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 601, Thu Apr 6 04:38:14 2000 UTC revision 796, Tue Mar 6 00:04:33 2001 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: Allen Leung
17    Date: Mon Mar  5 18:54:57 EST 2001
18    Tag: leunga-20010305-cut-support
19    
20    1. New support for alternative control-flow in MLTREE.
21       Currently we support
22    
23          FLOW_TO(CALL ...., [k1,...,kn])
24    
25       This is needed for 'cuts to' in C-- and try/handle-like constructs
26       in Moby
27    
28       New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
29    
30    2. Register Allocator
31    
32       Changes in interface [from Fermin, John]
33    
34    3. Alpha 8-bit SLL support [Fermin]
35    
36    4. All architectures
37    
38       A new module (ClusterExpandCopies) for expanding parallel copies.
39    
40    ----------------------------------------------------------------------
41    Name: Allen Leung
42    Date: 2001/02/27 23:07:00 EST
43    Tag: leunga-20010227-minor-stuff
44    
45    1. Alpha bug fix for CMOVNE
46    2. Handle mltree COND(..,FCMP ...,...)
47    3. Bug fix in simplifier
48    
49    ----------------------------------------------------------------------
50    Name: Matthias Blume
51    Date: 2001/01/30 17:50:00 EST
52    Tag: blume-20010130-sync
53    Description:
54    
55    This is just a minor update to sync my devel branch with the main brach.
56    The only visible change is the addition of some README files.
57    
58    ----------------------------------------------------------------------
59    Name: Matthias Blume
60    Date: 2001/01/12 23:30:00 JST
61    Tag: blume-20010112-bootfiles
62    Description:
63    
64    Made a new set of bootfiles that goes with the current state of the
65    repository.
66    
67    ----------------------------------------------------------------------
68    Name: Matthias Blume
69    Date: 2001/01/12 21:20:00 JST
70    Tag: blume-20010112-sync
71    Description:
72    
73    I am just flushing out some minor changes that had accumulated in
74    my private branch in order to sync with the main tree.  (This is
75    mainly because I had CVS trouble when trying to merge _into_ my
76    private branch.)
77    
78    Most people should be completely unaffected by this.
79    
80    ----------------------------------------------------------------------
81    Name: Allen Leung
82    Date: Thu Jan 11 21:03:00 EST 2001
83    Tag: leunga-20010111-labexp=mltree
84    Description:
85    
86    1.  Removed the type LabelExp and replace it by MLTree.
87    2.  Rewritten mltree-simplify with the pattern matcher tool.
88    3.  There were some bugs in alpha code generator which would break
89        64-bit code generation.
90    4.  Redo the tools to generate code with the
91    5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
92        are now generated by perl scripts.
93    
94    ----------------------------------------------------------------------
95    Name: Matthias Blume
96    Date: 2001/01/10 21:55:00 JST
97    Tag: blume-20010110-rcc
98    Description:
99    
100    The RCC stuff now seems to work (but only on the x86).
101    This required hacking of the c-calls interface (and -implementation) in
102    MLRISC.
103    
104    Normal compiler users should be unaffected.
105    
106    ----------------------------------------------------------------------
107    Name: Matthias Blume
108    Date: 2001/01/09 01:20:00 JST
109    Tag: blume-20010109-rcc
110    Description:
111    
112    This is a fairly big patch, flushing out a large number of pending
113    changes that I made to my development copy over the last couple of days.
114    
115    Of practical relevance at this moment is a workaround for a pickling
116    bug that Allen ran into the other day.  The cause of the bug itself is
117    still unknown and it might be hard to fix it properly, but the
118    workaround has some merits of its own (namely somewhat reducing pickling
119    overhead for certain libraries).  Therefore, I think this solution should
120    be satisfactory at this time.
121    
122    The rest of the changes (i.e., the vast majority) has to do with my
123    ongoing efforts of providing direct support for C function calls from
124    ML.  At the moment there is a new primop "RAW_CCALL", typing magic
125    in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
126    case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
127    translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
128    handle RCC.
129    
130    The last part (the changes to mlriscGen.sml) are still known to be
131    wrong on the x86 and not implemented on all other architectures.  But
132    the infrastructure is in place. I had to change a few functor
133    signatures in the backend to be able to route the CCalls interface
134    from MLRISC there, and I had to specialize the mltree type (on the
135    x86) to include the necessary extensions. (The extensions themselves
136    were already there and redy to go in MLRISC/x86).
137    
138    Everything should be very happy as soon as someone helps me with
139    mlriscGen.sml...
140    
141    In any case, nothing of this should matter to anyone as long as the
142    new primop is not being used (which is going to be the case unless you
143    find it where I hid it :). The rest of the compiler is completely
144    unaffected.
145    
146    ----------------------------------------------------------------------
147    Name: Matthias Blume
148    Date: 2001/01/05 00:30:00 JST
149    Tag: blume-20010105-primops
150    Description:
151    
152    Added some experimental support for work that I am doing right now.
153    These changes mostly concern added primops, but there is also a new
154    experimental C library in the runtime system (but currently not enabled
155    anywhere except on Linux/X86).
156    
157    In the course of adding primops (and playing with them), I discovered that
158    Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
159    broken.  (Zhong was very right he labeled this stuff as "major gross hack".)
160    To recover, I made type information in INL_PRIM mandatory and changed
161    prim.sml as well as built-in.sml accordingly.  The InLine structure now
162    has complete, correct type information (i.e., no bottom types).
163    
164    Since all these changes mean that we need new binfiles, I also bumped the
165    version number to 110.32.1.
166    
167    ----------------------------------------------------------------------
168    Name: Matthias Blume
169    Date: 2000/12/30 22:10:00 JST
170    Tag: blume-20001230-various
171    Description:
172    
173    Added proxy libraries for MLRISC and let MLRISC libraries refer
174    to each other using path anchors.  (See CM manual for explanation.)
175    
176    Updated CM documentation.
177    
178    Fixed some bugs in CM.
179    
180    Implemented "proxy" libraries (= syntactic sugar for CM).
181    
182    Added "-quiet" option to makeml and changed runtime system accordingly.
183    
184    Added cleanup handler for exportML to reset timers and compiler stats.
185    
186    ----------------------------------------------------------------------
187    Name: Lal George
188    Date: 2000/12/22 22:22:58 EST 2000
189    Tag: Release_110_32
190    Description:
191    
192            Infinite precision used throughout MLRISC.
193            see MLRISC/mltree/machine-int.sig
194    
195    ----------------------------------------------------------------------
196    Name: Matthias Blume
197    Date: 2000/12/22 23:16:00 JST
198    Tag: blume-20001222-warn
199    Description:
200    
201    Corrected wording and formatting of some CM warning message which I
202    broke in my previous patch.
203    
204    ----------------------------------------------------------------------
205    Name: Matthias Blume
206    Date: 2000/12/22 21:20:00 JST
207    Tag: blume-20001222-anchorenv
208    Description:
209    
210    Fixed CM's handling of anchor environments in connection with CMB.make.
211    
212    ----------------------------------------------------------------------
213    Name: Matthias Blume
214    Date: 2000/12/22 13:15:00 JST
215    Tag: blume-20001222-cleanup
216    Description:
217    
218    Removed src/cm/ffi which does not (and did not) belong here.
219    
220    ----------------------------------------------------------------------
221    Name: Matthias Blume
222    Date: 2000/12/21 23:55:00 JST
223    Tag: blume-20001221-exn
224    Description:
225    
226    Probably most important: CM no longer silently swallows all exceptions
227    in the compiler.
228    Plus: some other minor CM changes.  For example, CM now reports some
229    sizes for generated binfiles (code, data, envpickle, lambdapickle).
230    
231    ----------------------------------------------------------------------
232    Name: Matthias Blume
233    Date: 2000/12/15 00:01:05 JST
234    Tag: blume-20001215-dirtool
235    Description:
236    
237    - "dir" tool added.
238    - improvements and cleanup to Tools structure
239    - documentation updates
240    
241    ----------------------------------------------------------------------
242    Name: Allen Leung
243    Date: Thu Dec 14 03:45:24 EST 2000
244    Description:
245    Tag:  leunga-20001214-int-inf
246    Description:
247    
248       In IntInf, added these standard functions, which are missing from our
249    implementation:
250    
251        andb : int * int -> int
252        xorb : int * int -> int
253        orb  : int * int -> int
254        notb : int -> int
255         <<   : int * word -> int
256        ~>>  : int * word -> int
257    
258       Not tested, I hope they are correct.
259    
260    ----------------------------------------------------------------------
261    Name: Allen Leung
262    Date: Fri Dec  8 19:23:26 EST 2000
263    Description:
264    Tag:  leunga-20001208-nowhere
265    Description:
266    
267      Slight improvements to the 'nowhere' tool to handle OR-patterns,
268    to generate better error messages etc.  Plus a brief manual.
