1716    
1717          barnesHut  3.158  3.015  4.75%    1+1       0+0   2.980  2.922   2.00%
1718              boyer  6.152  5.708  7.77%    0+0       0+0   0.218  0.213   2.34%
1719       count-graphs  1.168  1.120  4.32%    0+0       0+0  22.705 23.073  -1.60%
1720                fft  0.877  0.792 10.74%    1+3       1+3   0.602  0.587   2.56%
1721        knuthBendix  3.180  2.857 11.32%    0+0       0+0   0.675  0.662   2.02%
1722             lexgen  6.190  5.290 17.01%    0+0       0+0   0.913  0.788  15.86%
1723               life  0.803  0.703 14.22%   25+25      0+0   0.153  0.140   9.52%
1724              logic  2.048  2.007  2.08%    6+6       1+1   4.133  4.008   3.12%
1725         mandelbrot  0.077  0.080 -4.17%    0+0       0+0   0.765  0.712   7.49%
1726             mlyacc 22.932 20.937  9.53%  154+181    32+57  0.468  0.430   8.91%
1727            nucleic  5.183  5.060  2.44%    2+2       0+0   0.125  0.120   4.17%
1728      ratio-regions  3.357  3.142  6.84%    0+0       0+0  116.225 113.173 2.70%
1729                ray  1.283  1.290 -0.52%    0+0       0+0   2.887  2.855   1.11%
1730             simple  6.307  6.032  4.56%   28+30      5+7   3.705  3.658   1.28%
1731                tsp  0.888  0.862  3.09%    0+0       0+0   7.040  6.893   2.13%
1732               vliw 24.378 23.455  3.94%  106+127    25+45  2.758  2.707   1.91%
1733      --------------------------------------------------------------------------
1734       Average                     6.12%                                   4.09%
1735    
1736          SPARC        Compilation Time     Spill+Reload      Run Time
1737                     110.25  New            110.25    New   110.25  New
1738    
1739          barnesHut  3.778  3.592  5.20%    2+2       0+0   3.648  3.453    5.65%
1740              boyer  6.632  6.110  8.54%    0+0       0+0   0.258  0.242    6.90%
1741       count-graphs  1.435  1.325  8.30%    0+0       0+0  33.672 34.737   -3.07%
1742                fft  0.980  0.940  4.26%    3+9       2+6   0.838  0.827    1.41%
1743        knuthBendix  3.590  3.138 14.39%    0+0       0+0   0.962  0.967   -0.52%
1744             lexgen  6.593  6.072  8.59%    1+1       0+0   1.077  1.078   -0.15%
1745               life  0.972  0.868 11.90%   26+26      0+0   0.143  0.140    2.38%
1746              logic  2.525  2.387  5.80%    7+7       1+1   5.625  5.158    9.05%
1747         mandelbrot  0.090  0.093 -3.57%    0+0       0+0   0.855  0.728   17.39%
1748             mlyacc 26.732 23.827 12.19%  162+189    32+57  0.550  0.560   -1.79%
1749            nucleic  6.233  6.197  0.59%    3+3       0+0   0.163  0.173   -5.77%
1750      ratio-regions  3.780  3.507  7.79%    0+0       0+0 133.993 131.035   2.26%
1751                ray  1.595  1.550  2.90%    1+1       0+0   3.440  3.418    0.63%
1752             simple  6.972  6.487  7.48%   29+32      5+7   3.523  3.525   -0.05%
1753                tsp  1.115  1.063  4.86%    0+0       0+0   7.393  7.265    1.77%
1754               vliw 27.765 24.818 11.87%  110+135    25+45  2.265  2.135    6.09%
1755      ----------------------------------------------------------------------------
1756       Average                     6.94%                                    2.64%
1757    
1758          X86          Compilation Time     Spill+Reload      Run Time
1759                     110.25  New            110.25    New   110.25  New
1760    
1761          barnesHut  5.530  5.420  2.03%  593+893   597+915   3.532  3.440   2.66%
1762              boyer  8.768  7.747 13.19%  493+199   301+289   0.327  0.297  10.11%
1763       count-graphs  2.040  2.010  1.49%  298+394   315+457  26.578 28.660  -7.26%
1764                fft  1.327  1.302  1.92%  112+209   115+210   1.055  0.962   9.71%
1765        knuthBendix  5.218  5.475 -4.69%  451+598   510+650   0.928  0.932  -0.36%
1766             lexgen  9.970  9.623  3.60% 1014+841  1157+885   0.947  0.928   1.97%
1767               life  1.183  1.183  0.00%  162+182   145+148   0.127  0.103  22.58%
1768              logic  3.285  3.512 -6.45%  514+684   591+836   5.682  5.577   1.88%
1769         mandelbrot  0.147  0.143  2.33%   38+41     33+54    0.703  0.690   1.93%
1770             mlyacc 35.457 32.763  8.22% 3496+4564 3611+4860  0.552  0.550   0.30%
1771            nucleic  7.100  6.888  3.07%  239+168   201+158   0.175  0.173   0.96%
1772      ratio-regions  6.388  6.843 -6.65% 1182+257   981+300  120.142 120.345 -0.17%
1773                ray  2.332  2.338 -0.29%  346+398   402+494   3.593  3.540   1.51%
1774             simple  9.912  9.903  0.08% 1475+941  1579+1168  3.057  3.178  -3.83%
1775                tsp  1.623  1.532  5.98%  266+200   250+211   8.045  7.878   2.12%
1776               vliw 33.947 35.470 -4.29% 2629+2774 2877+3171  2.072  1.890   9.61%
1777      ----------------------------------------------------------------------------
1778       Average                     1.22%                                     3.36%
1779    
1780    ----------------------------------------------------------------------
1781    Name: Allen Leung
1782    Date: 2000/03/23 16:25:00
1783    Tag: leunga-20000323-fix_x86_alpha
1784    Description:
1785    
1786    1. X86 fixes/changes
1787    
1788       a.  The old code generated for SETcc was completely wrong.
