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 585, Wed Mar 29 23:55:35 2000 UTC revision 813, Fri Apr 20 15:13:12 2001 UTC
# Line 13  Line 13 
13  Description:  Description:
14    
15  ----------------------------------------------------------------------  ----------------------------------------------------------------------
16    Name: Matthias Blume
17    Date: 2001/04/20 11:10:00 EDT
18    Tag: blume-20010420-inMLflag
19    Description:
20    
21     - added vp_limitPtrMask to vproc-state.h
22       (for use by the raw-C-calls mechanism to implement proper interrupt
23        handling)
24     - made the ML compiler aware of various data-structure offsets so it
25       can generate code for accessing the vp_inML flag and vp_limitPtrMask
26     - tweaked mlriscGen.sml to have it emit interrupt-handling code for
27       raw C-calls
28    
29    ----------------------------------------------------------------------
30    Name: Lal George
31    Date: 2001/04/20 09:15:28 EDT
32    Tag: george-20010420-macosX
33    Description:
34    
35     - Changes to port to Mac OS X; Darwin.
36    
37     - In the process I found that sqrt was broken on the PPC, because the
38       fsqrt instruction is not implemented.
39    
40    ----------------------------------------------------------------------
41    Name: Matthias Blume
42    Date: 2001/04/18 12:45:00 EDT
43    Tag: blume-20010418-ccalls
44    Description:
45    
46     - fixed two off-by-4 errors in the x86-specific c-calls implementation
47       (this bug prevented structure arguments containing pointers from being
48        passed correctly)
49     - changed the raw-C-call code in mlriscGen.sml in such a way that
50       structure arguments are represented as a pointer to the beginning
51       of the structure (instead of having a series of synthesized arguments,
52       one for each structure member)
53    
54     - made makeml script's verbosity level configurable via environment
55       variable (MAKEML_VERBOSITY)
56    
57     - eliminated placeholder implementations for f32l, w16s, i16s, and f32s
58       in rawmem-x86.sml; we are now using the real thing
59    
60    ----------------------------------------------------------------------
61    Name: Matthias Blume
62    Date: 2001/03/22 16:25:00 EST
63    Tag: blume-20010322-bootfiles
64    Description:
65    
66    Created a new set of bootfiles (for your automatic installation convenience).
67    
68    ----------------------------------------------------------------------
69    Name: Matthias Blume
70    Date: 2001/03/22 15:10:00 EST
71    Tag: blume-20010322-rawmem-parcm
72    Description:
73    
74    1. All "raw memory access" primitives for the new FFI are implemented now
75       (at least on the x86).
76    2. Some further cleanup of CM's parallel make mechanism.
77    
78    ----------------------------------------------------------------------
79    Name: Matthias Blume
80    Date: 2001/03/19 17:53:00 EST
81    Tag: blume-20010319-parallel
82    Description:
83    
84    Parallel make (using compile servers) now works again.
85    
86    To this end, CM.stabilize and CMB.make have been modified to work in
87    two passes when compile servers are attached:
88       1. Compile everything, do not perform stabilization; this pass
89          uses compile servers
90       2. Stabilize everything; this pass does not use compile servers
91    If there are no compile servers, the two passes are combined into one
92    (as before).  Splitting the passes increases the inherent parallelism
93    in the dependency graph because the entire graph including all
94    libraries is available at the same time.  This, in turn, improves
95    server utilization.  The downside is that the master process will
96    have to do some extra work after compilation is done (because for
97    technical reasons it must re-read all the binfiles during stabilization).
98    
99    ----------------------------------------------------------------------
100    Name: Matthias Blume
101    Date: 2001/03/16 12:22:00 EST
102    Tag: blume-20010316-bootfiles
103    Description:
104    
105    Created a new set of bootfiles (for your automatic installation convenience).
106    
107    ----------------------------------------------------------------------
108    Name: Matthias Blume
109    Date: 2001/03/16 11:00:00 EST
110    Tag: blume-20010316-MLTREE-fixup
111    Description:
112    
113    This is a minor fixup for an (untagged) earlier commit by Allen.
114    (A file was missing).
115    
116    ----------------------------------------------------------------------
117    Name: Allen Leung
118    Date: Mon Mar  5 18:54:57 EST 2001
119    Tag: leunga-20010305-cut-support
120    
121    1. New support for alternative control-flow in MLTREE.
122       Currently we support
123    
124          FLOW_TO(CALL ...., [k1,...,kn])
125    
126       This is needed for 'cuts to' in C-- and try/handle-like constructs
127       in Moby
128    
129       New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
130    
131    2. Register Allocator
132    
133       Changes in interface [from Fermin, John]
134    
135    3. Alpha 8-bit SLL support [Fermin]
136    
137    4. All architectures
138    
139       A new module (ClusterExpandCopies) for expanding parallel copies.
140    
141    ----------------------------------------------------------------------
142    Name: Allen Leung
143    Date: 2001/02/27 23:07:00 EST
144    Tag: leunga-20010227-minor-stuff
145    
146    1. Alpha bug fix for CMOVNE
147    2. Handle mltree COND(..,FCMP ...,...)
148    3. Bug fix in simplifier
149    
150    ----------------------------------------------------------------------
151    Name: Matthias Blume
152    Date: 2001/01/30 17:50:00 EST
153    Tag: blume-20010130-sync
154    Description:
155    
156    This is just a minor update to sync my devel branch with the main brach.
157    The only visible change is the addition of some README files.
158    
159    ----------------------------------------------------------------------
160    Name: Matthias Blume
161    Date: 2001/01/12 23:30:00 JST
162    Tag: blume-20010112-bootfiles
163    Description:
164    
165    Made a new set of bootfiles that goes with the current state of the
166    repository.
167    
168    ----------------------------------------------------------------------
169    Name: Matthias Blume
170    Date: 2001/01/12 21:20:00 JST
171    Tag: blume-20010112-sync
172    Description:
173    
174    I am just flushing out some minor changes that had accumulated in
175    my private branch in order to sync with the main tree.  (This is
176    mainly because I had CVS trouble when trying to merge _into_ my
177    private branch.)
178    
179    Most people should be completely unaffected by this.
180    
181    ----------------------------------------------------------------------
182    Name: Allen Leung
183    Date: Thu Jan 11 21:03:00 EST 2001
184    Tag: leunga-20010111-labexp=mltree
185    Description:
186    
187    1.  Removed the type LabelExp and replace it by MLTree.
188    2.  Rewritten mltree-simplify with the pattern matcher tool.
189    3.  There were some bugs in alpha code generator which would break
190        64-bit code generation.
191    4.  Redo the tools to generate code with the
192    5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
193        are now generated by perl scripts.
194    
195    ----------------------------------------------------------------------
196    Name: Matthias Blume
197    Date: 2001/01/10 21:55:00 JST
198    Tag: blume-20010110-rcc
199    Description:
200    
201    The RCC stuff now seems to work (but only on the x86).
202    This required hacking of the c-calls interface (and -implementation) in
203    MLRISC.
204    
205    Normal compiler users should be unaffected.
206    
207    ----------------------------------------------------------------------
208    Name: Matthias Blume
209    Date: 2001/01/09 01:20:00 JST
210    Tag: blume-20010109-rcc
211    Description:
212    
213    This is a fairly big patch, flushing out a large number of pending
214    changes that I made to my development copy over the last couple of days.
215    
216    Of practical relevance at this moment is a workaround for a pickling
217    bug that Allen ran into the other day.  The cause of the bug itself is
218    still unknown and it might be hard to fix it properly, but the
219    workaround has some merits of its own (namely somewhat reducing pickling
220    overhead for certain libraries).  Therefore, I think this solution should
221    be satisfactory at this time.
222    
223    The rest of the changes (i.e., the vast majority) has to do with my
224    ongoing efforts of providing direct support for C function calls from
225    ML.  At the moment there is a new primop "RAW_CCALL", typing magic
226    in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
227    case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
228    translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
229    handle RCC.
230    
231    The last part (the changes to mlriscGen.sml) are still known to be
232    wrong on the x86 and not implemented on all other architectures.  But
233    the infrastructure is in place. I had to change a few functor
234    signatures in the backend to be able to route the CCalls interface
235    from MLRISC there, and I had to specialize the mltree type (on the
236    x86) to include the necessary extensions. (The extensions themselves
237    were already there and redy to go in MLRISC/x86).