269    
270    ----------------------------------------------------------------------
271    Name: Lal George
272    Date: 2000/12/08 09:54:02 EST 2000
273    Tag: Release_110_31
274    Description:
275    
276    - Version 110.31
277    ----------------------------------------------------------------------
278    Name: Allen Leung
279    Date: Thu Dec  7 22:01:04 EST 2000
280    Tag:  leunga-20001207-cell-monster-hack
281    Description:
282    
283    Major MLRISC internal changes.  Affect all clients.
284    Summary:
285    
286    1.  Type CELLS.cell = int is now replaced by a datatype.
287        As a result, the old regmap is now gone.  Almost all interfaces
288        in MLRISC change as a consequence.
289    
290    2.  A new brand version of machine description tool (v3.0) that generates
291        modules expecting the new interface.  The old version is removed.
292    
293    3.  The RA interface has been further abstracted into two new functors.
294        RISC_RA and X86RA.  These functors have much simpler interfaces.
295        [See also directory MLRISC/demo.]
296    
297    4.  Some other new source->source code generation tools are available:
298    
299        a. MLRISC/Tools/RewriteGen -- generate rewriters from rules.
300        b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules.
301           I use this tool to generate the peephole optimizers---with the new
302           cell type changes, peephole rules are becoming difficult to write
303           without conditional pattern matching.
304    
305    5.  More Intmap -> IntHashTable change.  Previous changes by Matthias didn't
306        cover the entire MLRISC source tree so many things broke.
307    
308    6.  CM files have been moved to the subdirectory MLRISC/cm.
309        They are moved because there are a lot of them and they clutter up the
310        root dir.
311    
312    7.  More detailed documentation to come...
313    
314        NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do
315        a makeml -rebuild first.  This is because of other other
316        changes that Matthias has made (see below).
317    
318    
319    ----------------------------------------------------------------------
320    Name: Matthias Blume
321    Date: 2000/11/30 23:12:00 JST
322    Tag: blume-20001130-filereorg
323    Description:
324    
325    Some manual updates and some file reorganizations in CM.
326    
327    ----------------------------------------------------------------------
328    Name: Matthias Blume
329    Date: 2000/11/24 17:45:00 JST
330    Tag: blume-20001124-link
331    Description:
332    
333    Drastically improved link traversal code for the case that the dynamic
334    value was already loaded at bootstrap time.  As a result, CM and CMB
335    now both load blazingly fast -- even on a very slow machine.  Also,
336    memory consumption has been further reduced by this.
337    
338    Warning: The format of the PIDMAP file has changed.  THerefore, to
339    bootstrap you have to do this:
340    
341    1. Run CMB.make
342    2. Make a symbolic link for the boot directory:
343         ln -s sml.boot.ARCH-OS xxx
344    3. "Rebuild" the boot directory:
345         ./makeml -boot xxx -rebuild sml ; rm xxx
346    4. Boot normally:
347          ./makeml
348    
349    ----------------------------------------------------------------------
350    Name: Matthias Blume
351    Date: 2000/11/21 21:20:00 JST
352    Tag: blume-20001121-tools
353    Description:
354    
355    Continued hacking on autoloading problem -- with success this time.
356    Also changed tool-plugin mechanism.  See new CM manual.
357    
358    ----------------------------------------------------------------------
359    Name: Matthias Blume
360    Date: 2000/11/19 14:30:00 JST
361    Tag:  blume-20001119-autoload
362    Description:
363    
364    Some hacking to make autoloading faster.  Success for CMB, no success
365    so far for CM.  There is a reduced structure CM' that autoloads faster.
366    (This is a temporary, non-documented hack to be eliminated again when
367    the general problem is solved.)
368    
369    ----------------------------------------------------------------------
370    Name: Matthias Blume
371    Date: 2000/11/17 14:10:00 JST
372    Tag: blume-20001117-pickle-lib
373    Description:
374    
375    1. Eliminated comp-lib.cm
376    2. Made pickle-lib.cm
377    3. Eliminated all uses of intset.sml (from comp-lib.cm)
378    4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
379       equivalent constructs from smlnj-lib.cm (INtHashTable).
380    5. Point 4. also goes for those uses of intmap.* in MLRISC.
381       Duplicated intmap modules thrown out.
382    6. Hunted down all duplicated SCC code and replaced it with
383       equivalent stuff (GraphSCCFn from smlnj-lib.cm).
384    7. Rewrote Feedback module.
385    8. Moved sortedlist.sml into viscomp-lib.cm.  Eventually it
386       should be thrown out and equivalent modules from smlnj-lib.cm
387       should be used (IntRedBlackSet, IntListSet, ...).
388    
389    Confirmed that compiler compiles to fixpoint.
390    
391    ----------------------------------------------------------------------
392    Name: Allen Leung
393    Date: 2000/11/10 18:00:00
394    Tag: leunga-20001110-new-x86-fp
395    
396    A new x86 floating point code generator has been added.
397    By default this is turned off.  To turn this on, do:
398    
399        CM.autoload "$smlnj/compiler.cm";
400        Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
401    
402    Changes:
403    
404    1.  Changed FTAN to FPTAN so that the assembly output is correct.
405    2.  Changed the extension callback for FTANGENT to generate:
406    
407              fptan
408              fstp  %st(0)
409        instead of
410              fptan
411              fstpl ftempmem
412    
413    3.  Numerous assembly fixes for x86.
414    
415    5.  Cleaned up the machine code output module x86/x86MC.sml and added
416        support for a whole bunch of instructions and addressing modes:
417    
418          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st, %st(n)
419          faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp  %st, %st(n)
420          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st(n), %st
421          fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
422          fxch %st(n)
423          fld %st(n)
424          fst %st(n)
425          fst mem
426          fstp %st(n)
427          fucom %st(n)
428          fucomp %st(n)
429    
430        All these are now generated when the fast fp mode is turned on.
431    
432    6.  Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
433    
434    ----------------------------------------------------------------------
435    Name: Matthias Blume
436    Date: 2000/11/09 11:20:00 JST
437    Tag: blume-20001109-scc
438    Description:
439    
440    Eliminated some code duplication:
441    
442    1. Added "where" clause to GraphSCCFn in SML/NJ Library.
443       (Otherwise the functor is useless.)
444    2. Used GraphSCCFn where SCCUtilFun was used previously.
445    3. Got rid of SCCUtilFun (in comp-lib.cm).
446    
447    ----------------------------------------------------------------------
448    Name: Lal George
449    Date: 2000/11/06 09:02:21 EST 2000
450    Tag: Release_110_30
451    Description:
452    
453    - Version 110.30
454    ----------------------------------------------------------------------
455    Name: Matthias Blume
456    Date: 2000/11/04 14:45:00
457    Tag: blume-20001104-mlbuild
458    Description:
459    
460    - Made ml-build faster on startup.
461    - Documentation fixes.
462    
463    ----------------------------------------------------------------------
464    Name: Matthias Blume
465    Date: 2000/11/02 17:00:00 JST
466    Tag: blume-20001102-condcomp
467    Description:
468    
469    - Small tweaks to pickler -- new BOOTFILES!
470    - Version bumped to 110.29.2.
471    - Added conditional compilation facility to init.cmi (see comment there).
472    ----------------------------------------------------------------------
473    Name: Allen Leung
474    Date: 2000/10/23 19:31:00
475    Tag: leunga-20001023-demo-ra
476    
477    1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
478    2. Test programs for the graph library updated
479    3. Some new MLRISC demo programs added
480    
481    ----------------------------------------------------------------------
482    Name: Matthias Blume
483    Date: 2000/08/31 22:15:00 JST
484    Tag: blume-20001017-errmsg
485    Description:
486    
487    More error message grief: Where there used to be no messages, there
488    now were some that had bogus error regions.  Fixed.
