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 605, Fri Apr 7 14:06:42 2000 UTC revision 735, Tue Nov 21 12:15:55 2000 UTC
# Line 11  Line 11 
11  Date:  Date:
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    
15    ----------------------------------------------------------------------
16    Name: Matthias Blume
17    Date: 2000/11/21 21:20:00 JST
18    Tag: blume-20001121-tools
19    Description:
20    
21    Continued hacking on autoloading problem -- with success this time.
22    Also changed tool-plugin mechanism.  See new CM manual.
23    
24    ----------------------------------------------------------------------
25    Name: Matthias Blume
26    Date: 2000/11/19 14:30:00 JST
27    Tag:  blume-20001119-autoload
28    Description:
29    
30    Some hacking to make autoloading faster.  Success for CMB, no success
31    so far for CM.  There is a reduced structure CM' that autoloads faster.
32    (This is a temporary, non-documented hack to be eliminated again when
33    the general problem is solved.)
34    
35    ----------------------------------------------------------------------
36    Name: Matthias Blume
37    Date: 2000/11/17 14:10:00 JST
38    Tag: blume-20001117-pickle-lib
39    Description:
40    
41    1. Eliminated comp-lib.cm
42    2. Made pickle-lib.cm
43    3. Eliminated all uses of intset.sml (from comp-lib.cm)
44    4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
45       equivalent constructs from smlnj-lib.cm (INtHashTable).
46    5. Point 4. also goes for those uses of intmap.* in MLRISC.
47       Duplicated intmap modules thrown out.
48    6. Hunted down all duplicated SCC code and replaced it with
49       equivalent stuff (GraphSCCFn from smlnj-lib.cm).
50    7. Rewrote Feedback module.
51    8. Moved sortedlist.sml into viscomp-lib.cm.  Eventually it
52       should be thrown out and equivalent modules from smlnj-lib.cm
53       should be used (IntRedBlackSet, IntListSet, ...).
54    
55    Confirmed that compiler compiles to fixpoint.
56    
57    ----------------------------------------------------------------------
58    Name: Allen Leung
59    Date: 2000/11/10 18:00:00
60    Tag: leunga-20001110-new-x86-fp
61    
62    A new x86 floating point code generator has been added.
63    By default this is turned off.  To turn this on, do:
64    
65        CM.autoload "$smlnj/compiler.cm";
66        Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
67    
68    Changes:
69    
70    1.  Changed FTAN to FPTAN so that the assembly output is correct.
71    2.  Changed the extension callback for FTANGENT to generate:
72    
73              fptan
74              fstp  %st(0)
75        instead of
76              fptan
77              fstpl ftempmem
78    
79    3.  Numerous assembly fixes for x86.
80    
81    5.  Cleaned up the machine code output module x86/x86MC.sml and added
82        support for a whole bunch of instructions and addressing modes:
83    
84          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st, %st(n)
85          faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp  %st, %st(n)
86          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st(n), %st
87          fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
88          fxch %st(n)
89          fld %st(n)
90          fst %st(n)
91          fst mem
92          fstp %st(n)
93          fucom %st(n)
94          fucomp %st(n)
95    
96        All these are now generated when the fast fp mode is turned on.
97    
98    6.  Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
99    
100    ----------------------------------------------------------------------
101    Name: Matthias Blume
102    Date: 2000/11/09 11:20:00 JST
103    Tag: blume-20001109-scc
104    Description:
105    
106    Eliminated some code duplication:
107    
108    1. Added "where" clause to GraphSCCFn in SML/NJ Library.
109       (Otherwise the functor is useless.)
110    2. Used GraphSCCFn where SCCUtilFun was used previously.
111    3. Got rid of SCCUtilFun (in comp-lib.cm).
112    
113    ----------------------------------------------------------------------
114    Name: Lal George
115    Date: 2000/11/06 09:02:21 EST 2000
116    Tag: Release_110_30
117    Description:
118    
119    - Version 110.30
120    ----------------------------------------------------------------------
121    Name: Matthias Blume
122    Date: 2000/11/04 14:45:00
123    Tag: blume-20001104-mlbuild
124    Description:
125    
126    - Made ml-build faster on startup.
127    - Documentation fixes.
128    
129    ----------------------------------------------------------------------
130    Name: Matthias Blume
131    Date: 2000/11/02 17:00:00 JST
132    Tag: blume-20001102-condcomp
133    Description:
134    
135    - Small tweaks to pickler -- new BOOTFILES!
136    - Version bumped to 110.29.2.
137    - Added conditional compilation facility to init.cmi (see comment there).
138    ----------------------------------------------------------------------
139    Name: Allen Leung
140    Date: 2000/10/23 19:31:00
141    Tag: leunga-20001023-demo-ra
142    
143    1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
144    2. Test programs for the graph library updated
145    3. Some new MLRISC demo programs added
146    
147    ----------------------------------------------------------------------
148    Name: Matthias Blume
149    Date: 2000/08/31 22:15:00 JST
150    Tag: blume-20001017-errmsg
151    Description:
152    
153    More error message grief: Where there used to be no messages, there
154    now were some that had bogus error regions.  Fixed.
155    
156    ----------------------------------------------------------------------
157    Name: Matthias Blume
158    Date: 2000/08/31 17:30:00 JST
159    Tag: blume-20001017-v110p29p1
160    Description:
161    
162    I made a version 110.29.1 with new bootfiles.
163    
164    Changes:  Modified pickler/unpickler for faster and leaner unpickling.
165              CM documentation changes and a small bugfix in CM's error reporting.
166    
167    ----------------------------------------------------------------------
168    Name: Lal George
169    Date: 2000/09/27 14:42:35 EDT
170    Tag: george-20000927-nodestatus
171    Description:
172    
173    Changed the type of the nodestatus, so that:
174    
175            SPILLED(~1)             is now SPILLED
176            SPILLED(m) where m>=0   is now MEMREG(m)
177            SPILLED(s) where s<~1   is now SPILL_LOC(~s)
178    
179    ----------------------------------------------------------------------
180    Name: Matthias Blume
181    Date: 2000/09/07 14:45:00 JST
182    Tag: blume-20000907-cmerrmsg
183    Description:
184    
185    Small tweak to CM to avoid getting ML syntax error messages twice.
186    
187    ----------------------------------------------------------------------
188    Name: Matthias Blume
189    Date: 2000/08/31 18:00:00 JST
190    Tag: blume-20000831-cvsbootfiles
191    Description:
192    
193    New URL for boot files (because the 110.29 files on the BL server do
194    now work correctly with my updated install scripts for yacc and lex).
195    
196    ----------------------------------------------------------------------
197    Name: Matthias Blume
198    Date: 2000/08/08 12:33:00 JST
199    Tag: blume-20000808-manual
200    Description:
201    
202    Tiny update to CM manual.
203    
204    ----------------------------------------------------------------------
205    Name: Allen Leung
206    Date: 2000/08/7 19:31:00
207    Tag: leunga-20000807-a-whole-bunch-of-stuff
208    
209      Moby, C--, SSA, x86, machine descriptions etc.  Should only affect C--
210    and Mobdy.