238    
239    Everything should be very happy as soon as someone helps me with
240    mlriscGen.sml...
241    
242    In any case, nothing of this should matter to anyone as long as the
243    new primop is not being used (which is going to be the case unless you
244    find it where I hid it :). The rest of the compiler is completely
245    unaffected.
246    
247    ----------------------------------------------------------------------
248    Name: Matthias Blume
249    Date: 2001/01/05 00:30:00 JST
250    Tag: blume-20010105-primops
251    Description:
252    
253    Added some experimental support for work that I am doing right now.
254    These changes mostly concern added primops, but there is also a new
255    experimental C library in the runtime system (but currently not enabled
256    anywhere except on Linux/X86).
257    
258    In the course of adding primops (and playing with them), I discovered that
259    Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
260    broken.  (Zhong was very right he labeled this stuff as "major gross hack".)
261    To recover, I made type information in INL_PRIM mandatory and changed
262    prim.sml as well as built-in.sml accordingly.  The InLine structure now
263    has complete, correct type information (i.e., no bottom types).
264    
265    Since all these changes mean that we need new binfiles, I also bumped the
266    version number to 110.32.1.
267    
268    ----------------------------------------------------------------------
269    Name: Matthias Blume
270    Date: 2000/12/30 22:10:00 JST
271    Tag: blume-20001230-various
272    Description:
273    
274    Added proxy libraries for MLRISC and let MLRISC libraries refer
275    to each other using path anchors.  (See CM manual for explanation.)
276    
277    Updated CM documentation.
278    
279    Fixed some bugs in CM.
280    
281    Implemented "proxy" libraries (= syntactic sugar for CM).
282    
283    Added "-quiet" option to makeml and changed runtime system accordingly.
284    
285    Added cleanup handler for exportML to reset timers and compiler stats.
286    
287    ----------------------------------------------------------------------
288    Name: Lal George
289    Date: 2000/12/22 22:22:58 EST 2000
290    Tag: Release_110_32
291    Description:
292    
293            Infinite precision used throughout MLRISC.
294            see MLRISC/mltree/machine-int.sig
295    
296    ----------------------------------------------------------------------
297    Name: Matthias Blume
298    Date: 2000/12/22 23:16:00 JST
299    Tag: blume-20001222-warn
300    Description:
301    
302    Corrected wording and formatting of some CM warning message which I
303    broke in my previous patch.
304    
305    ----------------------------------------------------------------------
306    Name: Matthias Blume
307    Date: 2000/12/22 21:20:00 JST
308    Tag: blume-20001222-anchorenv
309    Description:
310    
311    Fixed CM's handling of anchor environments in connection with CMB.make.
312    
313    ----------------------------------------------------------------------
314    Name: Matthias Blume
315    Date: 2000/12/22 13:15:00 JST
316    Tag: blume-20001222-cleanup
317    Description:
318    
319    Removed src/cm/ffi which does not (and did not) belong here.
320    
321    ----------------------------------------------------------------------
322    Name: Matthias Blume
323    Date: 2000/12/21 23:55:00 JST
324    Tag: blume-20001221-exn
325    Description:
326    
327    Probably most important: CM no longer silently swallows all exceptions
328    in the compiler.
329    Plus: some other minor CM changes.  For example, CM now reports some
330    sizes for generated binfiles (code, data, envpickle, lambdapickle).
331    
332    ----------------------------------------------------------------------
333    Name: Matthias Blume
334    Date: 2000/12/15 00:01:05 JST
335    Tag: blume-20001215-dirtool
336    Description:
337    
338    - "dir" tool added.
339    - improvements and cleanup to Tools structure
340    - documentation updates
341    
342    ----------------------------------------------------------------------
343    Name: Allen Leung
344    Date: Thu Dec 14 03:45:24 EST 2000
345    Description:
346    Tag:  leunga-20001214-int-inf
347    Description:
348    
349       In IntInf, added these standard functions, which are missing from our
350    implementation:
351    
352        andb : int * int -> int
353        xorb : int * int -> int
354        orb  : int * int -> int
355        notb : int -> int
356         <<   : int * word -> int
357        ~>>  : int * word -> int
358    
359       Not tested, I hope they are correct.
360    
361    ----------------------------------------------------------------------
362    Name: Allen Leung
363    Date: Fri Dec  8 19:23:26 EST 2000
364    Description:
365    Tag:  leunga-20001208-nowhere
366    Description:
367    
368      Slight improvements to the 'nowhere' tool to handle OR-patterns,
369    to generate better error messages etc.  Plus a brief manual.
370    
371    ----------------------------------------------------------------------
372    Name: Lal George
373    Date: 2000/12/08 09:54:02 EST 2000
374    Tag: Release_110_31
375    Description:
376    
377    - Version 110.31
378    ----------------------------------------------------------------------
379    Name: Allen Leung
380    Date: Thu Dec  7 22:01:04 EST 2000
381    Tag:  leunga-20001207-cell-monster-hack
382    Description:
383    
384    Major MLRISC internal changes.  Affect all clients.
385    Summary:
386    
387    1.  Type CELLS.cell = int is now replaced by a datatype.
388        As a result, the old regmap is now gone.  Almost all interfaces
389        in MLRISC change as a consequence.
390    
391    2.  A new brand version of machine description tool (v3.0) that generates
392        modules expecting the new interface.  The old version is removed.
393    
394    3.  The RA interface has been further abstracted into two new functors.
395        RISC_RA and X86RA.  These functors have much simpler interfaces.
396        [See also directory MLRISC/demo.]
397    
398    4.  Some other new source->source code generation tools are available:
399    
400        a. MLRISC/Tools/RewriteGen -- generate rewriters from rules.
401        b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules.
402           I use this tool to generate the peephole optimizers---with the new
403           cell type changes, peephole rules are becoming difficult to write
404           without conditional pattern matching.
405    
406    5.  More Intmap -> IntHashTable change.  Previous changes by Matthias didn't
407        cover the entire MLRISC source tree so many things broke.
408    
409    6.  CM files have been moved to the subdirectory MLRISC/cm.
410        They are moved because there are a lot of them and they clutter up the
411        root dir.
412    
413    7.  More detailed documentation to come...
414    
415        NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do
416        a makeml -rebuild first.  This is because of other other
417        changes that Matthias has made (see below).
418    
419    
420    ----------------------------------------------------------------------
421    Name: Matthias Blume
422    Date: 2000/11/30 23:12:00 JST
423    Tag: blume-20001130-filereorg
424    Description:
425    
426    Some manual updates and some file reorganizations in CM.
427    
428    ----------------------------------------------------------------------
429    Name: Matthias Blume
430    Date: 2000/11/24 17:45:00 JST
431    Tag: blume-20001124-link
432    Description:
433    
434    Drastically improved link traversal code for the case that the dynamic
435    value was already loaded at bootstrap time.  As a result, CM and CMB
436    now both load blazingly fast -- even on a very slow machine.  Also,
437    memory consumption has been further reduced by this.
438    
439    Warning: The format of the PIDMAP file has changed.  THerefore, to
440    bootstrap you have to do this:
441    
442    1. Run CMB.make
443    2. Make a symbolic link for the boot directory:
444         ln -s sml.boot.ARCH-OS xxx
445    3. "Rebuild" the boot directory:
446         ./makeml -boot xxx -rebuild sml ; rm xxx
447    4. Boot normally:
448          ./makeml
449    
450    ----------------------------------------------------------------------
451    Name: Matthias Blume
452    Date: 2000/11/21 21:20:00 JST
453    Tag: blume-20001121-tools
454    Description:
455    
456    Continued hacking on autoloading problem -- with success this time.
457    Also changed tool-plugin mechanism.  See new CM manual.
458    
459    ----------------------------------------------------------------------
460    Name: Matthias Blume
461    Date: 2000/11/19 14:30:00 JST
462    Tag:  blume-20001119-autoload
463    Description:
464    
465    Some hacking to make autoloading faster.  Success for CMB, no success
466    so far for CM.  There is a reduced structure CM' that autoloads faster.