489    
490    ----------------------------------------------------------------------
491    Name: Matthias Blume
492    Date: 2000/08/31 17:30:00 JST
493    Tag: blume-20001017-v110p29p1
494    Description:
495    
496    I made a version 110.29.1 with new bootfiles.
497    
498    Changes:  Modified pickler/unpickler for faster and leaner unpickling.
499              CM documentation changes and a small bugfix in CM's error reporting.
500    
501    ----------------------------------------------------------------------
502    Name: Lal George
503    Date: 2000/09/27 14:42:35 EDT
504    Tag: george-20000927-nodestatus
505    Description:
506    
507    Changed the type of the nodestatus, so that:
508    
509            SPILLED(~1)             is now SPILLED
510            SPILLED(m) where m>=0   is now MEMREG(m)
511            SPILLED(s) where s<~1   is now SPILL_LOC(~s)
512    
513    ----------------------------------------------------------------------
514    Name: Matthias Blume
515    Date: 2000/09/07 14:45:00 JST
516    Tag: blume-20000907-cmerrmsg
517    Description:
518    
519    Small tweak to CM to avoid getting ML syntax error messages twice.
520    
521    ----------------------------------------------------------------------
522    Name: Matthias Blume
523    Date: 2000/08/31 18:00:00 JST
524    Tag: blume-20000831-cvsbootfiles
525    Description:
526    
527    New URL for boot files (because the 110.29 files on the BL server do
528    now work correctly with my updated install scripts for yacc and lex).
529    
530    ----------------------------------------------------------------------
531    Name: Matthias Blume
532    Date: 2000/08/08 12:33:00 JST
533    Tag: blume-20000808-manual
534    Description:
535    
536    Tiny update to CM manual.
537    
538    ----------------------------------------------------------------------
539    Name: Allen Leung
540    Date: 2000/08/7 19:31:00
541    Tag: leunga-20000807-a-whole-bunch-of-stuff
542    
543      Moby, C--, SSA, x86, machine descriptions etc.  Should only affect C--
544    and Mobdy.
545    
546    1.  x86
547    
548       a.  Fixes to peephole module by John and Dan.
549       b.  Assembly fix to SETcc by Allen.
550       c.  Fix to c-call by John.
551       d.  Fix to spilling by John.  (This one deals with the missing FSTPT case)
552       e.  Instruction selection optimization to SETcc as suggested by John.
553    
554           For example,
555    
556            MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
557    
558           should generate:
559    
560            MOVL a, x
561            SUBL b, x
562            SHRL 31, x
563    
564    2.  IR stuff
565    
566         A bunch of new DJ-graph related algorithms added.  These
567         speed up SSA construction.
568    
569    3.  SSA + Scheduling
570    
571         Added code for SSA and scheduling to the repository
572    
573    ----------------------------------------------------------------------
574    Name: Lal George
575    Date: 2000/07/27 11:53:14 EDT
576    
577    Tag: lal-20000727-linux-ppc
578    Description:
579    
580     Made changes to support Linux PPC.
581     p.s. I have confirmation that the 110.29 boot files work fine.
582    
583    ----------------------------------------------------------------------
584    Name: Matthias Blume
585    Date: 2000/07/27 17:40:00 JST
586    Tag: blume-20000727-scripts
587    Description:
588    
589    !!!! WARNING !!!!
590    You must recompile the runtime system!
591    !!!! WARNING !!!!
592    
593    This is basically another round of script-enhancements:
594    
595    1. sml, ml-build, and ml-makedepend accept options -D and -U to define
596       and undefine CM preprocessor symbols.
597    
598    2. ml-build avoids generating a new heap image if it finds that the
599       existing one is still ok.  (The condition is that no ML file had to
600       be recompiled and all ML files are found to be older that the heap
601       file.)
602    
603       To make this work smoothly, I also hacked the runtime system as
604       well as SMLofNJ.SysInfo to get access to the heap image suffix
605       (.sparc-solaris, ...) that is currently being used.
606    
607       Moreover, the signature of CM.mk_standalone has changed.  See the
608       CM manual.
609    
610    3. ml-makedepend accepts additional options -n, -a, and -o.  (See the
611       CM manual for details.)
612    
613    4. More CM manual updates:
614        - all of the above has been documented.
615        - there is now a section describing the (CM-related) command line
616          arguments that are accepted by the "sml" command
617    
618    ----------------------------------------------------------------------
619    Name: Matthias Blume
620    Date: 2000/07/25 16:20:00 JST
621    Tag: blume-20000725-makedepend
622    Description:
623    
624    Added a script called ml-makedepend.  This can be used in makefiles
625    for Unix' make in a way very similar to the "makedepend" command for
626    C.
627    
628    The script internally uses function CM.sources.
629    
630    Synopsis:
631    
632        ml-makedepend [-f makefile] cmfile targetname
633    
634    The default for the makefile is "makefile" (or "Makefile" should
635    "makefile" not exist).
636    
637    ml-makedepend adds a cmfile/targetname-specific section to this
638    makefile (after removing the previous version of this section).  The
639    section contains a single dependency specification with targetname on
640    the LHS (targetname is an arbitrary name), and a list of files derived
641    from the cmfile on the RHS.  Some of the files on the RHS are
642    ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
643    "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
644    path names.  The makefile writer is responsible for making sure that
645    these variables have correct at the time "make" is invoked.
646    
647    ----------------------------------------------------------------------
648    Name: Matthias Blume
649    Date: 2000/07/22 23:30:00 JST
650    Tag: blume-20000722-urlupdate
651    Description:
652    
653    Changed BOOT and config/srcarchiveurl to point to BL server:
654    
655        ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
656    
657    ----------------------------------------------------------------------
658    Name: Matthias Blume
659    Date: 2000/07/18 18:00:00 JST
660    Tag: blume-20000718-Version_110_29
661    Description:
662    
663    1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
664    
665    2. Updated config/version to 110.29
666    
667    3. Updated config/srcarchiveurl
668    
669    3. New boot files!
670       ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
671    
672    ----------------------------------------------------------------------
673    Name: Matthias Blume
674    Date: 2000/07/11 13:58:00 JST
675    Tag: blume-20000711-doctypo
676    Description:
677    
678    Fixed a few typos in CM manual.
679    
680    ----------------------------------------------------------------------
681    Name: Allen Leung
682    Date: 2000/06/15 00:38:00
683    Tag: leunga-20000704-sparc-x86
684    
685    1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
686    2. fix to x86 RET bug [found by Dan Grossman]
687    3. sparc assembly bug fix for ticc instructions [found by Fermin]
688    
689       Affects c-- and moby only
690    
691    ----------------------------------------------------------------------
692    Name: Matthias Blume
693    Date: 2000/07/04 15:26:00
694    Tag: blume-20000704-trigger
695    Description:
696    
697    1. Improvements to CM manual.
698    2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
699       of getting a back-trace.  The function, when called, raises an
700       internal exception which explicitly carries the full back-trace history,
701       so it is unaffected by any intervening handle-raise pairs ("trivial"
702       or not).  The interactive loop will print that history once it arrives
703       at top level.
704       Short of having all exceptions implicitly carry the full history, the
705       recommended way of using this facility is:
706         - compile your program with instrumentation "on"
707         - run it, when it raises an exception, look at the history
708         - if the history is "cut off" because of some handler, go and modify
709           your program so that it explicitly calls BTrace.trigger
710         - recompile (still instrumented), and rerun; look at the full history
711    
712    ----------------------------------------------------------------------
713    Name: Matthias Blume
714    Date: 2000/07/03 15:36:00 JST
715    Tag: blume-20000702-manual
716    Description:
717    
718    Small corrections and updates to CM manual.
719    
720    ----------------------------------------------------------------------
721    Name: Matthias Blume
722    Date: 2000/06/29 16:04:00 JST
723    Tag: blume-20000629-yacctool
724    Description:
725    
726    Changes:
727    
728    1. Class "mlyacc" now takes separate arguments to pass options to
729       generated .sml- and .sig-files independently.
730    2. Corresponding CM manual updates.
731    3. BTrace module now also reports call sites.  (However, for loop clusters
732       it only shows from where the cluster was entered.)  There are associated
733       modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
734    
735    ----------------------------------------------------------------------
736    Name: Matthias Blume
737    Date: 2000/06/27 16:51:00 JST
738    Tag: blume-20000627-noweb
739    Description:
740    
741    Changes:
742    
743     1. Implemented "subdir" and "witness" options for noweb tool.
744        This caused some slight internal changes in CM's tool implementation.
745     2. Fixed bug in "tool plugin" mechanism.  This is essentially cleaning
746        some remaining issues from earlier path anchor changes.
747     3. Updated CM manual accordingly.
748    
749     4. Changed implementation of back-tracing so that I now consider it
750        ready for prime-time.
751    
752        In particular, you don't have to explicitly trigger the back-trace
753        anymore.  Instead, if you are running BTrace-instrumented code and
754        there is an uncaught exception (regardless of whether or not it was
755        raised in instrumented code), the top-level evalloop will print
756        the back-trace.
757    
758        Features:
759    
760          - Instrumented and uninstrumented code work together seemlessly.
761            (Of course, uninstrumented code is never mentioned in actual
762             back-traces.)