211    
212    1.  x86
213    
214       a.  Fixes to peephole module by John and Dan.
215       b.  Assembly fix to SETcc by Allen.
216       c.  Fix to c-call by John.
217       d.  Fix to spilling by John.  (This one deals with the missing FSTPT case)
218       e.  Instruction selection optimization to SETcc as suggested by John.
219    
220           For example,
221    
222            MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
223    
224           should generate:
225    
226            MOVL a, x
227            SUBL b, x
228            SHRL 31, x
229    
230    2.  IR stuff
231    
232         A bunch of new DJ-graph related algorithms added.  These
233         speed up SSA construction.
234    
235    3.  SSA + Scheduling
236    
237         Added code for SSA and scheduling to the repository
238    
239    ----------------------------------------------------------------------
240    Name: Lal George
241    Date: 2000/07/27 11:53:14 EDT
242    
243    Tag: lal-20000727-linux-ppc
244    Description:
245    
246     Made changes to support Linux PPC.
247     p.s. I have confirmation that the 110.29 boot files work fine.
248    
249    ----------------------------------------------------------------------
250    Name: Matthias Blume
251    Date: 2000/07/27 17:40:00 JST
252    Tag: blume-20000727-scripts
253    Description:
254    
255    !!!! WARNING !!!!
256    You must recompile the runtime system!
257    !!!! WARNING !!!!
258    
259    This is basically another round of script-enhancements:
260    
261    1. sml, ml-build, and ml-makedepend accept options -D and -U to define
262       and undefine CM preprocessor symbols.
263    
264    2. ml-build avoids generating a new heap image if it finds that the
265       existing one is still ok.  (The condition is that no ML file had to
266       be recompiled and all ML files are found to be older that the heap
267       file.)
268    
269       To make this work smoothly, I also hacked the runtime system as
270       well as SMLofNJ.SysInfo to get access to the heap image suffix
271       (.sparc-solaris, ...) that is currently being used.
272    
273       Moreover, the signature of CM.mk_standalone has changed.  See the
274       CM manual.
275    
276    3. ml-makedepend accepts additional options -n, -a, and -o.  (See the
277       CM manual for details.)
278    
279    4. More CM manual updates:
280        - all of the above has been documented.
281        - there is now a section describing the (CM-related) command line
282          arguments that are accepted by the "sml" command
283    
284    ----------------------------------------------------------------------
285    Name: Matthias Blume
286    Date: 2000/07/25 16:20:00 JST
287    Tag: blume-20000725-makedepend
288    Description:
289    
290    Added a script called ml-makedepend.  This can be used in makefiles
291    for Unix' make in a way very similar to the "makedepend" command for
292    C.
293    
294    The script internally uses function CM.sources.
295    
296    Synopsis:
297    
298        ml-makedepend [-f makefile] cmfile targetname
299    
300    The default for the makefile is "makefile" (or "Makefile" should
301    "makefile" not exist).
302    
303    ml-makedepend adds a cmfile/targetname-specific section to this
304    makefile (after removing the previous version of this section).  The
305    section contains a single dependency specification with targetname on
306    the LHS (targetname is an arbitrary name), and a list of files derived
307    from the cmfile on the RHS.  Some of the files on the RHS are
308    ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
309    "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
310    path names.  The makefile writer is responsible for making sure that
311    these variables have correct at the time "make" is invoked.
312    
313    ----------------------------------------------------------------------
314    Name: Matthias Blume
315    Date: 2000/07/22 23:30:00 JST
316    Tag: blume-20000722-urlupdate
317    Description:
318    
319    Changed BOOT and config/srcarchiveurl to point to BL server:
320    
321        ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
322    
323    ----------------------------------------------------------------------
324    Name: Matthias Blume
325    Date: 2000/07/18 18:00:00 JST
326    Tag: blume-20000718-Version_110_29
327    Description:
328    
329    1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
330    
331    2. Updated config/version to 110.29
332    
333    3. Updated config/srcarchiveurl
334    
335    3. New boot files!
336       ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
337    
338    ----------------------------------------------------------------------
339    Name: Matthias Blume
340    Date: 2000/07/11 13:58:00 JST
341    Tag: blume-20000711-doctypo
342    Description:
343    
344    Fixed a few typos in CM manual.
345    
346    ----------------------------------------------------------------------
347    Name: Allen Leung
348    Date: 2000/06/15 00:38:00
349    Tag: leunga-20000704-sparc-x86
350    
351    1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
352    2. fix to x86 RET bug [found by Dan Grossman]
353    3. sparc assembly bug fix for ticc instructions [found by Fermin]
354    
355       Affects c-- and moby only
356    
357    ----------------------------------------------------------------------
358    Name: Matthias Blume
359    Date: 2000/07/04 15:26:00
360    Tag: blume-20000704-trigger
361    Description:
362    
363    1. Improvements to CM manual.
364    2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
365       of getting a back-trace.  The function, when called, raises an
366       internal exception which explicitly carries the full back-trace history,
367       so it is unaffected by any intervening handle-raise pairs ("trivial"
368       or not).  The interactive loop will print that history once it arrives
369       at top level.
370       Short of having all exceptions implicitly carry the full history, the
371       recommended way of using this facility is:
372         - compile your program with instrumentation "on"
373         - run it, when it raises an exception, look at the history
374         - if the history is "cut off" because of some handler, go and modify
375           your program so that it explicitly calls BTrace.trigger
376         - recompile (still instrumented), and rerun; look at the full history
377    
378    ----------------------------------------------------------------------
379    Name: Matthias Blume
380    Date: 2000/07/03 15:36:00 JST
381    Tag: blume-20000702-manual
382    Description:
383    
384    Small corrections and updates to CM manual.
385    
386    ----------------------------------------------------------------------
387    Name: Matthias Blume
388    Date: 2000/06/29 16:04:00 JST
389    Tag: blume-20000629-yacctool
390    Description:
391    
392    Changes:
393    
394    1. Class "mlyacc" now takes separate arguments to pass options to
395       generated .sml- and .sig-files independently.
396    2. Corresponding CM manual updates.
397    3. BTrace module now also reports call sites.  (However, for loop clusters
398       it only shows from where the cluster was entered.)  There are associated
399       modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
400    
401    ----------------------------------------------------------------------
402    Name: Matthias Blume
403    Date: 2000/06/27 16:51:00 JST
404    Tag: blume-20000627-noweb
405    Description:
406    
407    Changes:
408    
409     1. Implemented "subdir" and "witness" options for noweb tool.
410        This caused some slight internal changes in CM's tool implementation.
411     2. Fixed bug in "tool plugin" mechanism.  This is essentially cleaning
412        some remaining issues from earlier path anchor changes.
413     3. Updated CM manual accordingly.
414    
415     4. Changed implementation of back-tracing so that I now consider it
416        ready for prime-time.
417    
418        In particular, you don't have to explicitly trigger the back-trace
419        anymore.  Instead, if you are running BTrace-instrumented code and
420        there is an uncaught exception (regardless of whether or not it was
421        raised in instrumented code), the top-level evalloop will print
422        the back-trace.