467    (This is a temporary, non-documented hack to be eliminated again when
468    the general problem is solved.)
469    
470    ----------------------------------------------------------------------
471    Name: Matthias Blume
472    Date: 2000/11/17 14:10:00 JST
473    Tag: blume-20001117-pickle-lib
474    Description:
475    
476    1. Eliminated comp-lib.cm
477    2. Made pickle-lib.cm
478    3. Eliminated all uses of intset.sml (from comp-lib.cm)
479    4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
480       equivalent constructs from smlnj-lib.cm (INtHashTable).
481    5. Point 4. also goes for those uses of intmap.* in MLRISC.
482       Duplicated intmap modules thrown out.
483    6. Hunted down all duplicated SCC code and replaced it with
484       equivalent stuff (GraphSCCFn from smlnj-lib.cm).
485    7. Rewrote Feedback module.
486    8. Moved sortedlist.sml into viscomp-lib.cm.  Eventually it
487       should be thrown out and equivalent modules from smlnj-lib.cm
488       should be used (IntRedBlackSet, IntListSet, ...).
489    
490    Confirmed that compiler compiles to fixpoint.
491    
492    ----------------------------------------------------------------------
493    Name: Allen Leung
494    Date: 2000/11/10 18:00:00
495    Tag: leunga-20001110-new-x86-fp
496    
497    A new x86 floating point code generator has been added.
498    By default this is turned off.  To turn this on, do:
499    
500        CM.autoload "$smlnj/compiler.cm";
501        Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
502    
503    Changes:
504    
505    1.  Changed FTAN to FPTAN so that the assembly output is correct.
506    2.  Changed the extension callback for FTANGENT to generate:
507    
508              fptan
509              fstp  %st(0)
510        instead of
511              fptan
512              fstpl ftempmem
513    
514    3.  Numerous assembly fixes for x86.
515    
516    5.  Cleaned up the machine code output module x86/x86MC.sml and added
517        support for a whole bunch of instructions and addressing modes:
518    
519          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st, %st(n)
520          faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp  %st, %st(n)
521          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st(n), %st
522          fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
523          fxch %st(n)
524          fld %st(n)
525          fst %st(n)
526          fst mem
527          fstp %st(n)
528          fucom %st(n)
529          fucomp %st(n)
530    
531        All these are now generated when the fast fp mode is turned on.
532    
533    6.  Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
534    
535    ----------------------------------------------------------------------
536    Name: Matthias Blume
537    Date: 2000/11/09 11:20:00 JST
538    Tag: blume-20001109-scc
539    Description:
540    
541    Eliminated some code duplication:
542    
543    1. Added "where" clause to GraphSCCFn in SML/NJ Library.
544       (Otherwise the functor is useless.)
545    2. Used GraphSCCFn where SCCUtilFun was used previously.
546    3. Got rid of SCCUtilFun (in comp-lib.cm).
547    
548    ----------------------------------------------------------------------
549    Name: Lal George
550    Date: 2000/11/06 09:02:21 EST 2000
551    Tag: Release_110_30
552    Description:
553    
554    - Version 110.30
555    ----------------------------------------------------------------------
556    Name: Matthias Blume
557    Date: 2000/11/04 14:45:00
558    Tag: blume-20001104-mlbuild
559    Description:
560    
561    - Made ml-build faster on startup.
562    - Documentation fixes.
563    
564    ----------------------------------------------------------------------
565    Name: Matthias Blume
566    Date: 2000/11/02 17:00:00 JST
567    Tag: blume-20001102-condcomp
568    Description:
569    
570    - Small tweaks to pickler -- new BOOTFILES!
571    - Version bumped to 110.29.2.
572    - Added conditional compilation facility to init.cmi (see comment there).
573    ----------------------------------------------------------------------
574    Name: Allen Leung
575    Date: 2000/10/23 19:31:00
576    Tag: leunga-20001023-demo-ra
577    
578    1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
579    2. Test programs for the graph library updated
580    3. Some new MLRISC demo programs added
581    
582    ----------------------------------------------------------------------
583    Name: Matthias Blume
584    Date: 2000/08/31 22:15:00 JST
585    Tag: blume-20001017-errmsg
586    Description:
587    
588    More error message grief: Where there used to be no messages, there
589    now were some that had bogus error regions.  Fixed.
590    
591    ----------------------------------------------------------------------
592    Name: Matthias Blume
593    Date: 2000/08/31 17:30:00 JST
594    Tag: blume-20001017-v110p29p1
595    Description:
596    
597    I made a version 110.29.1 with new bootfiles.
598    
599    Changes:  Modified pickler/unpickler for faster and leaner unpickling.
600              CM documentation changes and a small bugfix in CM's error reporting.
601    
602    ----------------------------------------------------------------------
603    Name: Lal George
604    Date: 2000/09/27 14:42:35 EDT
605    Tag: george-20000927-nodestatus
606    Description:
607    
608    Changed the type of the nodestatus, so that:
609    
610            SPILLED(~1)             is now SPILLED
611            SPILLED(m) where m>=0   is now MEMREG(m)
612            SPILLED(s) where s<~1   is now SPILL_LOC(~s)
613    
614    ----------------------------------------------------------------------
615    Name: Matthias Blume
616    Date: 2000/09/07 14:45:00 JST
617    Tag: blume-20000907-cmerrmsg
618    Description:
619    
620    Small tweak to CM to avoid getting ML syntax error messages twice.
621    
622    ----------------------------------------------------------------------
623    Name: Matthias Blume
624    Date: 2000/08/31 18:00:00 JST
625    Tag: blume-20000831-cvsbootfiles
626    Description:
627    
628    New URL for boot files (because the 110.29 files on the BL server do
629    now work correctly with my updated install scripts for yacc and lex).
630    
631    ----------------------------------------------------------------------
632    Name: Matthias Blume
633    Date: 2000/08/08 12:33:00 JST
634    Tag: blume-20000808-manual
635    Description:
636    
637    Tiny update to CM manual.
638    
639    ----------------------------------------------------------------------
640    Name: Allen Leung
641    Date: 2000/08/7 19:31:00
642    Tag: leunga-20000807-a-whole-bunch-of-stuff
643    
644      Moby, C--, SSA, x86, machine descriptions etc.  Should only affect C--
645    and Mobdy.
646    
647    1.  x86
648    
649       a.  Fixes to peephole module by John and Dan.
650       b.  Assembly fix to SETcc by Allen.
651       c.  Fix to c-call by John.
652       d.  Fix to spilling by John.  (This one deals with the missing FSTPT case)
653       e.  Instruction selection optimization to SETcc as suggested by John.
654    
655           For example,
656    
657            MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
658    
659           should generate:
660    
661            MOVL a, x
662            SUBL b, x
663            SHRL 31, x
664    
665    2.  IR stuff
666    
667         A bunch of new DJ-graph related algorithms added.  These
668         speed up SSA construction.
669    
670    3.  SSA + Scheduling
671    
672         Added code for SSA and scheduling to the repository
673    
674    ----------------------------------------------------------------------
675    Name: Lal George
676    Date: 2000/07/27 11:53:14 EDT
677    
678    Tag: lal-20000727-linux-ppc
679    Description:
680    
681     Made changes to support Linux PPC.
682     p.s. I have confirmation that the 110.29 boot files work fine.
683    
684    ----------------------------------------------------------------------
685    Name: Matthias Blume
686    Date: 2000/07/27 17:40:00 JST
687    Tag: blume-20000727-scripts
688    Description:
689    
690    !!!! WARNING !!!!
691    You must recompile the runtime system!
692    !!!! WARNING !!!!
693    
694    This is basically another round of script-enhancements:
695    
696    1. sml, ml-build, and ml-makedepend accept options -D and -U to define
697       and undefine CM preprocessor symbols.
698    
699    2. ml-build avoids generating a new heap image if it finds that the
700       existing one is still ok.  (The condition is that no ML file had to
701       be recompiled and all ML files are found to be older that the heap
702       file.)
703    
704       To make this work smoothly, I also hacked the runtime system as
705       well as SMLofNJ.SysInfo to get access to the heap image suffix
706       (.sparc-solaris, ...) that is currently being used.