763    
764          - Asymptotic time- and space-complexity of instrumented code is
765            equal to that of uninstrumented code.  (This means that
766            tail-recursion is preserved by the instrumentation phase.)
767    
768          - Modules whose code has been instrumented in different sessions
769            work together without problem.
770    
771          - There is no penalty whatsoever on uninstrumented code.
772    
773          - There is no penalty on "raise" expressions, even in
774            instrumented code.
775    
776        A potential bug (or perhaps it is a feature, too):
777    
778          A back-trace reaches no further than the outermost instrumented
779          non-trivial "raise".  Here, a "trivial" raise is one that is the
780          sole RHS of a "handle" rule.  Thus, back-traces reach trough
781    
782               <exp> handle e => raise e
783    
784          and even
785    
786               <exp> handle Foo => raise Bar
787    
788          and, of course, through
789    
790               <exp> handle Foo => ...
791    
792         if the exception was not Foo.
793    
794         Back-traces always reach right through any un-instrumented code
795         including any of its "handle" expressions, trivial or not.
796    
797       To try this out, do the following:
798    
799         - Erase all existing binfiles for your program.
800           (You may keep binfiles for those modules where you think you
801            definitely don't need back-tracing.)
802         - Turn on back-trace instrumentation:
803              SMLofNJ.Internals.BTrace.mode (SOME true);
804         - Recompile your program.  (I.e., run "CM.make" or "use".)
805         - You may now turn instrumentation off again (if you want):
806              SMLofNJ.Internals.BTrace.mode (SOME false);
807         - Run your program as usual.  If it raises an exception that
808           reaches the interactive toplevel, then a back-trace will
809           automatically be printed.  After that, the toplevel loop
810           will print the exception history as usual.
811    
812    ----------------------------------------------------------------------
813    Name: Matthias Blume
814    Date: 2000/06/26 09:56:46 JST
815    Tag: blume-20000626-setup
816    Description:
817    
818    CM: - setup-parameter to "sml" added; this can be used to run arbitrary
819          ML code before and after compiling a file (e.g., to set compiler
820          flags)
821    
822    Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
823              - associated changes to btrace.sml (BTrace instrumentation pass)
824              - cleaner implementation of btimp.sml (BTrace tracing and report
825                module)
826    
827    CM manual: * new path encoding documented
828               * description of setup-parameter to "sml" added
829    
830    The biggest user-visible change to back-tracing is that it is no
831    longer necessary to compile all traced modules within the same
832    session.  (This was a real limitation.)
833    
834    ----------------------------------------------------------------------
835    Name: Matthias Blume
836    Date: 2000/06/24 12:40:00 JST
837    Tag: blume-20000624-startup
838    Description:
839    
840    Fixes startup slowdown problem.  (I was calling SrcPath.sync a _tad_
841    bit too often -- to put it mildly. :)
842    
843    ----------------------------------------------------------------------
844    Name: Matthias Blume
845    Date: 2000/06/23 18:20:00 JST
846    Tag: blume-20000623-btrace
847    Description:
848    
849    This updates adds a backtrace facility to aid programmers in debugging
850    their programs.  This involves the following changes:
851    
852    1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
853       keeping track of the current call stack.  When programs are compiled
854       in a special mode, the compiler will insert calls to these hooks
855       into the user program.
856       "Hook" means that it is possible for different implementations of
857       back-tracing to register themselves (at different times).
858    
859    2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
860       as an Absyn.dec->Absyn.dec rewrite.  Normally this phase is turned off.
861       It can be turned on using this call:
862         SMLofNJ.Internals.BTrace.mode (SOME true);
863       Turning it off again:
864         SMLofNJ.Internals.BTrace.mode (SOME false);
865       Querying the current status:
866         SMLofNJ.Internals.BTrace.mode NONE;
867       Annotated programs are about twice as big as normal ones, and they
868       run a factor of 2 to 4 slower with a dummy back-trace plugin (one
869       where all hooks do nothing).  The slowdown with a plugin that is
870       actually useful (such as the one supplied by default) is even greater,
871       but in the case of the default plugin it is still only an constant
872       factor (amortized).
873    
874    3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
875       with a sub-structure BTrace for controlling back-tracing.  In particular,
876       the above-mentioned function "mode" controls whether the annotation
877       phase is invoked by the compiler.  Another important function is
878       "trigger": when called it aborts the current execution and causes
879       the top-level loop to print a full back-trace.
880    
881    4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
882       for back-tracing.  It keeps track of the dynamic call stack and in
883       addition to that it keeps a partial history at each "level" of that
884       stack.  For example, if a tail-calls b, b tail-calls c, and c tail-calls
885       d and b (at separate times, dynamically), then the report will show:
886    
887       GOTO   d
888             /c
889       GOTO  \b
890       CALL   a
891    
892       This shows that there was an initial non-tail call of a, then a
893       tail-call to b or c, looping behavior in a cluster of functions that
894       consist of b and c, and then a goto from that cluster (i.e., either from
895       b or from c) to d.
896    
897       Note that (depending on the user program) the amount of information
898       that the back-trace module has to keep track of at each level is bounded
899       by a constant.  Thus, the whole implementation has the same asymptotical
900       complexity as the original program (both in space and in time).
901    
902    5. compiler/TopLevel/interact/evalloop.sml has been modified to
903       handle the special exception SMLofNJ.Internals.BTrace.BTrace
904       which is raised by the "trigger" function mentioned above.
905    
906    Notes on usage:
907    
908    - Annotated code works well together with unannotated code:
909    Unannotated calls simply do not show up at all in the backtrace.
910    
911    - It is not a good idea to let modules that were annotated during
912    different sessions run at the same time.  This is because the compiler
913    chooses small integers to identify individual functions, and there
914    will be clashes if different modules were compiled in separate sessions.
915    (Nothing will crash, and you will even be told about the clashes, but
916    back-trace information will in general not be useful.)
917    
918    - Back-tracing can be confused by callcc and capture.
919    
920    - The only way of getting a back-trace right now is to explicitly
921    invoke the "trigger" function from your user program.  Eventually, we
922    should make every exception carry back-trace information (if
923    available).  But since this creates more overhead at "raise"-time
924    (similar to the current exnHistory overhead), I have not yet
925    implemented this.  (The implementation will be rather easy.)  With
926    exceptions carrying back-trace information, this facility will be even
927    more useful because users don't need to modify their programs...
928    
929    - While it is possible to compile the compiler with back-trace
930    annotations turned on (I did it to get some confidence in
931    correctness), you must make absolutely sure that core.sml and
932    btimp.sml are compiled WITHOUT annotation!  (core.sml cannot actually
933    be compiled with annotation because there is no core access yet, but
934    if you compile btimp.sml with annotation, then the system will go into
935    an infinite recursion and crash.)
936    Since CM currently does not know about BTrace, the only way to turn
937    annotations on and off for different modules of the compiler is to
938    interrupt CMB.make, change the settings, and re-invoke it.  Of course,
939    this is awkward and clumsy.
940    
941    Sample sessions:
942    
943    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
944    - SMLofNJ.Internals.BTrace.mode (SOME true);
945    [autoloading]
946    [autoloading done]
947    val it = false : bool
948    - structure X = struct
949    -     fun main n = let
950    -         fun a (x, 0) = d x
951    -           | a (x, n) = b (x, n - 1)
952    -         and b (x, n) = c (x, n)
953    -         and c (x, n) = a (x, n)
954    -         and d x = e (x, 3)
955    -         and e (x, 0) = f x
956    -           | e (x, n) = e (x, n - 1)
957    -         and f 0 = SMLofNJ.Internals.BTrace.trigger ()
958    -           | f n = n * g (n - 1)
959    -         and g n = a (n, 3)
960    -     in
961    -         f n
962    -     end
963    - end;
964    structure X : sig val main : int -> int end
965    - X.main 3;
966    *** BACK-TRACE ***
967    GOTO   stdIn:4.2-13.20: X.main[2].f
968    GOTO-( stdIn:4.2-13.20: X.main[2].e
969    GOTO   stdIn:4.2-13.20: X.main[2].d
970         / stdIn:4.2-13.20: X.main[2].a
971         | stdIn:4.2-13.20: X.main[2].b
972    GOTO-\ stdIn:4.2-13.20: X.main[2].c
973    CALL   stdIn:4.2-13.20: X.main[2].g
974    GOTO   stdIn:4.2-13.20: X.main[2].f
975    GOTO-( stdIn:4.2-13.20: X.main[2].e
976    GOTO   stdIn:4.2-13.20: X.main[2].d
977         / stdIn:4.2-13.20: X.main[2].a
978         | stdIn:4.2-13.20: X.main[2].b
979    GOTO-\ stdIn:4.2-13.20: X.main[2].c
980    CALL   stdIn:4.2-13.20: X.main[2].g
981    GOTO   stdIn:4.2-13.20: X.main[2].f
982    GOTO-( stdIn:4.2-13.20: X.main[2].e
983    GOTO   stdIn:4.2-13.20: X.main[2].d
984         / stdIn:4.2-13.20: X.main[2].a
985         | stdIn:4.2-13.20: X.main[2].b
986    GOTO-\ stdIn:4.2-13.20: X.main[2].c
987    CALL   stdIn:4.2-13.20: X.main[2].g
988    GOTO   stdIn:4.2-13.20: X.main[2].f
989    CALL   stdIn:2.15-17.4: X.main[2]
990    -
991    
992    (Note that because of a FLINt bug the above code currently does not
993    compile without BTrace turned on.)
