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

Legend:
Removed from v.572  
changed lines
  Added in v.715

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