423    
424        Features:
425    
426          - Instrumented and uninstrumented code work together seemlessly.
427            (Of course, uninstrumented code is never mentioned in actual
428             back-traces.)
429    
430          - Asymptotic time- and space-complexity of instrumented code is
431            equal to that of uninstrumented code.  (This means that
432            tail-recursion is preserved by the instrumentation phase.)
433    
434          - Modules whose code has been instrumented in different sessions
435            work together without problem.
436    
437          - There is no penalty whatsoever on uninstrumented code.
438    
439          - There is no penalty on "raise" expressions, even in
440            instrumented code.
441    
442        A potential bug (or perhaps it is a feature, too):
443    
444          A back-trace reaches no further than the outermost instrumented
445          non-trivial "raise".  Here, a "trivial" raise is one that is the
446          sole RHS of a "handle" rule.  Thus, back-traces reach trough
447    
448               <exp> handle e => raise e
449    
450          and even
451    
452               <exp> handle Foo => raise Bar
453    
454          and, of course, through
455    
456               <exp> handle Foo => ...
457    
458         if the exception was not Foo.
459    
460         Back-traces always reach right through any un-instrumented code
461         including any of its "handle" expressions, trivial or not.
462    
463       To try this out, do the following:
464    
465         - Erase all existing binfiles for your program.
466           (You may keep binfiles for those modules where you think you
467            definitely don't need back-tracing.)
468         - Turn on back-trace instrumentation:
469              SMLofNJ.Internals.BTrace.mode (SOME true);
470         - Recompile your program.  (I.e., run "CM.make" or "use".)
471         - You may now turn instrumentation off again (if you want):
472              SMLofNJ.Internals.BTrace.mode (SOME false);
473         - Run your program as usual.  If it raises an exception that
474           reaches the interactive toplevel, then a back-trace will
475           automatically be printed.  After that, the toplevel loop
476           will print the exception history as usual.
477    
478    ----------------------------------------------------------------------
479    Name: Matthias Blume
480    Date: 2000/06/26 09:56:46 JST
481    Tag: blume-20000626-setup
482    Description:
483    
484    CM: - setup-parameter to "sml" added; this can be used to run arbitrary
485          ML code before and after compiling a file (e.g., to set compiler
486          flags)
487    
488    Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
489              - associated changes to btrace.sml (BTrace instrumentation pass)
490              - cleaner implementation of btimp.sml (BTrace tracing and report
491                module)
492    
493    CM manual: * new path encoding documented
494               * description of setup-parameter to "sml" added
495    
496    The biggest user-visible change to back-tracing is that it is no
497    longer necessary to compile all traced modules within the same
498    session.  (This was a real limitation.)
499    
500    ----------------------------------------------------------------------
501    Name: Matthias Blume
502    Date: 2000/06/24 12:40:00 JST
503    Tag: blume-20000624-startup
504    Description:
505    
506    Fixes startup slowdown problem.  (I was calling SrcPath.sync a _tad_
507    bit too often -- to put it mildly. :)
508    
509    ----------------------------------------------------------------------
510    Name: Matthias Blume
511    Date: 2000/06/23 18:20:00 JST
512    Tag: blume-20000623-btrace
513    Description:
514    
515    This updates adds a backtrace facility to aid programmers in debugging
516    their programs.  This involves the following changes:
517    
518    1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
519       keeping track of the current call stack.  When programs are compiled
520       in a special mode, the compiler will insert calls to these hooks
521       into the user program.
522       "Hook" means that it is possible for different implementations of
523       back-tracing to register themselves (at different times).
524    
525    2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
526       as an Absyn.dec->Absyn.dec rewrite.  Normally this phase is turned off.
527       It can be turned on using this call:
528         SMLofNJ.Internals.BTrace.mode (SOME true);
529       Turning it off again:
530         SMLofNJ.Internals.BTrace.mode (SOME false);
531       Querying the current status:
532         SMLofNJ.Internals.BTrace.mode NONE;
533       Annotated programs are about twice as big as normal ones, and they
534       run a factor of 2 to 4 slower with a dummy back-trace plugin (one
535       where all hooks do nothing).  The slowdown with a plugin that is
536       actually useful (such as the one supplied by default) is even greater,
537       but in the case of the default plugin it is still only an constant
538       factor (amortized).
539    
540    3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
541       with a sub-structure BTrace for controlling back-tracing.  In particular,
542       the above-mentioned function "mode" controls whether the annotation
543       phase is invoked by the compiler.  Another important function is
544       "trigger": when called it aborts the current execution and causes
545       the top-level loop to print a full back-trace.
546    
547    4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
548       for back-tracing.  It keeps track of the dynamic call stack and in
549       addition to that it keeps a partial history at each "level" of that
550       stack.  For example, if a tail-calls b, b tail-calls c, and c tail-calls
551       d and b (at separate times, dynamically), then the report will show:
552    
553       GOTO   d
554             /c
555       GOTO  \b
556       CALL   a
557    
558       This shows that there was an initial non-tail call of a, then a
559       tail-call to b or c, looping behavior in a cluster of functions that
560       consist of b and c, and then a goto from that cluster (i.e., either from
561       b or from c) to d.
562    
563       Note that (depending on the user program) the amount of information
564       that the back-trace module has to keep track of at each level is bounded
565       by a constant.  Thus, the whole implementation has the same asymptotical
566       complexity as the original program (both in space and in time).
567    
568    5. compiler/TopLevel/interact/evalloop.sml has been modified to
569       handle the special exception SMLofNJ.Internals.BTrace.BTrace
570       which is raised by the "trigger" function mentioned above.
571    
572    Notes on usage:
573    
574    - Annotated code works well together with unannotated code:
575    Unannotated calls simply do not show up at all in the backtrace.
576    
577    - It is not a good idea to let modules that were annotated during
578    different sessions run at the same time.  This is because the compiler
579    chooses small integers to identify individual functions, and there
580    will be clashes if different modules were compiled in separate sessions.
581    (Nothing will crash, and you will even be told about the clashes, but
582    back-trace information will in general not be useful.)
583    
584    - Back-tracing can be confused by callcc and capture.
585    
586    - The only way of getting a back-trace right now is to explicitly
587    invoke the "trigger" function from your user program.  Eventually, we
588    should make every exception carry back-trace information (if
589    available).  But since this creates more overhead at "raise"-time
590    (similar to the current exnHistory overhead), I have not yet
591    implemented this.  (The implementation will be rather easy.)  With
592    exceptions carrying back-trace information, this facility will be even
593    more useful because users don't need to modify their programs...
594    
595    - While it is possible to compile the compiler with back-trace
596    annotations turned on (I did it to get some confidence in
597    correctness), you must make absolutely sure that core.sml and
598    btimp.sml are compiled WITHOUT annotation!  (core.sml cannot actually
599    be compiled with annotation because there is no core access yet, but
600    if you compile btimp.sml with annotation, then the system will go into
601    an infinite recursion and crash.)
602    Since CM currently does not know about BTrace, the only way to turn
603    annotations on and off for different modules of the compiler is to
604    interrupt CMB.make, change the settings, and re-invoke it.  Of course,
605    this is awkward and clumsy.