994    
995    Here is another example, using my modified Tiger compiler:
996    
997    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
998    - SMLofNJ.Internals.BTrace.mode (SOME true);
999    [autoloading]
1000    [autoloading done]
1001    val it = false : bool
1002    - CM.make "sources.cm";
1003    [autoloading]
1004    ...
1005    [autoloading done]
1006    [scanning sources.cm]
1007    [parsing (sources.cm):parse.sml]
1008    [creating directory CM/SKEL ...]
1009    [parsing (sources.cm):tiger.lex.sml]
1010    ...
1011    [wrote CM/sparc-unix/semant.sml]
1012    [compiling (sources.cm):main.sml]
1013    [wrote CM/sparc-unix/main.sml]
1014    [New bindings added.]
1015    val it = true : bool
1016    - Main.compile ("../testcases/merge.tig", "foo.out");
1017    *** BACK-TRACE ***
1018    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
1019    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1020    CALL   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1021    GOTO   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1022    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1023    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1024    CALL   lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
1025         / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
1026    CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
1027    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1028    CALL   lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
1029    CALL   main.sml:1.18-118.4: Main.compile[2]
1030    -
1031    
1032    ----------------------------------------------------------------------
1033    Name: Matthias Blumen
1034    Date: 2000/06/21 18:00:00 JST
1035    Tag: blume-20000621-manual
1036    Description:
1037    
1038    CM manual update: Path environments documented.
1039    
1040    ----------------------------------------------------------------------
1041    Name: Matthias Blume
1042    Date: 2000/06/19 13:40:00
1043    Tag: blume-20000619-manual
1044    Description:
1045    
1046    CM manual and system/README update.  This only covers the fact that
1047    there are no more implicit anchors.  (Path environments and the "bind"
1048    option to "cm" have yet to be documented.)
1049    
1050    ----------------------------------------------------------------------
1051    Name: Matthias Blume
1052    Date: 2000/06/19 11:05:00 JST
1053    Tag: blume-20000619-chdir-bugfix
1054    Description:
1055    
1056    Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
1057    
1058    ----------------------------------------------------------------------
1059    Name: Matthias Blume
1060    Date: 2000/06/18 22:00:10 JST
1061    Tag: blume-20000618-implicit-anchors-really-gone
1062    Description:
1063    
1064    I updates the previous HISTORY entry where I forgot to mention that
1065    implicit anchors are no longer with us.
1066    
1067    The current update also gets rid of the (now useless) controller
1068    CM.Control.implicit_anchors.
1069    
1070    ----------------------------------------------------------------------
1071    Name: Matthias Blume
1072    Date: 2000/06/16 17:30:00 JST
1073    Tag: blume-20000616-anchorenv
1074    Description:
1075    
1076    This patch implements the long anticipated (just kidding :) "anchor
1077    environment" mechanism.  In the course of doing this, I also
1078    re-implemented CM's internal "SrcPath" module from scratch.  The new
1079    one should be more robust in certain boundary cases.  In any case, it
1080    is a lot cleaner than its predecessor (IMHO).
1081    
1082    This time, although there is yet another boot file format change, I
1083    kept the unpickler backward-compatible.  As a result, no new bootfiles
1084    are necessary and bootstrapping is straightforward.  (You cannot read
1085    new bootfiles into an old system, but the other way around is no
1086    problem.)
1087    
1088    Visible changes:
1089    
1090    ** 0. Implicit path anchors (without the leading $-symbol) are no
1091    longer recognized at all. This means that such path names are not
1092    illegal either.  For example, the name basis.cm simply refers to a
1093    local file called "basis.cm" (i.e, the name is an ordinary path
1094    relative to .cm-files directory).  Or, to put it differently, only
1095    names that start with $ are anchored paths.
1096    
1097    ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
1098    vanished.
1099    
1100    John (Reppy) had critizised this as soon as I originally proposed and
1101    implemented it, but at that time I did not really deeply believe
1102    him. :) Now I came full-circle because I need the $<singlearc> syntax
1103    in another place where it cannot be seen as an abbreviation for
1104    $/<singlearc>.  To avoid the confusion, $<singlearc> now means what it
1105    seems to mean (i.e., it "expands" into the corresponding anchor
1106    value).
1107    
1108    However, when paths are used as members in CM description files, it
1109    continues to be true that there must be at least another arc after the
1110    anchor.  This is now enforced separately during semantic analysis
1111    (i.e., from a lexical/syntactical point of view, the notation is ok.)
1112    
1113    ** 2. The "cm" class now accepts an option "bind".  The option's value
1114    is a sub-option list of precisely two items -- one labeled "anchor"
1115    and the other one labeled "value".  As you might expect, "anchor" is
1116    used to specify an anchor name to be bound, and "value" specifies what
1117    the anchor is being bound to.
1118    
1119    The value must be a directory name and can be given in either standard
1120    syntax (including the possibility that it is itself an anchored path)
1121    or native syntax.
1122    
1123    Examples:
1124    
1125       foo.cm (bind:(anchor:bar value:$mystuff/bar))
1126       lib.cm (bind:(anchor:a value:"H:\\x\\y\\z"))  (* only works under windows *)
1127    
1128    and so on.
1129    
1130    The meaning of this is that the .cm-file will be processed with an
1131    augmented anchor environment where the given anchor(s) is/are bound to
1132    the given values(s).
1133    
1134    The rationale for having this feature is this: Suppose you are trying
1135    to use two different (already stable) libraries a.cm and b.cm (that
1136    you perhaps didn't write yourself).  Further, suppose each of these
1137    two libraries internally uses its own auxiliary library $aux/lib.cm.
1138    Normally you would now have a problem because the anchor "lib" can not
1139    be bound to more than one value globally.  Therefore, the project that
1140    uses both a.cm and b.cm must locally redirect the anchor to some other
1141    place:
1142    
1143       a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
1144       b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
1145    
1146    This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
1147    /usr/lib/smlnj/b-stuff/aux.cm, respectively.
1148    
1149    Hard-wiring path names is a bit inflexible (and CM will verbosely warn
1150    you when you do so at the time of CM.stabilize).  Therefore, you can
1151    also use an anchored path as the value:
1152    
1153      a.cm (bind:(anchor:lib value:$a-lib))
1154      b.cm (bind:(anchor:lib value:$b-lib))
1155    
1156    Now you can globally configure (using the usual CM.Anchor.anchor or
1157    pathconfig machinery) bindings for "a-lib" and "b-lib".  Since "lib"
1158    itself is always locally bound, setting it globally is no longer
1159    meaningful or necessary (but it does not hurt either).  In fact, "lib"
1160    can still be used as a global anchor for separate purposes.  As a
1161    matter of fact, one can locally define "lib" in terms of a global
1162    "lib":
1163    
1164      a.cm (bind:(anchor:lib value:$lib/a))
1165      b.cm (bind:(anchor:lib value:$lib/b))
1166    
1167    ** 3: The encoding of path names has changed.  This affects the way
1168    path names are shown in CM's progress report and also the internal
1169    protocol encoding used for parallel make.
1170    
1171    The encoding now uses one or more ':'-separated segments.  Each
1172    segments corresponds to a file that has been specified relative to the
1173    file given by its preceding segment.  The first segment is either
1174    relative to the CWD, absolute, or anchored.  Each segment itself is
1175    basically a Unix pathname; all segments but the first are relative.
1176    
1177    Example:
1178    
1179       $foo/bar/baz.cm:a/b/c.sml
1180    
1181    This path denotes the file bar/a/b/c.sml relative to the directory
1182    denoted by anchor "foo".  Notice that the encoding also includes
1183    baz.cm which is the .cm-file that listed a/b/c.sml.  As usual, such
1184    paths are resolved relative to the .cm-files directory, so baz.cm must
1185    be ignored to get the "real" pathname.