707    
708       Moreover, the signature of CM.mk_standalone has changed.  See the
709       CM manual.
710    
711    3. ml-makedepend accepts additional options -n, -a, and -o.  (See the
712       CM manual for details.)
713    
714    4. More CM manual updates:
715        - all of the above has been documented.
716        - there is now a section describing the (CM-related) command line
717          arguments that are accepted by the "sml" command
718    
719    ----------------------------------------------------------------------
720    Name: Matthias Blume
721    Date: 2000/07/25 16:20:00 JST
722    Tag: blume-20000725-makedepend
723    Description:
724    
725    Added a script called ml-makedepend.  This can be used in makefiles
726    for Unix' make in a way very similar to the "makedepend" command for
727    C.
728    
729    The script internally uses function CM.sources.
730    
731    Synopsis:
732    
733        ml-makedepend [-f makefile] cmfile targetname
734    
735    The default for the makefile is "makefile" (or "Makefile" should
736    "makefile" not exist).
737    
738    ml-makedepend adds a cmfile/targetname-specific section to this
739    makefile (after removing the previous version of this section).  The
740    section contains a single dependency specification with targetname on
741    the LHS (targetname is an arbitrary name), and a list of files derived
742    from the cmfile on the RHS.  Some of the files on the RHS are
743    ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
744    "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
745    path names.  The makefile writer is responsible for making sure that
746    these variables have correct at the time "make" is invoked.
747    
748    ----------------------------------------------------------------------
749    Name: Matthias Blume
750    Date: 2000/07/22 23:30:00 JST
751    Tag: blume-20000722-urlupdate
752    Description:
753    
754    Changed BOOT and config/srcarchiveurl to point to BL server:
755    
756        ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
757    
758    ----------------------------------------------------------------------
759    Name: Matthias Blume
760    Date: 2000/07/18 18:00:00 JST
761    Tag: blume-20000718-Version_110_29
762    Description:
763    
764    1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
765    
766    2. Updated config/version to 110.29
767    
768    3. Updated config/srcarchiveurl
769    
770    3. New boot files!
771       ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
772    
773    ----------------------------------------------------------------------
774    Name: Matthias Blume
775    Date: 2000/07/11 13:58:00 JST
776    Tag: blume-20000711-doctypo
777    Description:
778    
779    Fixed a few typos in CM manual.
780    
781    ----------------------------------------------------------------------
782    Name: Allen Leung
783    Date: 2000/06/15 00:38:00
784    Tag: leunga-20000704-sparc-x86
785    
786    1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
787    2. fix to x86 RET bug [found by Dan Grossman]
788    3. sparc assembly bug fix for ticc instructions [found by Fermin]
789    
790       Affects c-- and moby only
791    
792    ----------------------------------------------------------------------
793    Name: Matthias Blume
794    Date: 2000/07/04 15:26:00
795    Tag: blume-20000704-trigger
796    Description:
797    
798    1. Improvements to CM manual.
799    2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
800       of getting a back-trace.  The function, when called, raises an
801       internal exception which explicitly carries the full back-trace history,
802       so it is unaffected by any intervening handle-raise pairs ("trivial"
803       or not).  The interactive loop will print that history once it arrives
804       at top level.
805       Short of having all exceptions implicitly carry the full history, the
806       recommended way of using this facility is:
807         - compile your program with instrumentation "on"
808         - run it, when it raises an exception, look at the history
809         - if the history is "cut off" because of some handler, go and modify
810           your program so that it explicitly calls BTrace.trigger
811         - recompile (still instrumented), and rerun; look at the full history
812    
813    ----------------------------------------------------------------------
814    Name: Matthias Blume
815    Date: 2000/07/03 15:36:00 JST
816    Tag: blume-20000702-manual
817    Description:
818    
819    Small corrections and updates to CM manual.
820    
821    ----------------------------------------------------------------------
822    Name: Matthias Blume
823    Date: 2000/06/29 16:04:00 JST
824    Tag: blume-20000629-yacctool
825    Description:
826    
827    Changes:
828    
829    1. Class "mlyacc" now takes separate arguments to pass options to
830       generated .sml- and .sig-files independently.
831    2. Corresponding CM manual updates.
832    3. BTrace module now also reports call sites.  (However, for loop clusters
833       it only shows from where the cluster was entered.)  There are associated
834       modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
835    
836    ----------------------------------------------------------------------
837    Name: Matthias Blume
838    Date: 2000/06/27 16:51:00 JST
839    Tag: blume-20000627-noweb
840    Description:
841    
842    Changes:
843    
844     1. Implemented "subdir" and "witness" options for noweb tool.
845        This caused some slight internal changes in CM's tool implementation.
846     2. Fixed bug in "tool plugin" mechanism.  This is essentially cleaning
847        some remaining issues from earlier path anchor changes.
848     3. Updated CM manual accordingly.
849    
850     4. Changed implementation of back-tracing so that I now consider it
851        ready for prime-time.
852    
853        In particular, you don't have to explicitly trigger the back-trace
854        anymore.  Instead, if you are running BTrace-instrumented code and
855        there is an uncaught exception (regardless of whether or not it was
856        raised in instrumented code), the top-level evalloop will print
857        the back-trace.
858    
859        Features:
860    
861          - Instrumented and uninstrumented code work together seemlessly.
862            (Of course, uninstrumented code is never mentioned in actual
863             back-traces.)
864    
865          - Asymptotic time- and space-complexity of instrumented code is
866            equal to that of uninstrumented code.  (This means that
867            tail-recursion is preserved by the instrumentation phase.)
868    
869          - Modules whose code has been instrumented in different sessions
870            work together without problem.
871    
872          - There is no penalty whatsoever on uninstrumented code.
873    
874          - There is no penalty on "raise" expressions, even in
875            instrumented code.
876    
877        A potential bug (or perhaps it is a feature, too):
878    
879          A back-trace reaches no further than the outermost instrumented
880          non-trivial "raise".  Here, a "trivial" raise is one that is the
881          sole RHS of a "handle" rule.  Thus, back-traces reach trough
882    
883               <exp> handle e => raise e
884    
885          and even
886    
887               <exp> handle Foo => raise Bar
888    
889          and, of course, through
890    
891               <exp> handle Foo => ...
892    
893         if the exception was not Foo.
894    
895         Back-traces always reach right through any un-instrumented code
896         including any of its "handle" expressions, trivial or not.
897    
898       To try this out, do the following:
899    
900         - Erase all existing binfiles for your program.
901           (You may keep binfiles for those modules where you think you
902            definitely don't need back-tracing.)
903         - Turn on back-trace instrumentation:
904              SMLofNJ.Internals.BTrace.mode (SOME true);
905         - Recompile your program.  (I.e., run "CM.make" or "use".)
906         - You may now turn instrumentation off again (if you want):
907              SMLofNJ.Internals.BTrace.mode (SOME false);
908         - Run your program as usual.  If it raises an exception that
909           reaches the interactive toplevel, then a back-trace will
910           automatically be printed.  After that, the toplevel loop
911           will print the exception history as usual.
912    
913    ----------------------------------------------------------------------
914    Name: Matthias Blume
915    Date: 2000/06/26 09:56:46 JST
916    Tag: blume-20000626-setup
917    Description:
918    
919    CM: - setup-parameter to "sml" added; this can be used to run arbitrary
920          ML code before and after compiling a file (e.g., to set compiler
921          flags)
922    
923    Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
924              - associated changes to btrace.sml (BTrace instrumentation pass)
925              - cleaner implementation of btimp.sml (BTrace tracing and report
926                module)
927    
928    CM manual: * new path encoding documented
929               * description of setup-parameter to "sml" added
930    
931    The biggest user-visible change to back-tracing is that it is no
932    longer necessary to compile all traced modules within the same
933    session.  (This was a real limitation.)