606    
607    Sample sessions:
608    
609    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
610    - SMLofNJ.Internals.BTrace.mode (SOME true);
611    [autoloading]
612    [autoloading done]
613    val it = false : bool
614    - structure X = struct
615    -     fun main n = let
616    -         fun a (x, 0) = d x
617    -           | a (x, n) = b (x, n - 1)
618    -         and b (x, n) = c (x, n)
619    -         and c (x, n) = a (x, n)
620    -         and d x = e (x, 3)
621    -         and e (x, 0) = f x
622    -           | e (x, n) = e (x, n - 1)
623    -         and f 0 = SMLofNJ.Internals.BTrace.trigger ()
624    -           | f n = n * g (n - 1)
625    -         and g n = a (n, 3)
626    -     in
627    -         f n
628    -     end
629    - end;
630    structure X : sig val main : int -> int end
631    - X.main 3;
632    *** BACK-TRACE ***
633    GOTO   stdIn:4.2-13.20: X.main[2].f
634    GOTO-( stdIn:4.2-13.20: X.main[2].e
635    GOTO   stdIn:4.2-13.20: X.main[2].d
636         / stdIn:4.2-13.20: X.main[2].a
637         | stdIn:4.2-13.20: X.main[2].b
638    GOTO-\ stdIn:4.2-13.20: X.main[2].c
639    CALL   stdIn:4.2-13.20: X.main[2].g
640    GOTO   stdIn:4.2-13.20: X.main[2].f
641    GOTO-( stdIn:4.2-13.20: X.main[2].e
642    GOTO   stdIn:4.2-13.20: X.main[2].d
643         / stdIn:4.2-13.20: X.main[2].a
644         | stdIn:4.2-13.20: X.main[2].b
645    GOTO-\ stdIn:4.2-13.20: X.main[2].c
646    CALL   stdIn:4.2-13.20: X.main[2].g
647    GOTO   stdIn:4.2-13.20: X.main[2].f
648    GOTO-( stdIn:4.2-13.20: X.main[2].e
649    GOTO   stdIn:4.2-13.20: X.main[2].d
650         / stdIn:4.2-13.20: X.main[2].a
651         | stdIn:4.2-13.20: X.main[2].b
652    GOTO-\ stdIn:4.2-13.20: X.main[2].c
653    CALL   stdIn:4.2-13.20: X.main[2].g
654    GOTO   stdIn:4.2-13.20: X.main[2].f
655    CALL   stdIn:2.15-17.4: X.main[2]
656    -
657    
658    (Note that because of a FLINt bug the above code currently does not
659    compile without BTrace turned on.)
660    
661    Here is another example, using my modified Tiger compiler:
662    
663    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
664    - SMLofNJ.Internals.BTrace.mode (SOME true);
665    [autoloading]
666    [autoloading done]
667    val it = false : bool
668    - CM.make "sources.cm";
669    [autoloading]
670    ...
671    [autoloading done]
672    [scanning sources.cm]
673    [parsing (sources.cm):parse.sml]
674    [creating directory CM/SKEL ...]
675    [parsing (sources.cm):tiger.lex.sml]
676    ...
677    [wrote CM/sparc-unix/semant.sml]
678    [compiling (sources.cm):main.sml]
679    [wrote CM/sparc-unix/main.sml]
680    [New bindings added.]
681    val it = true : bool
682    - Main.compile ("../testcases/merge.tig", "foo.out");
683    *** BACK-TRACE ***
684    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
685    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
686    CALL   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
687    GOTO   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
688    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
689    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
690    CALL   lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
691         / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
692    CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
693    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
694    CALL   lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
695    CALL   main.sml:1.18-118.4: Main.compile[2]
696    -
697    
698    ----------------------------------------------------------------------
699    Name: Matthias Blumen
700    Date: 2000/06/21 18:00:00 JST
701    Tag: blume-20000621-manual
702    Description:
703    
704    CM manual update: Path environments documented.
705    
706    ----------------------------------------------------------------------
707    Name: Matthias Blume
708    Date: 2000/06/19 13:40:00
709    Tag: blume-20000619-manual
710    Description:
711    
712    CM manual and system/README update.  This only covers the fact that
713    there are no more implicit anchors.  (Path environments and the "bind"
714    option to "cm" have yet to be documented.)
715    
716    ----------------------------------------------------------------------
717    Name: Matthias Blume
718    Date: 2000/06/19 11:05:00 JST
719    Tag: blume-20000619-chdir-bugfix
720    Description:
721    
722    Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
723    
724    ----------------------------------------------------------------------
725    Name: Matthias Blume
726    Date: 2000/06/18 22:00:10 JST
727    Tag: blume-20000618-implicit-anchors-really-gone
728    Description:
729    
730    I updates the previous HISTORY entry where I forgot to mention that
731    implicit anchors are no longer with us.
732    
733    The current update also gets rid of the (now useless) controller
734    CM.Control.implicit_anchors.
735    
736    ----------------------------------------------------------------------
737    Name: Matthias Blume
738    Date: 2000/06/16 17:30:00 JST
739    Tag: blume-20000616-anchorenv
740    Description:
741    
742    This patch implements the long anticipated (just kidding :) "anchor
743    environment" mechanism.  In the course of doing this, I also
744    re-implemented CM's internal "SrcPath" module from scratch.  The new
745    one should be more robust in certain boundary cases.  In any case, it
746    is a lot cleaner than its predecessor (IMHO).
747    
748    This time, although there is yet another boot file format change, I
749    kept the unpickler backward-compatible.  As a result, no new bootfiles
750    are necessary and bootstrapping is straightforward.  (You cannot read
751    new bootfiles into an old system, but the other way around is no
752    problem.)
753    
754    Visible changes:
755    
756    ** 0. Implicit path anchors (without the leading $-symbol) are no
757    longer recognized at all. This means that such path names are not
758    illegal either.  For example, the name basis.cm simply refers to a
759    local file called "basis.cm" (i.e, the name is an ordinary path
760    relative to .cm-files directory).  Or, to put it differently, only
761    names that start with $ are anchored paths.
762    
763    ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
764    vanished.
765    
766    John (Reppy) had critizised this as soon as I originally proposed and
767    implemented it, but at that time I did not really deeply believe
768    him. :) Now I came full-circle because I need the $<singlearc> syntax
769    in another place where it cannot be seen as an abbreviation for
770    $/<singlearc>.  To avoid the confusion, $<singlearc> now means what it
771    seems to mean (i.e., it "expands" into the corresponding anchor
772    value).
773    
774    However, when paths are used as members in CM description files, it
775    continues to be true that there must be at least another arc after the
776    anchor.  This is now enforced separately during semantic analysis
777    (i.e., from a lexical/syntactical point of view, the notation is ok.)
778    
779    ** 2. The "cm" class now accepts an option "bind".  The option's value
780    is a sub-option list of precisely two items -- one labeled "anchor"
781    and the other one labeled "value".  As you might expect, "anchor" is
782    used to specify an anchor name to be bound, and "value" specifies what
783    the anchor is being bound to.