1186    
1187    To make this fact more obvious, CM puts the names of such "virtual
1188    arcs" into parentheses when they appear in progress reports. (No
1189    parentheses will appear in the internal protocol encoding.)  Thus,
1190    what you really see is:
1191    
1192      $foo/bar/(baz.cm):a/b/c.sml
1193    
1194    I find this notation to be much more informative than before.
1195    
1196    Another new feature of the encoding is that special characters
1197    including parentheses, colons, (back)slashes, and white space are
1198    written as \ddd (where ddd is the decimal encoding of the character).
1199    
1200    *** The CM manual still needs to be updated.
1201    
1202    ----------------------------------------------------------------------
1203    Name: Allen Leung
1204    Date: 2000/06/15 00:38:00
1205    Tag: leunga-20000615-x86-peephole
1206    
1207    x86 Peephole fix by Fermin.  Affects c-- and moby only.
1208    
1209    ----------------------------------------------------------------------
1210    Name: Matthias Blume
1211    Date: 2000/06/12 11:40:00
1212    Tag: blume-20000612-parmakefix
1213    Description:
1214    
1215    More cleanup after changing the file naming scheme: This time I
1216    repaired the parallel make mechanism for CMB.make which I broke earlier.
1217    
1218    ----------------------------------------------------------------------
1219    Name: Allen Leung
1220    Date: 2000/06/09 01:25:00
1221    Tag: leunga-20000609-various
1222    
1223    None of these things should affect normal SML/NJ operations
1224    
1225    1. Peephole improvements provided by Fermin (c--)
1226    2. New annotation DEFUSE for adding extra dependence (moby)
1227    3. New X86 LOCK instructions (moby)
1228    4. New machine description language for reservation tables (scheduling)
1229    5. Fixes to various optimization/analysis modules (branch chaining, dominator
1230       trees etc.)
1231    6. I've changed the CM files so that they can work with versions
1232       110.0.6, 110.25 and 110.28
1233    
1234    ----------------------------------------------------------------------
1235    Name: Matthias Blume
1236    Date: 2000/06/09 12:40:00
1237    Tag: blume-20000609-log
1238    Description:
1239    
1240    - Removed all(?) remaining RCS Log entries from sources.
1241    
1242    - Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
1243      anchored paths).
1244    
1245    ----------------------------------------------------------------------
1246    Name: Matthias Blume
1247    Date: 2000/06/07 17:00:00 JST
1248    Tag: blume-20000607-no-implicit-anchors
1249    Description:
1250    
1251    1. This update changes the default setting for
1252    CM.Control.implicit_anchors from true to false.  This means that
1253    implicit anchors are no longer permitted by default.  I also tried to
1254    make sure that nothing else still relies on implicit anchors.
1255    (This is the next step on the schedule towards a CM that does not even
1256    have the notion of implicit anchors anymore.)
1257    
1258    2. More CM manual updates.
1259    
1260    3. I managed to track down and fix the pickling bug I mentioned last
1261    time.  Because of the previously existing workaround, this entails no
1262    immediate practical changes.
1263    
1264    ----------------------------------------------------------------------
1265    Name: Matthias Blume
1266    Date: 2000/06/06 11:15:00 JST
1267    Tag: blume-20000606-lazierpickle
1268    Description:
1269    
1270    !!!! NEW BOOT FILES !!!!
1271    
1272    * The main purpose of this update is to make library pickles lazier in
1273    order to reduce the initial space penalty for autoloading a library.
1274    As a result, it is now possible to have $smlnj/compiler.cm
1275    pre-registered.  This should take care of the many complaints or
1276    inquiries about missing structure Compiler.  This required changes to
1277    CM's internal data structures and small tweaks to some algorithms.
1278    
1279    As a neat additional effect, it is no longer necessary (for the sake
1280    of lean heap image files) to distinguish between a "minimal" CM and a
1281    "full" CM.  Now, there is only one CM (i.e., the "full" version:
1282    $smlnj/cm.cm aka $smlnj/cm/full.cm), and it is always available at the
1283    interactive top level. ($smlnj/cm/minimal.cm is gone.)
1284    
1285    To make the life of compiler-hackers easier, "makeml" now also
1286    pre-registers $smlnj/cmb.cm (aka $smlnj/cmb/current.cm).  In other
1287    words, after you bootstrap a new sml for the first time, you will not
1288    have to autoload $smlnj/cmb.cm again afterwards.  (The first time
1289    around you will still have to do it, though.)
1290    
1291    * A second change consists of major updates to the CM manual.  There
1292    are now several appendices with summary information and also a full
1293    specification of the CM description file syntax.
1294    
1295    * In directory src/system I added the script "allcross".  This script
1296    invokes sml and cross-compiles the compiler for all supported
1297    architectures.  (Useful when providing a new set of boot files.)
1298    
1299    * There seems to be a latent bug in my "lazy pickles" mechanism.  I
1300    added a small tweak to pickle-util.sml to work around this problem,
1301    but it is not a proper fix yet.  I will investigate further.  (The
1302    effect of the bug was an inflation of library pickle size.)
1303    
1304    * Version number increased to 110.28.1 (to avoid compatibility problems).
1305    
1306    ----------------------------------------------------------------------
1307    Name: Allen Leung
1308    Date: 2000/05/25 17:28 EDT
1309    Tag: leunga-20000525-ra
1310    Description:
1311    
1312      Fixed a bug in freezing phase of the register allocator.
1313    
1314    ----------------------------------------------------------------------
1315    Name: Allen Leung
1316    Date: 2000/05/15 22:53 EDT
1317    Tag: leunga-20000515-alpha-x86-ra
1318    Description:
1319    
1320      1. Alpha
1321    
1322          Slight cleanup.  Removed the instruction SGNXL
1323    
1324      2. X86
1325    
1326          Added the following instructions to the instruction set:
1327    
1328            ROLx, RORx,
1329            BTx, BTSx, BTLx, BTRx,
1330            XCHGx, and variants with the LOCK prefix
1331    
1332      3. Register Allocation
1333    
1334          The module ra-rewrite-with-renaming has been improved.
1335    
1336      These have no effect on SML/NJ.
1337    
1338    ----------------------------------------------------------------------
1339    Name: Matthias Blume
1340    Date: 2000/05/15 16:20:00 JST
1341    Tag: blume-20000515-lightrebuild
1342    Description:
1343    
1344    1. I added an alternative to "-rebuild" to "makeml".  The difference is
1345       that prior to calling CMB.make' the CM-variable "LIGHT" will be
1346       defined.  In effect, the command will not build any cross-compiler
1347       backends and therefore finish more quickly.
1348    
1349       The "fixpt" script also takes a "-light" switch to be able to use
1350       this new facility while compiling for a fixpoint.
1351    
1352    2. I replaced all mentions of anchored paths in group owner specifications
1353       with simple relative paths (usually starting with "..").
1354       The rationale is that a library's internal workings should not be
1355       compromised by the lack of some anchor.  (An anchor is necessary
1356       for someone who wants to refer to the library by an anchored path,
1357       but it should not be necessary to build the same library in the first
1358       place.)
1359    
1360    3. I changed the way CM's tool mechanism determines the shell command
1361       string used for things like ml-yacc etc. so that it does not break
1362       when CM.Control.implicit_anchors is turned off.
1363    
1364    ----------------------------------------------------------------------
1365    Name: Matthias Blume
1366    Date: 2000/05/12 18:20:00 JST
1367    Tag: blume-20000512-ml-build
1368    Description:
1369    
1370    Fixed a bug in config/_ml-build that prevented ml-yacc and ml-lex from
1371    getting installed properly (by config/install.sh).
1372    
1373    ----------------------------------------------------------------------
1374    Name: Matthias Blume
1375    Date: 2000/05/12 17:30:00 JST
1376    Tag: blume-20000512-anchors
1377    Description:
1378    
1379    !!! NEW BOOT FILES !!!
1380    
1381    This change is in preparation of fading out support for "implicitly
1382    anchored path names".  I went through all sources and used the
1383    explicit (and relatively new) $-notation.  See system/README and the
1384    CM manual for more info on this.
1385    
1386    I also modified the anchoring scheme for some things such as "smlnj",
1387    "MLRISC", "cm", etc. to take advantage of the fact that explicit
1388    anchors are more expressive: anchor name and first arc do not have to
1389    coincide.  This entails the following user-visible change:
1390    
1391    You have to write $smlnj/foo/bar instead of smlnj/foo/bar.  In
1392    particular, when you fire up sml with a command-line argument, say,
1393    e.g.:
1394    
1395       sml '$smlnj/cmb.cm'
1396    
1397    At the ML toplevel prompt:
1398    
1399       CM.autoload "$smlnj/cmb.cm";
1400    
1401    There is also a new controller in CM.Control that can be used to turn
1402    off all remaining support for implicit anchors by saying:
1403    
1404        CM.autoload "$smlnj/
1405        #set CM.Control.implicit_anchors false;
1406    
1407    This causes CM to reject implicitly anchored paths.  This is (for the
1408    time being) less permissive than the "final" version where there will
1409    be no more such implicit anchors and relative paths will be just that:
1410    relative.