934    
935    ----------------------------------------------------------------------
936    Name: Matthias Blume
937    Date: 2000/06/24 12:40:00 JST
938    Tag: blume-20000624-startup
939    Description:
940    
941    Fixes startup slowdown problem.  (I was calling SrcPath.sync a _tad_
942    bit too often -- to put it mildly. :)
943    
944    ----------------------------------------------------------------------
945    Name: Matthias Blume
946    Date: 2000/06/23 18:20:00 JST
947    Tag: blume-20000623-btrace
948    Description:
949    
950    This updates adds a backtrace facility to aid programmers in debugging
951    their programs.  This involves the following changes:
952    
953    1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
954       keeping track of the current call stack.  When programs are compiled
955       in a special mode, the compiler will insert calls to these hooks
956       into the user program.
957       "Hook" means that it is possible for different implementations of
958       back-tracing to register themselves (at different times).
959    
960    2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
961       as an Absyn.dec->Absyn.dec rewrite.  Normally this phase is turned off.
962       It can be turned on using this call:
963         SMLofNJ.Internals.BTrace.mode (SOME true);
964       Turning it off again:
965         SMLofNJ.Internals.BTrace.mode (SOME false);
966       Querying the current status:
967         SMLofNJ.Internals.BTrace.mode NONE;
968       Annotated programs are about twice as big as normal ones, and they
969       run a factor of 2 to 4 slower with a dummy back-trace plugin (one
970       where all hooks do nothing).  The slowdown with a plugin that is
971       actually useful (such as the one supplied by default) is even greater,
972       but in the case of the default plugin it is still only an constant
973       factor (amortized).
974    
975    3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
976       with a sub-structure BTrace for controlling back-tracing.  In particular,
977       the above-mentioned function "mode" controls whether the annotation
978       phase is invoked by the compiler.  Another important function is
979       "trigger": when called it aborts the current execution and causes
980       the top-level loop to print a full back-trace.
981    
982    4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
983       for back-tracing.  It keeps track of the dynamic call stack and in
984       addition to that it keeps a partial history at each "level" of that
985       stack.  For example, if a tail-calls b, b tail-calls c, and c tail-calls
986       d and b (at separate times, dynamically), then the report will show:
987    
988       GOTO   d
989             /c
990       GOTO  \b
991       CALL   a
992    
993       This shows that there was an initial non-tail call of a, then a
994       tail-call to b or c, looping behavior in a cluster of functions that
995       consist of b and c, and then a goto from that cluster (i.e., either from
996       b or from c) to d.
997    
998       Note that (depending on the user program) the amount of information
999       that the back-trace module has to keep track of at each level is bounded
1000       by a constant.  Thus, the whole implementation has the same asymptotical
1001       complexity as the original program (both in space and in time).
1002    
1003    5. compiler/TopLevel/interact/evalloop.sml has been modified to
1004       handle the special exception SMLofNJ.Internals.BTrace.BTrace
1005       which is raised by the "trigger" function mentioned above.
1006    
1007    Notes on usage:
1008    
1009    - Annotated code works well together with unannotated code:
1010    Unannotated calls simply do not show up at all in the backtrace.
1011    
1012    - It is not a good idea to let modules that were annotated during
1013    different sessions run at the same time.  This is because the compiler
1014    chooses small integers to identify individual functions, and there
1015    will be clashes if different modules were compiled in separate sessions.
1016    (Nothing will crash, and you will even be told about the clashes, but
1017    back-trace information will in general not be useful.)
1018    
1019    - Back-tracing can be confused by callcc and capture.
1020    
1021    - The only way of getting a back-trace right now is to explicitly
1022    invoke the "trigger" function from your user program.  Eventually, we
1023    should make every exception carry back-trace information (if
1024    available).  But since this creates more overhead at "raise"-time
1025    (similar to the current exnHistory overhead), I have not yet
1026    implemented this.  (The implementation will be rather easy.)  With
1027    exceptions carrying back-trace information, this facility will be even
1028    more useful because users don't need to modify their programs...
1029    
1030    - While it is possible to compile the compiler with back-trace
1031    annotations turned on (I did it to get some confidence in
1032    correctness), you must make absolutely sure that core.sml and
1033    btimp.sml are compiled WITHOUT annotation!  (core.sml cannot actually
1034    be compiled with annotation because there is no core access yet, but
1035    if you compile btimp.sml with annotation, then the system will go into
1036    an infinite recursion and crash.)
1037    Since CM currently does not know about BTrace, the only way to turn
1038    annotations on and off for different modules of the compiler is to
1039    interrupt CMB.make, change the settings, and re-invoke it.  Of course,
1040    this is awkward and clumsy.
1041    
1042    Sample sessions:
1043    
1044    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
1045    - SMLofNJ.Internals.BTrace.mode (SOME true);
1046    [autoloading]
1047    [autoloading done]
1048    val it = false : bool
1049    - structure X = struct
1050    -     fun main n = let
1051    -         fun a (x, 0) = d x
1052    -           | a (x, n) = b (x, n - 1)
1053    -         and b (x, n) = c (x, n)
1054    -         and c (x, n) = a (x, n)
1055    -         and d x = e (x, 3)
1056    -         and e (x, 0) = f x
1057    -           | e (x, n) = e (x, n - 1)
1058    -         and f 0 = SMLofNJ.Internals.BTrace.trigger ()
1059    -           | f n = n * g (n - 1)
1060    -         and g n = a (n, 3)
1061    -     in
1062    -         f n
1063    -     end
1064    - end;
1065    structure X : sig val main : int -> int end
1066    - X.main 3;
1067    *** BACK-TRACE ***
1068    GOTO   stdIn:4.2-13.20: X.main[2].f
1069    GOTO-( stdIn:4.2-13.20: X.main[2].e
1070    GOTO   stdIn:4.2-13.20: X.main[2].d
1071         / stdIn:4.2-13.20: X.main[2].a
1072         | stdIn:4.2-13.20: X.main[2].b
1073    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1074    CALL   stdIn:4.2-13.20: X.main[2].g
1075    GOTO   stdIn:4.2-13.20: X.main[2].f
1076    GOTO-( stdIn:4.2-13.20: X.main[2].e
1077    GOTO   stdIn:4.2-13.20: X.main[2].d
1078         / stdIn:4.2-13.20: X.main[2].a
1079         | stdIn:4.2-13.20: X.main[2].b
1080    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1081    CALL   stdIn:4.2-13.20: X.main[2].g
1082    GOTO   stdIn:4.2-13.20: X.main[2].f
1083    GOTO-( stdIn:4.2-13.20: X.main[2].e
1084    GOTO   stdIn:4.2-13.20: X.main[2].d
1085         / stdIn:4.2-13.20: X.main[2].a
1086         | stdIn:4.2-13.20: X.main[2].b
1087    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1088    CALL   stdIn:4.2-13.20: X.main[2].g
1089    GOTO   stdIn:4.2-13.20: X.main[2].f
1090    CALL   stdIn:2.15-17.4: X.main[2]
1091    -
1092    
1093    (Note that because of a FLINt bug the above code currently does not
1094    compile without BTrace turned on.)
1095    
1096    Here is another example, using my modified Tiger compiler:
1097    
1098    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
1099    - SMLofNJ.Internals.BTrace.mode (SOME true);
1100    [autoloading]
1101    [autoloading done]
1102    val it = false : bool
1103    - CM.make "sources.cm";
1104    [autoloading]
1105    ...
1106    [autoloading done]
1107    [scanning sources.cm]
1108    [parsing (sources.cm):parse.sml]
1109    [creating directory CM/SKEL ...]
1110    [parsing (sources.cm):tiger.lex.sml]
1111    ...
1112    [wrote CM/sparc-unix/semant.sml]
1113    [compiling (sources.cm):main.sml]
1114    [wrote CM/sparc-unix/main.sml]
1115    [New bindings added.]
1116    val it = true : bool
1117    - Main.compile ("../testcases/merge.tig", "foo.out");
1118    *** BACK-TRACE ***
1119    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
1120    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1121    CALL   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1122    GOTO   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1123    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1124    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1125    CALL   lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
1126         / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
1127    CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
1128    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1129    CALL   lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
1130    CALL   main.sml:1.18-118.4: Main.compile[2]
1131    -
1132    
1133    ----------------------------------------------------------------------
1134    Name: Matthias Blumen
1135    Date: 2000/06/21 18:00:00 JST
1136    Tag: blume-20000621-manual
1137    Description:
1138    
1139    CM manual update: Path environments documented.
1140    
1141    ----------------------------------------------------------------------
1142    Name: Matthias Blume
1143    Date: 2000/06/19 13:40:00
1144    Tag: blume-20000619-manual
1145    Description:
1146    
1147    CM manual and system/README update.  This only covers the fact that
1148    there are no more implicit anchors.  (Path environments and the "bind"
1149    option to "cm" have yet to be documented.)