1789           The Intel optimization guide is VERY misleading.
1790    
1791    2. ALPHA fixes/changes
1792    
1793       a.  Added the instructions LDBU, LDWU, STB, STW as per Fermin's suggestion.
1794       b.  Added a new mode byteWordLoadStores to the functor parameter to Alpha()
1795       c.  Added reassociation code for address computation.
1796    
1797    ----------------------------------------------------------------------
1798    Name: Allen Leung
1799    Date: 2000/03/22 01:23:00
1800    Tag: leunga-20000322-fix_x86_hppa_ra
1801    Description:
1802    
1803    1. X86 fixes/changes
1804    
1805       a.  x86Rewrite bug with MUL3 (found by Lal)
1806       b.  Added the instructions FSTS, FSTL
1807    
1808    2. PA-RISC fixes/changes
1809    
1810       a.  B label should not be a delay slot candidate!  Why did this work?
1811       b.  ADDT(32, REG(32, r), LI n) now generates one instruction instead of two,
1812           as it should be.
1813       c.  The assembly syntax for fstds and fstdd was wrong.
1814       d.  Added the composite instruction COMICLR/LDO, which is the immediate
1815           operand variant of COMCLR/LDO.
1816    
1817    3. Generic MLRISC
1818    
1819       a.  shuffle.sml rewritten to be slightly more efficient
1820       b.  DIV bug in mltree-simplify fixed (found by Fermin)
1821    
1822    4. Register Allocator
1823    
1824       a.  I now release the interference graph earlier during spilling.
1825           May improve memory usage.
1826    
1827    ----------------------------------------------------------------------
1828    Name: Matthias Blume
1829    Date: 2000/03/14 14:15:32
1830    Tag: blume_main_v110p26p1_2
1831    Description:
1832    
1833    1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
1834    additional argument called "template" which is an optional string that
1835    specifies the layout of the tool command line.  See the CM manual for
1836    explanation.
1837    
1838    2. A special-purpose tool can be "registered" by simply dropping the
1839    corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
1840    directory where the .cm file lives that uses this tool.  (The
1841    behavior/misfeature until now was to look for the tool description
1842    files in the current working directory.)  As before, tool description
1843    files could also be anchored -- in which case they can live anywhere
1844    they like.  Following the recent e-mail discussion, this change should
1845    make it easier to have special-purpose tools that are shipped together
1846    with the sources of the program that uses them.
1847    
1848    ----------------------------------------------------------------------
1849    Name: Matthias Blume
1850    Date: 2000/03/10 07:48:34
1851    Tag: blume_main_v110p26p1_1
1852    Description:
1853    
1854    I added a re-written version of Dave's fixpt script to src/system.
1855    Changes relative to the original version:
1856      - sh-ified (not everybody has ksh)
1857      - automatically figures out which architecture it runs on
1858      - uses ./makeml a bit more cleverly
1859      - never invokes ./installml (and, thus, does not clobber your
1860        good and working installation of sml in case something goes wrong)
1861      - accepts max iteration count using option "-iter <n>"
1862      - accepts a "base" name using option "-base <base>"
1863    
1864    It does not build any extraneous heap images but directly rebuilds
1865    bin- and boot-hierarchies using makeml's "-rebuild" switch. Finally,
1866    it can incorporate existing bin- and boot- hierarchies.  For example,
1867    suppose the base is set to "sml" (which is the default).  Then it
1868    successively builds
1869    
1870            sml.bin.<arch>-unix and sml.boot.<arch>-unix
1871    then    sml1.bin.<arch>-unix and sml1.boot.<arch>-unix
1872    then    sml2.bin.<arch>-unix and sml2.boot.<arch>-unix
1873    ...