784    
785    The value must be a directory name and can be given in either standard
786    syntax (including the possibility that it is itself an anchored path)
787    or native syntax.
788    
789    Examples:
790    
791       foo.cm (bind:(anchor:bar value:$mystuff/bar))
792       lib.cm (bind:(anchor:a value:"H:\\x\\y\\z"))  (* only works under windows *)
793    
794    and so on.
795    
796    The meaning of this is that the .cm-file will be processed with an
797    augmented anchor environment where the given anchor(s) is/are bound to
798    the given values(s).
799    
800    The rationale for having this feature is this: Suppose you are trying
801    to use two different (already stable) libraries a.cm and b.cm (that
802    you perhaps didn't write yourself).  Further, suppose each of these
803    two libraries internally uses its own auxiliary library $aux/lib.cm.
804    Normally you would now have a problem because the anchor "lib" can not
805    be bound to more than one value globally.  Therefore, the project that
806    uses both a.cm and b.cm must locally redirect the anchor to some other
807    place:
808    
809       a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
810       b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
811    
812    This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
813    /usr/lib/smlnj/b-stuff/aux.cm, respectively.
814    
815    Hard-wiring path names is a bit inflexible (and CM will verbosely warn
816    you when you do so at the time of CM.stabilize).  Therefore, you can
817    also use an anchored path as the value:
818    
819      a.cm (bind:(anchor:lib value:$a-lib))
820      b.cm (bind:(anchor:lib value:$b-lib))
821    
822    Now you can globally configure (using the usual CM.Anchor.anchor or
823    pathconfig machinery) bindings for "a-lib" and "b-lib".  Since "lib"
824    itself is always locally bound, setting it globally is no longer
825    meaningful or necessary (but it does not hurt either).  In fact, "lib"
826    can still be used as a global anchor for separate purposes.  As a
827    matter of fact, one can locally define "lib" in terms of a global
828    "lib":
829    
830      a.cm (bind:(anchor:lib value:$lib/a))
831      b.cm (bind:(anchor:lib value:$lib/b))
832    
833    ** 3: The encoding of path names has changed.  This affects the way
834    path names are shown in CM's progress report and also the internal
835    protocol encoding used for parallel make.
836    
837    The encoding now uses one or more ':'-separated segments.  Each
838    segments corresponds to a file that has been specified relative to the
839    file given by its preceding segment.  The first segment is either
840    relative to the CWD, absolute, or anchored.  Each segment itself is
841    basically a Unix pathname; all segments but the first are relative.
842    
843    Example:
844    
845       $foo/bar/baz.cm:a/b/c.sml
846    
847    This path denotes the file bar/a/b/c.sml relative to the directory
848    denoted by anchor "foo".  Notice that the encoding also includes
849    baz.cm which is the .cm-file that listed a/b/c.sml.  As usual, such
850    paths are resolved relative to the .cm-files directory, so baz.cm must
851    be ignored to get the "real" pathname.
852    
853    To make this fact more obvious, CM puts the names of such "virtual
854    arcs" into parentheses when they appear in progress reports. (No
855    parentheses will appear in the internal protocol encoding.)  Thus,
856    what you really see is:
857    
858      $foo/bar/(baz.cm):a/b/c.sml
859    
860    I find this notation to be much more informative than before.
861    
862    Another new feature of the encoding is that special characters
863    including parentheses, colons, (back)slashes, and white space are
864    written as \ddd (where ddd is the decimal encoding of the character).
865    
866    *** The CM manual still needs to be updated.
867    
868    ----------------------------------------------------------------------
869    Name: Allen Leung
870    Date: 2000/06/15 00:38:00
871    Tag: leunga-20000615-x86-peephole
872    
873    x86 Peephole fix by Fermin.  Affects c-- and moby only.
874    
875    ----------------------------------------------------------------------
876    Name: Matthias Blume
877    Date: 2000/06/12 11:40:00
878    Tag: blume-20000612-parmakefix
879    Description:
880    
881    More cleanup after changing the file naming scheme: This time I
882    repaired the parallel make mechanism for CMB.make which I broke earlier.
883    
884    ----------------------------------------------------------------------
885    Name: Allen Leung
886    Date: 2000/06/09 01:25:00
887    Tag: leunga-20000609-various
888    
889    None of these things should affect normal SML/NJ operations
890    
891    1. Peephole improvements provided by Fermin (c--)
892    2. New annotation DEFUSE for adding extra dependence (moby)
893    3. New X86 LOCK instructions (moby)
894    4. New machine description language for reservation tables (scheduling)
895    5. Fixes to various optimization/analysis modules (branch chaining, dominator
896       trees etc.)
897    6. I've changed the CM files so that they can work with versions
898       110.0.6, 110.25 and 110.28
899    
900    ----------------------------------------------------------------------
901    Name: Matthias Blume
902    Date: 2000/06/09 12:40:00
903    Tag: blume-20000609-log
904    Description:
905    
906    - Removed all(?) remaining RCS Log entries from sources.
907    
908    - Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
909      anchored paths).
910    
911    ----------------------------------------------------------------------
912    Name: Matthias Blume
913    Date: 2000/06/07 17:00:00 JST
914    Tag: blume-20000607-no-implicit-anchors
915    Description:
916    
917    1. This update changes the default setting for
918    CM.Control.implicit_anchors from true to false.  This means that
919    implicit anchors are no longer permitted by default.  I also tried to
920    make sure that nothing else still relies on implicit anchors.
921    (This is the next step on the schedule towards a CM that does not even
922    have the notion of implicit anchors anymore.)
923    
924    2. More CM manual updates.
925    
926    3. I managed to track down and fix the pickling bug I mentioned last
927    time.  Because of the previously existing workaround, this entails no
928    immediate practical changes.
929    
930    ----------------------------------------------------------------------
931    Name: Matthias Blume
932    Date: 2000/06/06 11:15:00 JST
933    Tag: blume-20000606-lazierpickle
934    Description:
935    
936    !!!! NEW BOOT FILES !!!!
937    
938    * The main purpose of this update is to make library pickles lazier in
939    order to reduce the initial space penalty for autoloading a library.
940    As a result, it is now possible to have $smlnj/compiler.cm
941    pre-registered.  This should take care of the many complaints or
942    inquiries about missing structure Compiler.  This required changes to
943    CM's internal data structures and small tweaks to some algorithms.
944    
945    As a neat additional effect, it is no longer necessary (for the sake
946    of lean heap image files) to distinguish between a "minimal" CM and a
947    "full" CM.  Now, there is only one CM (i.e., the "full" version:
948    $smlnj/cm.cm aka $smlnj/cm/full.cm), and it is always available at the
949    interactive top level. ($smlnj/cm/minimal.cm is gone.)
950    
951    To make the life of compiler-hackers easier, "makeml" now also
952    pre-registers $smlnj/cmb.cm (aka $smlnj/cmb/current.cm).  In other
953    words, after you bootstrap a new sml for the first time, you will not
954    have to autoload $smlnj/cmb.cm again afterwards.  (The first time
955    around you will still have to do it, though.)
956    
957    * A second change consists of major updates to the CM manual.  There
958    are now several appendices with summary information and also a full
959    specification of the CM description file syntax.