1150    
1151    ----------------------------------------------------------------------
1152    Name: Matthias Blume
1153    Date: 2000/06/19 11:05:00 JST
1154    Tag: blume-20000619-chdir-bugfix
1155    Description:
1156    
1157    Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
1158    
1159    ----------------------------------------------------------------------
1160    Name: Matthias Blume
1161    Date: 2000/06/18 22:00:10 JST
1162    Tag: blume-20000618-implicit-anchors-really-gone
1163    Description:
1164    
1165    I updates the previous HISTORY entry where I forgot to mention that
1166    implicit anchors are no longer with us.
1167    
1168    The current update also gets rid of the (now useless) controller
1169    CM.Control.implicit_anchors.
1170    
1171    ----------------------------------------------------------------------
1172    Name: Matthias Blume
1173    Date: 2000/06/16 17:30:00 JST
1174    Tag: blume-20000616-anchorenv
1175    Description:
1176    
1177    This patch implements the long anticipated (just kidding :) "anchor
1178    environment" mechanism.  In the course of doing this, I also
1179    re-implemented CM's internal "SrcPath" module from scratch.  The new
1180    one should be more robust in certain boundary cases.  In any case, it
1181    is a lot cleaner than its predecessor (IMHO).
1182    
1183    This time, although there is yet another boot file format change, I
1184    kept the unpickler backward-compatible.  As a result, no new bootfiles
1185    are necessary and bootstrapping is straightforward.  (You cannot read
1186    new bootfiles into an old system, but the other way around is no
1187    problem.)
1188    
1189    Visible changes:
1190    
1191    ** 0. Implicit path anchors (without the leading $-symbol) are no
1192    longer recognized at all. This means that such path names are not
1193    illegal either.  For example, the name basis.cm simply refers to a
1194    local file called "basis.cm" (i.e, the name is an ordinary path
1195    relative to .cm-files directory).  Or, to put it differently, only
1196    names that start with $ are anchored paths.
1197    
1198    ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
1199    vanished.
1200    
1201    John (Reppy) had critizised this as soon as I originally proposed and
1202    implemented it, but at that time I did not really deeply believe
1203    him. :) Now I came full-circle because I need the $<singlearc> syntax
1204    in another place where it cannot be seen as an abbreviation for
1205    $/<singlearc>.  To avoid the confusion, $<singlearc> now means what it
1206    seems to mean (i.e., it "expands" into the corresponding anchor
1207    value).
1208    
1209    However, when paths are used as members in CM description files, it
1210    continues to be true that there must be at least another arc after the
1211    anchor.  This is now enforced separately during semantic analysis
1212    (i.e., from a lexical/syntactical point of view, the notation is ok.)
1213    
1214    ** 2. The "cm" class now accepts an option "bind".  The option's value
1215    is a sub-option list of precisely two items -- one labeled "anchor"
1216    and the other one labeled "value".  As you might expect, "anchor" is
1217    used to specify an anchor name to be bound, and "value" specifies what
1218    the anchor is being bound to.
1219    
1220    The value must be a directory name and can be given in either standard
1221    syntax (including the possibility that it is itself an anchored path)
1222    or native syntax.
1223    
1224    Examples:
1225    
1226       foo.cm (bind:(anchor:bar value:$mystuff/bar))
1227       lib.cm (bind:(anchor:a value:"H:\\x\\y\\z"))  (* only works under windows *)
1228    
1229    and so on.
1230    
1231    The meaning of this is that the .cm-file will be processed with an
1232    augmented anchor environment where the given anchor(s) is/are bound to
1233    the given values(s).
1234    
1235    The rationale for having this feature is this: Suppose you are trying
1236    to use two different (already stable) libraries a.cm and b.cm (that
1237    you perhaps didn't write yourself).  Further, suppose each of these
1238    two libraries internally uses its own auxiliary library $aux/lib.cm.
1239    Normally you would now have a problem because the anchor "lib" can not
1240    be bound to more than one value globally.  Therefore, the project that
1241    uses both a.cm and b.cm must locally redirect the anchor to some other
1242    place:
1243    
1244       a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
1245       b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
1246    
1247    This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
1248    /usr/lib/smlnj/b-stuff/aux.cm, respectively.
1249    
1250    Hard-wiring path names is a bit inflexible (and CM will verbosely warn
1251    you when you do so at the time of CM.stabilize).  Therefore, you can
1252    also use an anchored path as the value:
1253    
1254      a.cm (bind:(anchor:lib value:$a-lib))
1255      b.cm (bind:(anchor:lib value:$b-lib))
1256    
1257    Now you can globally configure (using the usual CM.Anchor.anchor or
1258    pathconfig machinery) bindings for "a-lib" and "b-lib".  Since "lib"
1259    itself is always locally bound, setting it globally is no longer
1260    meaningful or necessary (but it does not hurt either).  In fact, "lib"
1261    can still be used as a global anchor for separate purposes.  As a
1262    matter of fact, one can locally define "lib" in terms of a global
1263    "lib":
1264    
1265      a.cm (bind:(anchor:lib value:$lib/a))
1266      b.cm (bind:(anchor:lib value:$lib/b))
1267    
1268    ** 3: The encoding of path names has changed.  This affects the way
1269    path names are shown in CM's progress report and also the internal
1270    protocol encoding used for parallel make.
1271    
1272    The encoding now uses one or more ':'-separated segments.  Each
1273    segments corresponds to a file that has been specified relative to the
1274    file given by its preceding segment.  The first segment is either
1275    relative to the CWD, absolute, or anchored.  Each segment itself is
1276    basically a Unix pathname; all segments but the first are relative.
1277    
1278    Example:
1279    
1280       $foo/bar/baz.cm:a/b/c.sml
1281    
1282    This path denotes the file bar/a/b/c.sml relative to the directory
1283    denoted by anchor "foo".  Notice that the encoding also includes
1284    baz.cm which is the .cm-file that listed a/b/c.sml.  As usual, such
1285    paths are resolved relative to the .cm-files directory, so baz.cm must
1286    be ignored to get the "real" pathname.
1287    
1288    To make this fact more obvious, CM puts the names of such "virtual
1289    arcs" into parentheses when they appear in progress reports. (No
1290    parentheses will appear in the internal protocol encoding.)  Thus,
1291    what you really see is:
1292    
1293      $foo/bar/(baz.cm):a/b/c.sml
1294    
1295    I find this notation to be much more informative than before.
1296    
1297    Another new feature of the encoding is that special characters
1298    including parentheses, colons, (back)slashes, and white space are
1299    written as \ddd (where ddd is the decimal encoding of the character).
1300    
1301    *** The CM manual still needs to be updated.
1302    
1303    ----------------------------------------------------------------------
1304    Name: Allen Leung
1305    Date: 2000/06/15 00:38:00
1306    Tag: leunga-20000615-x86-peephole
1307    
1308    x86 Peephole fix by Fermin.  Affects c-- and moby only.
1309    
1310    ----------------------------------------------------------------------
1311    Name: Matthias Blume
1312    Date: 2000/06/12 11:40:00
1313    Tag: blume-20000612-parmakefix
1314    Description:
1315    
1316    More cleanup after changing the file naming scheme: This time I
1317    repaired the parallel make mechanism for CMB.make which I broke earlier.
1318    
1319    ----------------------------------------------------------------------
1320    Name: Allen Leung
1321    Date: 2000/06/09 01:25:00
1322    Tag: leunga-20000609-various
1323    
1324    None of these things should affect normal SML/NJ operations
1325    
1326    1. Peephole improvements provided by Fermin (c--)
1327    2. New annotation DEFUSE for adding extra dependence (moby)
1328    3. New X86 LOCK instructions (moby)
1329    4. New machine description language for reservation tables (scheduling)
1330    5. Fixes to various optimization/analysis modules (branch chaining, dominator
1331       trees etc.)