1874    then    sml<n>.bin.<arch>-unix and sml<n>.boot.<arch>-unix
1875    
1876    and so on.  If any of these already exist, it will just use what's
1877    there.  In particular, many people will have the initial set of bin
1878    and boot files around, so this saves time for at least one full
1879    rebuild.  Having sets of the form <base><k>.{bin,boot}.<arch>-unix for
1880    <k>=1,2,... is normally not a good idea when invoking fixpt.  However,
1881    they might be the result of an earlier partial run of fixpt (which
1882    perhaps got accidentally killed).  In this case, fixpt will quickly
1883    move through what exists before continuing where it left off earlier,
1884    and, thus, saves a lot of time.
1885    
1886    ----------------------------------------------------------------------
1887    Name: Allen Leung
1888    Date: 00/03/10 02:20:00
1889    Tag: leunga-20000310-fix_x86_asm_ra
1890    Description:
1891    
1892    More assembly output problems involving the indexed addressing mode
1893    on the x86 have been found and corrected. Thanks to Fermin Reig for the
1894    fix.
1895    
1896    The interface and implementation of the register allocator have been changed
1897    slightly to accommodate the possibility to skip the register allocation
1898    phases completely and go directly to memory allocation.  This is needed
1899    for C-- use.
1900    
1901    ----------------------------------------------------------------------
1902    Name: Matthias Blume
1903    Date: 00/03/09 10:23:53
1904    Tag: blume_main_v110p26p1_0
1905    Description:
1906    
1907    * Complete re-organization of library names.  Many libraries have been
1908    consolidated so that they share the same path anchor.  For example,
1909    all MLRISC-related libraries are anchored at MLRISC, most libraries that
1910    are SML/NJ-specific are under "smlnj".  Notice that names like
1911    host-cmb.cm or host-compiler.cm no longer exist.  See system/README
1912    for a complete description of the new naming scheme.  Quick reference:
1913    
1914       host-cmb.cm        -> smlnj/cmb.cm
1915       host-compiler.cm   -> smlnj/compiler.cm
1916       full-cm.cm         -> smlnj/cm.cm
1917       <arch>-<os>.cm     -> smlnj/cmb/<arch>-<os>.cm
1918       <arch>-compiler.cm -> smlnj/compiler/<arch>.cm
1919    
1920    * Bug fixes in CM.
1921        - exceptions in user code are being passed through (i.e., reach top level)
1922        - more bugs in paranoia mode fixed
1923        - bug related to checking group owners fixed
1924    
1925    * New install.sh script that automagically fetches archive files:
1926      The new file config/srcarchiveurl must contain the URL of the
1927      (remote) directory that contains bin files (or other source archives).
1928      If install.sh does not find the archive locally, it tries to get
1929      it from that remote directory.
1930      This should simplify installation further:  For machines that have
1931      access to the internet, just fetch <version>-config.tgz, unpack it,
1932      edit config/targets, and go (run config/install.sh).  The script will
1933      fetch everything else that it might need all by itself.
1934    
1935      For CVS users, this mechanism is not relevant for source archives, but
1936      it is convenient for getting new sets of binfiles.
1937    
1938      Archives should be tar files compressed with either gzip, compress, or
1939      bzip2.  The script recognizes .tgz, .tar, tar.gz, tz, .tar.Z, and .tar.bz2.
1940    
1941    ----------------------------------------------------------------------
1942    Name: Matthias Blume
1943    Date: 2000/03/07 04:01:04
1944    Tag: blume_main_v110_26_2
1945    Description:
1946    - size info in BOOTLIST
1947         * no fixed upper limits for number of bootfiles or length of
1948           bootfile names in runtime
1949         * falling back to old behavior if no BOOTLIST size info found
1950    - allocation size heuristics in .run-sml
1951         * tries to read cache size from /proc/cpuinfo (this is important for
1952            small-cache Celeron systems!)
1953    - install.sh robustified
1954    - CM manual updates
1955    - paranoid mode
1956         * no more CMB.deliver() (i.e., all done by CMB.make())
1957         * can re-use existing sml.boot.* files
1958         * init.cmi now treated as library
1959         * library stamps for consistency checks
1960    - sml.boot.<arch>-<os>/PIDMAP file
1961         * This file is read by the CM startup code.  This is used to minimize
1962           the amount of dynamic state that needs to be stowed away for the
1963           purpose of sharing between interactive system and user code.
1964    - CM.Anchor.anchor instead of CM.Anchor.{set,cancel}
1965         * Upon request by Elsa.  Anchors now controlled by get-set-pair
1966           like most other CM state variables.
1967    - Compiler.CMSA eliminated
1968         * No longer supported by CM anyway.
1969    - fixed bugs in pickler that kept biting Stefan
1970         * past refs to past refs (was caused by the possibility that
1971           ad-hoc sharing is more discriminating than hash-cons sharing)
1972         * integer overflow on LargeInt.minInt
1973    - ml-{lex,yacc} build scripts now use new mechanism
1974      for building standalone programs
1975    - fixed several gcc -Wall warnings that were caused by missing header
1976      files, missing initializations, etc., in runtime (not all warnings
1977      eliminated, though)

Legend:
Removed from v.570  
changed lines
  Added in v.716

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