960    
961    * In directory src/system I added the script "allcross".  This script
962    invokes sml and cross-compiles the compiler for all supported
963    architectures.  (Useful when providing a new set of boot files.)
964    
965    * There seems to be a latent bug in my "lazy pickles" mechanism.  I
966    added a small tweak to pickle-util.sml to work around this problem,
967    but it is not a proper fix yet.  I will investigate further.  (The
968    effect of the bug was an inflation of library pickle size.)
969    
970    * Version number increased to 110.28.1 (to avoid compatibility problems).
971    
972    ----------------------------------------------------------------------
973    Name: Allen Leung
974    Date: 2000/05/25 17:28 EDT
975    Tag: leunga-20000525-ra
976    Description:
977    
978      Fixed a bug in freezing phase of the register allocator.
979    
980    ----------------------------------------------------------------------
981    Name: Allen Leung
982    Date: 2000/05/15 22:53 EDT
983    Tag: leunga-20000515-alpha-x86-ra
984    Description:
985    
986      1. Alpha
987    
988          Slight cleanup.  Removed the instruction SGNXL
989    
990      2. X86
991    
992          Added the following instructions to the instruction set:
993    
994            ROLx, RORx,
995            BTx, BTSx, BTLx, BTRx,
996            XCHGx, and variants with the LOCK prefix
997    
998      3. Register Allocation
999    
1000          The module ra-rewrite-with-renaming has been improved.
1001    
1002      These have no effect on SML/NJ.
1003    
1004    ----------------------------------------------------------------------
1005    Name: Matthias Blume
1006    Date: 2000/05/15 16:20:00 JST
1007    Tag: blume-20000515-lightrebuild
1008    Description:
1009    
1010    1. I added an alternative to "-rebuild" to "makeml".  The difference is
1011       that prior to calling CMB.make' the CM-variable "LIGHT" will be
1012       defined.  In effect, the command will not build any cross-compiler
1013       backends and therefore finish more quickly.
1014    
1015       The "fixpt" script also takes a "-light" switch to be able to use
1016       this new facility while compiling for a fixpoint.
1017    
1018    2. I replaced all mentions of anchored paths in group owner specifications
1019       with simple relative paths (usually starting with "..").
1020       The rationale is that a library's internal workings should not be
1021       compromised by the lack of some anchor.  (An anchor is necessary
1022       for someone who wants to refer to the library by an anchored path,
1023       but it should not be necessary to build the same library in the first
1024       place.)
1025    
1026    3. I changed the way CM's tool mechanism determines the shell command
1027       string used for things like ml-yacc etc. so that it does not break
1028       when CM.Control.implicit_anchors is turned off.
1029    
1030    ----------------------------------------------------------------------
1031    Name: Matthias Blume
1032    Date: 2000/05/12 18:20:00 JST
1033    Tag: blume-20000512-ml-build
1034    Description:
1035    
1036    Fixed a bug in config/_ml-build that prevented ml-yacc and ml-lex from
1037    getting installed properly (by config/install.sh).
1038    
1039    ----------------------------------------------------------------------
1040    Name: Matthias Blume
1041    Date: 2000/05/12 17:30:00 JST
1042    Tag: blume-20000512-anchors
1043    Description:
1044    
1045    !!! NEW BOOT FILES !!!
1046    
1047    This change is in preparation of fading out support for "implicitly
1048    anchored path names".  I went through all sources and used the
1049    explicit (and relatively new) $-notation.  See system/README and the
1050    CM manual for more info on this.
1051    
1052    I also modified the anchoring scheme for some things such as "smlnj",
1053    "MLRISC", "cm", etc. to take advantage of the fact that explicit
1054    anchors are more expressive: anchor name and first arc do not have to
1055    coincide.  This entails the following user-visible change:
1056    
1057    You have to write $smlnj/foo/bar instead of smlnj/foo/bar.  In
1058    particular, when you fire up sml with a command-line argument, say,
1059    e.g.:
1060    
1061       sml '$smlnj/cmb.cm'
1062    
1063    At the ML toplevel prompt:
1064    
1065       CM.autoload "$smlnj/cmb.cm";
1066    
1067    There is also a new controller in CM.Control that can be used to turn
1068    off all remaining support for implicit anchors by saying:
1069    
1070        CM.autoload "$smlnj/
1071        #set CM.Control.implicit_anchors false;
1072    
1073    This causes CM to reject implicitly anchored paths.  This is (for the
1074    time being) less permissive than the "final" version where there will
1075    be no more such implicit anchors and relative paths will be just that:
1076    relative.
1077    
1078    The next step (version after next version?) will be to make the
1079    default for CM.Control.implicit_anchors false.  After the dust has
1080    settled, I can then produce the "final" version of this...
1081    
1082    Note: Since bootstrapping is a bit tricky, I provided new boot files.
1083    
1084    ----------------------------------------------------------------------
1085    Name: Matthias Blume
1086    Date: 2000/05/11 16:30:00 JST
1087    Tag: blume-20000511-sources
1088    Description:
1089    
1090    The main change is that I added function CM.sources as a generalized
1091    version of the earlier CM.makedepend.  This entails the following
1092    additional changes:
1093    
1094      - CM.makedepend has been dropped.
1095    
1096      - CM manual has been updated.
1097    
1098      - TOOLS signature and API have been changed.
1099    
1100    ----------------------------------------------------------------------
1101    Name: Allen Leung
1102    Date: 2000/05/10 21:17 EDT
1103    Tag: leunga-20000510-moby-c--ssa
1104    Description:
1105    
1106      Various bug fixes and new features for C--, Moby and MLRISC optimizations.
1107    None of these affect SML/NJ.
1108    
1109    1. Register Allocation
1110    
1111        a. A new ra spilling module (ra/ra-spill-with-renaming) is implemented.
1112           This module tries to remove local (i.e. basic block level) redundancies
1113           during spilling.
1114    
1115        b. A new framework for performing region based register allocation.
1116           Not yet entirely functional.
1117    
1118    2. X86
1119    
1120       a. DefUse for POP was missing the stack pointer [found by Lal]
1121       b. Reload for CALL was incorrect in X86Spill [found by John]
1122       c. Various fixes in X86Spill so that it can be used correctly for
1123          the new spilling module.
1124    
1125    3. SSA/IR
1126    
1127       a. New module ir/dj-dataflow.sml implements elimination based
1128          data flow analysis.
1129    
1130    4. MLRiscGen
1131    
1132       a. Fix for gc type annotation
1133    
1134    5. MDGen
1135    
1136       Various fixes for machine description -> ml code translation.  For ssa
1137       only.
1138    
1139    ----------------------------------------------------------------------
1140    Name: Allen Leung
1141    Date: 2000/05/08 22:17 EDT
1142    Tag: leunga-20000508-labexp
1143    Description:
1144    
1145      Fermin has found a few assembly problems with constant expressions
1146      generated in LabelExp.  Mostly, the problems involve extra parentheses,
1147      which choke on dumb assemblers.  This is his fix.