1332    6. I've changed the CM files so that they can work with versions
1333       110.0.6, 110.25 and 110.28
1334    
1335    ----------------------------------------------------------------------
1336    Name: Matthias Blume
1337    Date: 2000/06/09 12:40:00
1338    Tag: blume-20000609-log
1339    Description:
1340    
1341    - Removed all(?) remaining RCS Log entries from sources.
1342    
1343    - Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
1344      anchored paths).
1345    
1346    ----------------------------------------------------------------------
1347    Name: Matthias Blume
1348    Date: 2000/06/07 17:00:00 JST
1349    Tag: blume-20000607-no-implicit-anchors
1350    Description:
1351    
1352    1. This update changes the default setting for
1353    CM.Control.implicit_anchors from true to false.  This means that
1354    implicit anchors are no longer permitted by default.  I also tried to
1355    make sure that nothing else still relies on implicit anchors.
1356    (This is the next step on the schedule towards a CM that does not even
1357    have the notion of implicit anchors anymore.)
1358    
1359    2. More CM manual updates.
1360    
1361    3. I managed to track down and fix the pickling bug I mentioned last
1362    time.  Because of the previously existing workaround, this entails no
1363    immediate practical changes.
1364    
1365    ----------------------------------------------------------------------
1366    Name: Matthias Blume
1367    Date: 2000/06/06 11:15:00 JST
1368    Tag: blume-20000606-lazierpickle
1369    Description:
1370    
1371    !!!! NEW BOOT FILES !!!!
1372    
1373    * The main purpose of this update is to make library pickles lazier in
1374    order to reduce the initial space penalty for autoloading a library.
1375    As a result, it is now possible to have $smlnj/compiler.cm
1376    pre-registered.  This should take care of the many complaints or
1377    inquiries about missing structure Compiler.  This required changes to
1378    CM's internal data structures and small tweaks to some algorithms.
1379    
1380    As a neat additional effect, it is no longer necessary (for the sake
1381    of lean heap image files) to distinguish between a "minimal" CM and a
1382    "full" CM.  Now, there is only one CM (i.e., the "full" version:
1383    $smlnj/cm.cm aka $smlnj/cm/full.cm), and it is always available at the
1384    interactive top level. ($smlnj/cm/minimal.cm is gone.)
1385    
1386    To make the life of compiler-hackers easier, "makeml" now also
1387    pre-registers $smlnj/cmb.cm (aka $smlnj/cmb/current.cm).  In other
1388    words, after you bootstrap a new sml for the first time, you will not
1389    have to autoload $smlnj/cmb.cm again afterwards.  (The first time
1390    around you will still have to do it, though.)
1391    
1392    * A second change consists of major updates to the CM manual.  There
1393    are now several appendices with summary information and also a full
1394    specification of the CM description file syntax.
1395    
1396    * In directory src/system I added the script "allcross".  This script
1397    invokes sml and cross-compiles the compiler for all supported
1398    architectures.  (Useful when providing a new set of boot files.)
1399    
1400    * There seems to be a latent bug in my "lazy pickles" mechanism.  I
1401    added a small tweak to pickle-util.sml to work around this problem,
1402    but it is not a proper fix yet.  I will investigate further.  (The
1403    effect of the bug was an inflation of library pickle size.)
1404    
1405    * Version number increased to 110.28.1 (to avoid compatibility problems).
1406    
1407    ----------------------------------------------------------------------
1408    Name: Allen Leung
1409    Date: 2000/05/25 17:28 EDT
1410    Tag: leunga-20000525-ra
1411    Description:
1412    
1413      Fixed a bug in freezing phase of the register allocator.
1414    
1415    ----------------------------------------------------------------------
1416    Name: Allen Leung
1417    Date: 2000/05/15 22:53 EDT
1418    Tag: leunga-20000515-alpha-x86-ra
1419    Description:
1420    
1421      1. Alpha
1422    
1423          Slight cleanup.  Removed the instruction SGNXL
1424    
1425      2. X86
1426    
1427          Added the following instructions to the instruction set:
1428    
1429            ROLx, RORx,
1430            BTx, BTSx, BTLx, BTRx,
1431            XCHGx, and variants with the LOCK prefix
1432    
1433      3. Register Allocation
1434    
1435          The module ra-rewrite-with-renaming has been improved.
1436    
1437      These have no effect on SML/NJ.
1438    
1439    ----------------------------------------------------------------------
1440    Name: Matthias Blume
1441    Date: 2000/05/15 16:20:00 JST
1442    Tag: blume-20000515-lightrebuild
1443    Description:
1444    
1445    1. I added an alternative to "-rebuild" to "makeml".  The difference is
1446       that prior to calling CMB.make' the CM-variable "LIGHT" will be
1447       defined.  In effect, the command will not build any cross-compiler
1448       backends and therefore finish more quickly.
1449    
1450       The "fixpt" script also takes a "-light" switch to be able to use
1451       this new facility while compiling for a fixpoint.
1452    
1453    2. I replaced all mentions of anchored paths in group owner specifications
1454       with simple relative paths (usually starting with "..").
1455       The rationale is that a library's internal workings should not be
1456       compromised by the lack of some anchor.  (An anchor is necessary
1457       for someone who wants to refer to the library by an anchored path,
1458       but it should not be necessary to build the same library in the first
1459       place.)
1460    
1461    3. I changed the way CM's tool mechanism determines the shell command
1462       string used for things like ml-yacc etc. so that it does not break
1463       when CM.Control.implicit_anchors is turned off.
1464    
1465    ----------------------------------------------------------------------
1466    Name: Matthias Blume
1467    Date: 2000/05/12 18:20:00 JST
1468    Tag: blume-20000512-ml-build
1469    Description:
1470    
1471    Fixed a bug in config/_ml-build that prevented ml-yacc and ml-lex from
1472    getting installed properly (by config/install.sh).
1473    
1474    ----------------------------------------------------------------------
1475    Name: Matthias Blume
1476    Date: 2000/05/12 17:30:00 JST
1477    Tag: blume-20000512-anchors
1478    Description:
1479    
1480    !!! NEW BOOT FILES !!!
1481    
1482    This change is in preparation of fading out support for "implicitly
1483    anchored path names".  I went through all sources and used the
1484    explicit (and relatively new) $-notation.  See system/README and the
1485    CM manual for more info on this.
1486    
1487    I also modified the anchoring scheme for some things such as "smlnj",
1488    "MLRISC", "cm", etc. to take advantage of the fact that explicit
1489    anchors are more expressive: anchor name and first arc do not have to
1490    coincide.  This entails the following user-visible change:
1491    
1492    You have to write $smlnj/foo/bar instead of smlnj/foo/bar.  In
1493    particular, when you fire up sml with a command-line argument, say,
1494    e.g.:
1495    
1496       sml '$smlnj/cmb.cm'
1497    
1498    At the ML toplevel prompt:
1499    
1500       CM.autoload "$smlnj/cmb.cm";
1501    
1502    There is also a new controller in CM.Control that can be used to turn
1503    off all remaining support for implicit anchors by saying:
1504    
1505        CM.autoload "$smlnj/
1506        #set CM.Control.implicit_anchors false;
1507    
1508    This causes CM to reject implicitly anchored paths.  This is (for the
1509    time being) less permissive than the "final" version where there will
1510    be no more such implicit anchors and relative paths will be just that:
1511    relative.
1512    
1513    The next step (version after next version?) will be to make the
1514    default for CM.Control.implicit_anchors false.  After the dust has
1515    settled, I can then produce the "final" version of this...
1516    
1517    Note: Since bootstrapping is a bit tricky, I provided new boot files.
1518    
1519    ----------------------------------------------------------------------
1520    Name: Matthias Blume
1521    Date: 2000/05/11 16:30:00 JST
1522    Tag: blume-20000511-sources
1523    Description:
1524    
1525    The main change is that I added function CM.sources as a generalized
1526    version of the earlier CM.makedepend.  This entails the following
1527    additional changes:
1528    
1529      - CM.makedepend has been dropped.
1530    
1531      - CM manual has been updated.
1532    
1533      - TOOLS signature and API have been changed.