1411    
1412    The next step (version after next version?) will be to make the
1413    default for CM.Control.implicit_anchors false.  After the dust has
1414    settled, I can then produce the "final" version of this...
1415    
1416    Note: Since bootstrapping is a bit tricky, I provided new boot files.
1417    
1418    ----------------------------------------------------------------------
1419    Name: Matthias Blume
1420    Date: 2000/05/11 16:30:00 JST
1421    Tag: blume-20000511-sources
1422    Description:
1423    
1424    The main change is that I added function CM.sources as a generalized
1425    version of the earlier CM.makedepend.  This entails the following
1426    additional changes:
1427    
1428      - CM.makedepend has been dropped.
1429    
1430      - CM manual has been updated.
1431    
1432      - TOOLS signature and API have been changed.
1433    
1434    ----------------------------------------------------------------------
1435    Name: Allen Leung
1436    Date: 2000/05/10 21:17 EDT
1437    Tag: leunga-20000510-moby-c--ssa
1438    Description:
1439    
1440      Various bug fixes and new features for C--, Moby and MLRISC optimizations.
1441    None of these affect SML/NJ.
1442    
1443    1. Register Allocation
1444    
1445        a. A new ra spilling module (ra/ra-spill-with-renaming) is implemented.
1446           This module tries to remove local (i.e. basic block level) redundancies
1447           during spilling.
1448    
1449        b. A new framework for performing region based register allocation.
1450           Not yet entirely functional.
1451    
1452    2. X86
1453    
1454       a. DefUse for POP was missing the stack pointer [found by Lal]
1455       b. Reload for CALL was incorrect in X86Spill [found by John]
1456       c. Various fixes in X86Spill so that it can be used correctly for
1457          the new spilling module.
1458    
1459    3. SSA/IR
1460    
1461       a. New module ir/dj-dataflow.sml implements elimination based
1462          data flow analysis.
1463    
1464    4. MLRiscGen
1465    
1466       a. Fix for gc type annotation
1467    
1468    5. MDGen
1469    
1470       Various fixes for machine description -> ml code translation.  For ssa
1471       only.
1472    
1473    ----------------------------------------------------------------------
1474    Name: Allen Leung
1475    Date: 2000/05/08 22:17 EDT
1476    Tag: leunga-20000508-labexp
1477    Description:
1478    
1479      Fermin has found a few assembly problems with constant expressions
1480      generated in LabelExp.  Mostly, the problems involve extra parentheses,
1481      which choke on dumb assemblers.  This is his fix.
1482    
1483    ----------------------------------------------------------------------
1484    Name: Dave MacQueen
1485    Date: 2000/04/09 14:00 EDT
1486    Tag: dbm-20000502-Version_110_28
1487    Description:
1488    
1489    1. Updated src/compiler/TopLevel/main/version.sml to version 110.28
1490    
1491    2. Updated config/version to 110.28
1492    
1493    3. Updated config/srcarchiveurl
1494    
1495    3. New boot files!
1496       ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.28/
1497    
1498    ----------------------------------------------------------------------
1499    Name: Matthias Blume
1500    Date: 2000/05/01 19:05:00 JST
1501    Tag: blume-20000501-noweb
1502    Description:
1503    
1504    A new noweb tool has been added.  The existing system is entirely
1505    unaffected by this, but some CM users have asked for renewed noweb
1506    support.  Everything is documented in the CM manual.
1507    
1508    New (plugin) libraries:
1509    
1510       noweb-tool.cm
1511       nw-ext.cm
1512    
1513    ----------------------------------------------------------------------
1514    Name: Dave MacQueen
1515    Date: 2000/04/30 12:40PM EDT
1516    Tag: dbm-20000430-bug_fixes
1517    Description:
1518    
1519    1. Fix for bug 1498
1520       smlnj/src/system/Basis/Implementation/Unsafe/object.sig
1521       smlnj/src/system/Basis/Implementation/Unsafe/object.sml
1522         added toRealArray function
1523       smlnj/src/compiler/MiscUtil/print/ppobj.sml
1524         added check for tag Obj.RealArray to array printing case in ppObj
1525    
1526    2. Fix for bug 1510
1527       smlnj/src/compiler/Semant/types/typesutil.sml
1528         fixed definition of dummyargs (used by equalTycon) so that
1529         dummy args are distinct types
1530    
1531    ----------------------------------------------------------------------
1532    Name: Matthias Blume
1533    Date: 2000/04/30 01:00:00 JST
1534    Tag: blume-20000430-versions
1535    Description:
1536    
1537    1. CM version numbering added.  This is an implementation of Lal's
1538       proposal for adding version numbers and version checking to .cm
1539       files.  Lal said that his proposal was just that -- a proposal.
1540       For the time being I went ahead and implemented it so that people
1541       can comment on it.  Everything is completely backward-compatible
1542       (except for the stable library format, i.e., new bootfiles!).
1543    
1544       As usual, see the CM manual for details.
1545    
1546    2. An alternative syntax for anchored paths has been implemented.
1547       Dave has recently voiced the same concerns that I had when I did
1548       this, so there should be some support.  My take is that eventually
1549       I will let support for the current syntax (where anchors are
1550       "implicit") fade out in favor of the new, explicit syntax.
1551       In order to be backward-compatible, both old and new syntax are
1552       currently supported.
1553    
1554       Again, see the CM manual for details.
1555    
1556    3. Parallel make is trying to be slightly smarter:  When the master
1557       process finds a "bottleneck", i.e., when there is only one
1558       compilation unit that can be compiled and everybody else is
1559       waiting on it, then it will simply compile it directly instead
1560       of clumsily telling one of the slaves to do it.
1561    
1562    4. Support for "unsharing" added.  This is necessary in order to be
1563       able to have two different versions of the same library running
1564       at the same time (e.g., for trying out a new MLRISC while still
1565       having the old MLRISC linked into the current compiler, etc.)
1566       See the CM manual.
1567    
1568    5. Simple "makedepend" functionality added for generating Makefile
1569       dependency information.  (This is rather crude at the moment.
1570       Expect some changes here in the future.)
1571    
1572    6. ".fun" added as a recognized suffix for ML files. Also documented
1573       explicitly in the manual that the fallback behavior (unknown suffix
1574       -> ML file) is not an official feature!
1575    
1576    7. Small changes to the pickler for stable libraries.
1577    
1578    8. Several internal changes to CM (for cleanup/improvement).
1579    
1580    
1581    !!!! NEW BINFILES !!!!
1582    
1583    ----------------------------------------------------------------------
1584    Name: Matthias Blume
1585    Date: 2000/04/28 17:30:00 JST
1586    Tag: blume-20000428-pathconfig
1587    Description:
1588    
1589    1. I changed config/install.sh to remove duplicate entries from the
1590       lib/pathconfig file at the end.  Moreover, the final version of
1591       lib/pathconfig is sorted alphabetically.  The same (sorting) is done
1592       in src/system/installml.
1593    
1594    2. The config/install.sh script now consistently uses relative
1595       pathnames in lib/pathconfig whenever the anchor is in the lib
1596       directory.  (So far this was true for the libraries that come
1597       pre-compiled and bundled as part of the bootfiles but not for
1598       libraries that are compiled by the script itself.)
1599    
1600    ----------------------------------------------------------------------
1601    Name: Matthias Blume
1602    Date: 2000/04/26 13:10:00 JST
1603    Tag: blume-20000426-fun_suffix
1604    Description:
1605    
1606    Added ".fun" as a recognized file name suffix (for ML code).
1607    
1608    ----------------------------------------------------------------------
1609    Name: Allen Leung
1610    Date: 2000/04/25 17:00:00 EST
1611    Tag: leunga-20000425-alpha-ra
1612    Description:
1613    
1614    1. Alpha
1615    
1616        PSEUDOARITH was missing in AlphaRewrite.  This causes an endless loop
1617    in C--.
1618    
1619    2. RA
1620    
1621       Added a flag "ra-dump-size" to print out the size of the flowgraph
1622       and the interference graph.