1148    
1149    ----------------------------------------------------------------------
1150    Name: Dave MacQueen
1151    Date: 2000/04/09 14:00 EDT
1152    Tag: dbm-20000502-Version_110_28
1153    Description:
1154    
1155    1. Updated src/compiler/TopLevel/main/version.sml to version 110.28
1156    
1157    2. Updated config/version to 110.28
1158    
1159    3. Updated config/srcarchiveurl
1160    
1161    3. New boot files!
1162       ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.28/
1163    
1164    ----------------------------------------------------------------------
1165    Name: Matthias Blume
1166    Date: 2000/05/01 19:05:00 JST
1167    Tag: blume-20000501-noweb
1168    Description:
1169    
1170    A new noweb tool has been added.  The existing system is entirely
1171    unaffected by this, but some CM users have asked for renewed noweb
1172    support.  Everything is documented in the CM manual.
1173    
1174    New (plugin) libraries:
1175    
1176       noweb-tool.cm
1177       nw-ext.cm
1178    
1179    ----------------------------------------------------------------------
1180    Name: Dave MacQueen
1181    Date: 2000/04/30 12:40PM EDT
1182    Tag: dbm-20000430-bug_fixes
1183    Description:
1184    
1185    1. Fix for bug 1498
1186       smlnj/src/system/Basis/Implementation/Unsafe/object.sig
1187       smlnj/src/system/Basis/Implementation/Unsafe/object.sml
1188         added toRealArray function
1189       smlnj/src/compiler/MiscUtil/print/ppobj.sml
1190         added check for tag Obj.RealArray to array printing case in ppObj
1191    
1192    2. Fix for bug 1510
1193       smlnj/src/compiler/Semant/types/typesutil.sml
1194         fixed definition of dummyargs (used by equalTycon) so that
1195         dummy args are distinct types
1196    
1197    ----------------------------------------------------------------------
1198    Name: Matthias Blume
1199    Date: 2000/04/30 01:00:00 JST
1200    Tag: blume-20000430-versions
1201    Description:
1202    
1203    1. CM version numbering added.  This is an implementation of Lal's
1204       proposal for adding version numbers and version checking to .cm
1205       files.  Lal said that his proposal was just that -- a proposal.
1206       For the time being I went ahead and implemented it so that people
1207       can comment on it.  Everything is completely backward-compatible
1208       (except for the stable library format, i.e., new bootfiles!).
1209    
1210       As usual, see the CM manual for details.
1211    
1212    2. An alternative syntax for anchored paths has been implemented.
1213       Dave has recently voiced the same concerns that I had when I did
1214       this, so there should be some support.  My take is that eventually
1215       I will let support for the current syntax (where anchors are
1216       "implicit") fade out in favor of the new, explicit syntax.
1217       In order to be backward-compatible, both old and new syntax are
1218       currently supported.
1219    
1220       Again, see the CM manual for details.
1221    
1222    3. Parallel make is trying to be slightly smarter:  When the master
1223       process finds a "bottleneck", i.e., when there is only one
1224       compilation unit that can be compiled and everybody else is
1225       waiting on it, then it will simply compile it directly instead
1226       of clumsily telling one of the slaves to do it.
1227    
1228    4. Support for "unsharing" added.  This is necessary in order to be
1229       able to have two different versions of the same library running
1230       at the same time (e.g., for trying out a new MLRISC while still
1231       having the old MLRISC linked into the current compiler, etc.)
1232       See the CM manual.
1233    
1234    5. Simple "makedepend" functionality added for generating Makefile
1235       dependency information.  (This is rather crude at the moment.
1236       Expect some changes here in the future.)
1237    
1238    6. ".fun" added as a recognized suffix for ML files. Also documented
1239       explicitly in the manual that the fallback behavior (unknown suffix
1240       -> ML file) is not an official feature!
1241    
1242    7. Small changes to the pickler for stable libraries.
1243    
1244    8. Several internal changes to CM (for cleanup/improvement).
1245    
1246    
1247    !!!! NEW BINFILES !!!!
1248    
1249    ----------------------------------------------------------------------
1250    Name: Matthias Blume
1251    Date: 2000/04/28 17:30:00 JST
1252    Tag: blume-20000428-pathconfig
1253    Description:
1254    
1255    1. I changed config/install.sh to remove duplicate entries from the
1256       lib/pathconfig file at the end.  Moreover, the final version of
1257       lib/pathconfig is sorted alphabetically.  The same (sorting) is done
1258       in src/system/installml.
1259    
1260    2. The config/install.sh script now consistently uses relative
1261       pathnames in lib/pathconfig whenever the anchor is in the lib
1262       directory.  (So far this was true for the libraries that come
1263       pre-compiled and bundled as part of the bootfiles but not for
1264       libraries that are compiled by the script itself.)
1265    
1266    ----------------------------------------------------------------------
1267    Name: Matthias Blume
1268    Date: 2000/04/26 13:10:00 JST
1269    Tag: blume-20000426-fun_suffix
1270    Description:
1271    
1272    Added ".fun" as a recognized file name suffix (for ML code).
1273    
1274    ----------------------------------------------------------------------
1275    Name: Allen Leung
1276    Date: 2000/04/25 17:00:00 EST
1277    Tag: leunga-20000425-alpha-ra
1278    Description:
1279    
1280    1. Alpha
1281    
1282        PSEUDOARITH was missing in AlphaRewrite.  This causes an endless loop
1283    in C--.
1284    
1285    2. RA
1286    
1287       Added a flag "ra-dump-size" to print out the size of the flowgraph
1288       and the interference graph.
1289    
1290    ----------------------------------------------------------------------
1291    Name: Dave MacQueen
1292    Date: 2000/04/25/
1293    Tag: dbm-20000425-mlyacc_doc_examples
1294    Description:
1295      Updated mlyacc.tex sections 5 and 7 for SML '97 and CM.
1296      Updated all three examples in src/ml-yacc/examples to run
1297      under 110.* using CM.make.
1298    
1299    ----------------------------------------------------------------------
1300    Name: Allen Leung
1301    Date: 2000/04/20 23:04:00 EST
1302    Tag: leunga-20000420-ssa-c---stuff
1303    Description:
1304    
1305      This update synchronizes my repository with Yale's.  Most of these
1306    changes, however, do not affect SML/NJ at all (the RA is an exception).
1307    
1308    1. Register Allocator
1309    
1310       a. An improvement in the interference graph construction:
1311          Given a copy
1312    
1313                s <- t
1314    
1315          no interference edge between s and t is added for this definition of s.
1316    
1317       b. I've added two new spill heuristic modules that Fermin and I developed
1318          (in the new library RA.cm). These are unused in SML/NJ but maybe
1319          useful for others (Moby?)
1320    
1321    2. X86
1322    
1323       a. Various fixes in the backend provided by Fermin [C--] and Lal.
1324    
1325    3. Alpha
1326    
1327       a. Added the BSR instruction and code generation that goes with it [C--]
1328       b. Other fixes too numerous to recount provided by Fermin [C--]
1329    
1330    4. Regmaps
1331    
1332       a. The regmaps are not initialized with the identity physical bindings
1333          at creation time.  This is unneeded.
