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

Legend:
Removed from v.570  
changed lines
  Added in v.711

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