1623    
1624    ----------------------------------------------------------------------
1625    Name: Dave MacQueen
1626    Date: 2000/04/25/
1627    Tag: dbm-20000425-mlyacc_doc_examples
1628    Description:
1629      Updated mlyacc.tex sections 5 and 7 for SML '97 and CM.
1630      Updated all three examples in src/ml-yacc/examples to run
1631      under 110.* using CM.make.
1632    
1633    ----------------------------------------------------------------------
1634    Name: Allen Leung
1635    Date: 2000/04/20 23:04:00 EST
1636    Tag: leunga-20000420-ssa-c---stuff
1637    Description:
1638    
1639      This update synchronizes my repository with Yale's.  Most of these
1640    changes, however, do not affect SML/NJ at all (the RA is an exception).
1641    
1642    1. Register Allocator
1643    
1644       a. An improvement in the interference graph construction:
1645          Given a copy
1646    
1647                s <- t
1648    
1649          no interference edge between s and t is added for this definition of s.
1650    
1651       b. I've added two new spill heuristic modules that Fermin and I developed
1652          (in the new library RA.cm). These are unused in SML/NJ but maybe
1653          useful for others (Moby?)
1654    
1655    2. X86
1656    
1657       a. Various fixes in the backend provided by Fermin [C--] and Lal.
1658    
1659    3. Alpha
1660    
1661       a. Added the BSR instruction and code generation that goes with it [C--]
1662       b. Other fixes too numerous to recount provided by Fermin [C--]
1663    
1664    4. Regmaps
1665    
1666       a. The regmaps are not initialized with the identity physical bindings
1667          at creation time.  This is unneeded.
1668    
1669    5. MLRISC Optimizations
1670    
1671       a. The DJ-Graph module can now compute the iterated dominance frontiers
1672          intersects with liveness incrementally in linear time! Woohoo!
1673          This is now used in my new SSA construction algorithm.
1674    
1675       b. THe branch reorganization module is now smarter about linear chains of
1676          basic blocks.
1677    
1678    
1679    ----------------------------------------------------------------------
1680    Name: Matthias Blume
1681    Date: 2000/04/12 13:52:00 JST
1682    Tag: blume_main_v110p27_1
1683    Description:
1684    
1685    Changed install.sh script to handle archive files without version number
1686    and to use "boot.<arch>-<os>" instead of "sml.boot.<arch>-<os>" for the
1687    name of the boot file archive.
1688    
1689    ----------------------------------------------------------------------
1690    Name: Dave MacQueen
1691    Date: 2000/04/09 14:00 EDT
1692    Tag: dbm-20000410-Version_110_27
1693    Description:
1694    
1695    1. Updated src/compiler/TopLevel/main/version.sml to version 110.27
1696    
1697    2. Updated src/config/version to 110.27
1698    
1699    3. New boot files!
1700    
1701    ----------------------------------------------------------------------
1702    Name: Allen Leung
1703    Date: 2000/04/09 19:09:00 EST
1704    Tag: leunga-20000409-misc
1705    Description:
1706    
1707    1.  Yet another fix for x86 assembly for idivl, imull, mull and friends.
1708    
1709    2.  Miscellaneous improvements to MLRISC (unused in sml/nj)
1710    
1711    ----------------------------------------------------------------------
1712    Name: Stefan
1713    Date: 2000/04/07 10:00:00 EDT
1714    Tag: monnier-20000406-branch-handling
1715    Description:
1716    
1717    Improved handling of branches (mostly those generated from
1718    polymorphic equality), removed switchoff and changed the
1719    default optimization settings (more cpsopt and less flintopt).
1720    
1721    ----------------------------------------------------------------------
1722    Name: Allen Leung
1723    Date: 2000/04/06 01:30:00 EST
1724    Tag: leunga-20000406-peephole-x86-SSA-2
1725    Description:
1726    
1727       Forgot a few files.
1728    
1729  ----------------------------------------------------------------------  ----------------------------------------------------------------------
1730  Name: Allen Leung  Name: Allen Leung
1731  Date: 2000/04/06 00:36:00 EST  Date: 2000/04/06 00:36:00 EST
# Line 79  Line 1794 
1794    
1795  3.  Assembly  3.  Assembly
1796    
1797      When generating assemby, resolve the value of client defined constants,      When generating assembly, resolve the value of client defined constants,
1798      instead of generating symbolic values.  This is controlled by the      instead of generating symbolic values.  This is controlled by the
1799      new flag "asm-resolve-constants", which is default to true.      new flag "asm-resolve-constants", which is default to true.
1800    
# Line 102  Line 1817 
1817    
1818  To this end, I arranged that instead of "structure Core" as "structure  To this end, I arranged that instead of "structure Core" as "structure
1819  _Core" is bound in the pervasive environment.  Core access is done via  _Core" is bound in the pervasive environment.  Core access is done via
1820  _Core (which can never be accidentially rebound because _Core is not a  _Core (which can never be accidentally rebound because _Core is not a
1821  legal surface-syntax symbol).  legal surface-syntax symbol).
1822    
1823  The current solution is much cleaner because the core environment is  The current solution is much cleaner because the core environment is
# Line 112  Line 1827 
1827  with dynamic and symbolic parts of the core environment.  with dynamic and symbolic parts of the core environment.
1828    
1829  Remaining hackery (to bind the "magic" symbol _Core) is localized in the  Remaining hackery (to bind the "magic" symbol _Core) is localized in the
1830  compilation mananger's bootstrap compiler (actually: in the "init group"  compilation manager's bootstrap compiler (actually: in the "init group"
1831  handling).  See the comments in src/system/smlnj/init/init.cmi for  handling).  See the comments in src/system/smlnj/init/init.cmi for
1832  more details.  more details.
1833    
# Line 227  Line 1942 
1942     (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
1943     used).     used).
1944    
1945     This was done to accomodate the new "make" and "shell" tools which     This was done to accommodate the new "make" and "shell" tools which
1946     facilitate fairly seemless hookup to portions of code managed using     facilitate fairly seamless hookup to portions of code managed using
1947     Makefiles or Shell scripts.     Makefiles or Shell scripts.
1948    
1949     There are no classes "shared" or "private" anymore.  Instead, the     There are no classes "shared" or "private" anymore.  Instead, the
# Line 240  Line 1955 
1955    
1956     All existing tools are described in the CM manual.     All existing tools are described in the CM manual.
1957    
1958     - Slightly better error handling.  (CM now surpresses many followup     - Slightly better error handling.  (CM now suppresses many followup
1959     error messages that tended to be more annoying than helpful.)     error messages that tended to be more annoying than helpful.)
1960    
1961  2. Major changes to the compiler's static environment data structures.  2. Major changes to the compiler's static environment data structures.
# Line 374  Line 2089 
2089    
2090       I've changed andl to testl in the floating point test sequence       I've changed andl to testl in the floating point test sequence
2091       whenever appropriate.  The Intel optimization guide states that       whenever appropriate.  The Intel optimization guide states that
2092       testl is perferable to andl.       testl is preferable to andl.
2093    
2094  7. RA (x86 only)  7. RA (x86 only)
2095    
# Line 556  Line 2271 
2271    
2272  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
2273  additional argument called "template" which is an optional string that  additional argument called "template" which is an optional string that
2274  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
2275  explanation.  explanation.
2276    
2277  2. A special-purpose tool can be "regisitered" by simply dropping the  2. A special-purpose tool can be "registered" by simply dropping the
2278  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
2279  directory where the .cm file lives that uses this tool.  (The  directory where the .cm file lives that uses this tool.  (The
2280  behavior/misfeature until now was to look for the tool description  behavior/misfeature until now was to look for the tool description
# Line 603  Line 2318 
2318  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
2319  <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,
2320  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
2321  perhaps got accidentially killed).  In this case, fixpt will quickly  perhaps got accidentally killed).  In this case, fixpt will quickly
2322  move through what exists before continuing where it left off earlier,  move through what exists before continuing where it left off earlier,
2323  and, thus, saves a lot of time.  and, thus, saves a lot of time.
2324    
# Line 653  Line 2368 
2368    it from that remote directory.    it from that remote directory.
2369    This should simplify installation further:  For machines that have    This should simplify installation further:  For machines that have
2370    access to the internet, just fetch <version>-config.tgz, unpack it,    access to the internet, just fetch <version>-config.tgz, unpack it,
2371    edit config/targets, and go (run config/install.sh).  The scipt will    edit config/targets, and go (run config/install.sh).  The script will
2372    fetch everything else that it might need all by itself.    fetch everything else that it might need all by itself.
2373    
2374    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.601  
changed lines
  Added in v.796

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