1334    
1335    5. MLRISC Optimizations
1336    
1337       a. The DJ-Graph module can now compute the iterated dominance frontiers
1338          intersects with liveness incrementally in linear time! Woohoo!
1339          This is now used in my new SSA construction algorithm.
1340    
1341       b. THe branch reorganization module is now smarter about linear chains of
1342          basic blocks.
1343    
1344    
1345    ----------------------------------------------------------------------
1346    Name: Matthias Blume
1347    Date: 2000/04/12 13:52:00 JST
1348    Tag: blume_main_v110p27_1
1349    Description:
1350    
1351    Changed install.sh script to handle archive files without version number
1352    and to use "boot.<arch>-<os>" instead of "sml.boot.<arch>-<os>" for the
1353    name of the boot file archive.
1354    
1355    ----------------------------------------------------------------------
1356    Name: Dave MacQueen
1357    Date: 2000/04/09 14:00 EDT
1358    Tag: dbm-20000410-Version_110_27
1359    Description:
1360    
1361    1. Updated src/compiler/TopLevel/main/version.sml to version 110.27
1362    
1363    2. Updated src/config/version to 110.27
1364    
1365    3. New boot files!
1366    
1367    ----------------------------------------------------------------------
1368    Name: Allen Leung
1369    Date: 2000/04/09 19:09:00 EST
1370    Tag: leunga-20000409-misc
1371    Description:
1372    
1373    1.  Yet another fix for x86 assembly for idivl, imull, mull and friends.
1374    
1375    2.  Miscellaneous improvements to MLRISC (unused in sml/nj)
1376    
1377  ----------------------------------------------------------------------  ----------------------------------------------------------------------
1378  Name: Stefan  Name: Stefan
1379  Date: 2000/04/07 10:00:00 EDT  Date: 2000/04/07 10:00:00 EDT
# Line 97  Line 1460 
1460    
1461  3.  Assembly  3.  Assembly
1462    
1463      When generating assemby, resolve the value of client defined constants,      When generating assembly, resolve the value of client defined constants,
1464      instead of generating symbolic values.  This is controlled by the      instead of generating symbolic values.  This is controlled by the
1465      new flag "asm-resolve-constants", which is default to true.      new flag "asm-resolve-constants", which is default to true.
1466    
# Line 120  Line 1483 
1483    
1484  To this end, I arranged that instead of "structure Core" as "structure  To this end, I arranged that instead of "structure Core" as "structure
1485  _Core" is bound in the pervasive environment.  Core access is done via  _Core" is bound in the pervasive environment.  Core access is done via
1486  _Core (which can never be accidentially rebound because _Core is not a  _Core (which can never be accidentally rebound because _Core is not a
1487  legal surface-syntax symbol).  legal surface-syntax symbol).
1488    
1489  The current solution is much cleaner because the core environment is  The current solution is much cleaner because the core environment is
# Line 130  Line 1493 
1493  with dynamic and symbolic parts of the core environment.  with dynamic and symbolic parts of the core environment.
1494    
1495  Remaining hackery (to bind the "magic" symbol _Core) is localized in the  Remaining hackery (to bind the "magic" symbol _Core) is localized in the
1496  compilation mananger's bootstrap compiler (actually: in the "init group"  compilation manager's bootstrap compiler (actually: in the "init group"
1497  handling).  See the comments in src/system/smlnj/init/init.cmi for  handling).  See the comments in src/system/smlnj/init/init.cmi for
1498  more details.  more details.
1499    
# Line 245  Line 1608 
1608     (specified in the .cm file at each instance where the tool's class is     (specified in the .cm file at each instance where the tool's class is
1609     used).     used).
1610    
1611     This was done to accomodate the new "make" and "shell" tools which     This was done to accommodate the new "make" and "shell" tools which
1612     facilitate fairly seemless hookup to portions of code managed using     facilitate fairly seamless hookup to portions of code managed using
1613     Makefiles or Shell scripts.     Makefiles or Shell scripts.
1614    
1615     There are no classes "shared" or "private" anymore.  Instead, the     There are no classes "shared" or "private" anymore.  Instead, the
# Line 258  Line 1621 
1621    
1622     All existing tools are described in the CM manual.     All existing tools are described in the CM manual.
1623    
1624     - Slightly better error handling.  (CM now surpresses many followup     - Slightly better error handling.  (CM now suppresses many followup
1625     error messages that tended to be more annoying than helpful.)     error messages that tended to be more annoying than helpful.)
1626    
1627  2. Major changes to the compiler's static environment data structures.  2. Major changes to the compiler's static environment data structures.
# Line 392  Line 1755 
1755    
1756       I've changed andl to testl in the floating point test sequence       I've changed andl to testl in the floating point test sequence
1757       whenever appropriate.  The Intel optimization guide states that       whenever appropriate.  The Intel optimization guide states that
1758       testl is perferable to andl.       testl is preferable to andl.
1759    
1760  7. RA (x86 only)  7. RA (x86 only)
1761    
# Line 574  Line 1937 
1937    
1938  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
1939  additional argument called "template" which is an optional string that  additional argument called "template" which is an optional string that
1940  specifiel the layout of the tool command line.  See the CM manual for  specifies the layout of the tool command line.  See the CM manual for
1941  explanation.  explanation.
1942    
1943  2. A special-purpose tool can be "regisitered" by simply dropping the  2. A special-purpose tool can be "registered" by simply dropping the
1944  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
1945  directory where the .cm file lives that uses this tool.  (The  directory where the .cm file lives that uses this tool.  (The
1946  behavior/misfeature until now was to look for the tool description  behavior/misfeature until now was to look for the tool description
# Line 621  Line 1984 
1984  rebuild.  Having sets of the form <base><k>.{bin,boot}.<arch>-unix for  rebuild.  Having sets of the form <base><k>.{bin,boot}.<arch>-unix for
1985  <k>=1,2,... is normally not a good idea when invoking fixpt.  However,  <k>=1,2,... is normally not a good idea when invoking fixpt.  However,
1986  they might be the result of an earlier partial run of fixpt (which  they might be the result of an earlier partial run of fixpt (which
1987  perhaps got accidentially killed).  In this case, fixpt will quickly  perhaps got accidentally killed).  In this case, fixpt will quickly
1988  move through what exists before continuing where it left off earlier,  move through what exists before continuing where it left off earlier,
1989  and, thus, saves a lot of time.  and, thus, saves a lot of time.
1990    
# Line 671  Line 2034 
2034    it from that remote directory.    it from that remote directory.
2035    This should simplify installation further:  For machines that have    This should simplify installation further:  For machines that have
2036    access to the internet, just fetch <version>-config.tgz, unpack it,    access to the internet, just fetch <version>-config.tgz, unpack it,
2037    edit config/targets, and go (run config/install.sh).  The scipt will    edit config/targets, and go (run config/install.sh).  The script will
2038    fetch everything else that it might need all by itself.    fetch everything else that it might need all by itself.
2039    
2040    For CVS users, this mechanism is not relevant for source archives, but    For CVS users, this mechanism is not relevant for source archives, but

Legend:
Removed from v.605  
changed lines
  Added in v.735

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