1534    
1535    ----------------------------------------------------------------------
1536    Name: Allen Leung
1537    Date: 2000/05/10 21:17 EDT
1538    Tag: leunga-20000510-moby-c--ssa
1539    Description:
1540    
1541      Various bug fixes and new features for C--, Moby and MLRISC optimizations.
1542    None of these affect SML/NJ.
1543    
1544    1. Register Allocation
1545    
1546        a. A new ra spilling module (ra/ra-spill-with-renaming) is implemented.
1547           This module tries to remove local (i.e. basic block level) redundancies
1548           during spilling.
1549    
1550        b. A new framework for performing region based register allocation.
1551           Not yet entirely functional.
1552    
1553    2. X86
1554    
1555       a. DefUse for POP was missing the stack pointer [found by Lal]
1556       b. Reload for CALL was incorrect in X86Spill [found by John]
1557       c. Various fixes in X86Spill so that it can be used correctly for
1558          the new spilling module.
1559    
1560    3. SSA/IR
1561    
1562       a. New module ir/dj-dataflow.sml implements elimination based
1563          data flow analysis.
1564    
1565    4. MLRiscGen
1566    
1567       a. Fix for gc type annotation
1568    
1569    5. MDGen
1570    
1571       Various fixes for machine description -> ml code translation.  For ssa
1572       only.
1573    
1574    ----------------------------------------------------------------------
1575    Name: Allen Leung
1576    Date: 2000/05/08 22:17 EDT
1577    Tag: leunga-20000508-labexp
1578    Description:
1579    
1580      Fermin has found a few assembly problems with constant expressions
1581      generated in LabelExp.  Mostly, the problems involve extra parentheses,
1582      which choke on dumb assemblers.  This is his fix.
1583    
1584    ----------------------------------------------------------------------
1585    Name: Dave MacQueen
1586    Date: 2000/04/09 14:00 EDT
1587    Tag: dbm-20000502-Version_110_28
1588    Description:
1589    
1590    1. Updated src/compiler/TopLevel/main/version.sml to version 110.28
1591    
1592    2. Updated config/version to 110.28
1593    
1594    3. Updated config/srcarchiveurl
1595    
1596    3. New boot files!
1597       ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.28/
1598    
1599    ----------------------------------------------------------------------
1600    Name: Matthias Blume
1601    Date: 2000/05/01 19:05:00 JST
1602    Tag: blume-20000501-noweb
1603    Description:
1604    
1605    A new noweb tool has been added.  The existing system is entirely
1606    unaffected by this, but some CM users have asked for renewed noweb
1607    support.  Everything is documented in the CM manual.
1608    
1609    New (plugin) libraries:
1610    
1611       noweb-tool.cm
1612       nw-ext.cm
1613    
1614    ----------------------------------------------------------------------
1615    Name: Dave MacQueen
1616    Date: 2000/04/30 12:40PM EDT
1617    Tag: dbm-20000430-bug_fixes
1618    Description:
1619    
1620    1. Fix for bug 1498
1621       smlnj/src/system/Basis/Implementation/Unsafe/object.sig
1622       smlnj/src/system/Basis/Implementation/Unsafe/object.sml
1623         added toRealArray function
1624       smlnj/src/compiler/MiscUtil/print/ppobj.sml
1625         added check for tag Obj.RealArray to array printing case in ppObj
1626    
1627    2. Fix for bug 1510
1628       smlnj/src/compiler/Semant/types/typesutil.sml
1629         fixed definition of dummyargs (used by equalTycon) so that
1630         dummy args are distinct types
1631    
1632    ----------------------------------------------------------------------
1633    Name: Matthias Blume
1634    Date: 2000/04/30 01:00:00 JST
1635    Tag: blume-20000430-versions
1636    Description:
1637    
1638    1. CM version numbering added.  This is an implementation of Lal's
1639       proposal for adding version numbers and version checking to .cm
1640       files.  Lal said that his proposal was just that -- a proposal.
1641       For the time being I went ahead and implemented it so that people
1642       can comment on it.  Everything is completely backward-compatible
1643       (except for the stable library format, i.e., new bootfiles!).
1644    
1645       As usual, see the CM manual for details.
1646    
1647    2. An alternative syntax for anchored paths has been implemented.
1648       Dave has recently voiced the same concerns that I had when I did
1649       this, so there should be some support.  My take is that eventually
1650       I will let support for the current syntax (where anchors are
1651       "implicit") fade out in favor of the new, explicit syntax.
1652       In order to be backward-compatible, both old and new syntax are
1653       currently supported.
1654    
1655       Again, see the CM manual for details.
1656    
1657    3. Parallel make is trying to be slightly smarter:  When the master
1658       process finds a "bottleneck", i.e., when there is only one
1659       compilation unit that can be compiled and everybody else is
1660       waiting on it, then it will simply compile it directly instead
1661       of clumsily telling one of the slaves to do it.
1662    
1663    4. Support for "unsharing" added.  This is necessary in order to be
1664       able to have two different versions of the same library running
1665       at the same time (e.g., for trying out a new MLRISC while still
1666       having the old MLRISC linked into the current compiler, etc.)
1667       See the CM manual.
1668    
1669    5. Simple "makedepend" functionality added for generating Makefile
1670       dependency information.  (This is rather crude at the moment.
1671       Expect some changes here in the future.)
1672    
1673    6. ".fun" added as a recognized suffix for ML files. Also documented
1674       explicitly in the manual that the fallback behavior (unknown suffix
1675       -> ML file) is not an official feature!
1676    
1677    7. Small changes to the pickler for stable libraries.
1678    
1679    8. Several internal changes to CM (for cleanup/improvement).
1680    
1681    
1682    !!!! NEW BINFILES !!!!
1683    
1684    ----------------------------------------------------------------------
1685    Name: Matthias Blume
1686    Date: 2000/04/28 17:30:00 JST
1687    Tag: blume-20000428-pathconfig
1688    Description:
1689    
1690    1. I changed config/install.sh to remove duplicate entries from the
1691       lib/pathconfig file at the end.  Moreover, the final version of
1692       lib/pathconfig is sorted alphabetically.  The same (sorting) is done
1693       in src/system/installml.
1694    
1695    2. The config/install.sh script now consistently uses relative
1696       pathnames in lib/pathconfig whenever the anchor is in the lib
1697       directory.  (So far this was true for the libraries that come
1698       pre-compiled and bundled as part of the bootfiles but not for
1699       libraries that are compiled by the script itself.)
1700    
1701    ----------------------------------------------------------------------
1702    Name: Matthias Blume
1703    Date: 2000/04/26 13:10:00 JST
1704    Tag: blume-20000426-fun_suffix
1705    Description:
1706    
1707    Added ".fun" as a recognized file name suffix (for ML code).
1708    
1709    ----------------------------------------------------------------------
1710    Name: Allen Leung
1711    Date: 2000/04/25 17:00:00 EST
1712    Tag: leunga-20000425-alpha-ra
1713    Description:
1714    
1715    1. Alpha
1716    
1717        PSEUDOARITH was missing in AlphaRewrite.  This causes an endless loop
1718    in C--.
1719    
1720    2. RA
1721    
1722       Added a flag "ra-dump-size" to print out the size of the flowgraph
1723       and the interference graph.
1724    
1725    ----------------------------------------------------------------------
1726    Name: Dave MacQueen
1727    Date: 2000/04/25/
1728    Tag: dbm-20000425-mlyacc_doc_examples
1729    Description:
1730      Updated mlyacc.tex sections 5 and 7 for SML '97 and CM.
1731      Updated all three examples in src/ml-yacc/examples to run
1732      under 110.* using CM.make.
1733    
1734    ----------------------------------------------------------------------
1735    Name: Allen Leung
1736    Date: 2000/04/20 23:04:00 EST
1737    Tag: leunga-20000420-ssa-c---stuff
1738    Description:
1739    
1740      This update synchronizes my repository with Yale's.  Most of these
1741    changes, however, do not affect SML/NJ at all (the RA is an exception).
1742    
1743    1. Register Allocator
1744    
1745       a. An improvement in the interference graph construction:
1746          Given a copy
1747    
1748                s <- t
1749    
1750          no interference edge between s and t is added for this definition of s.
1751    
1752       b. I've added two new spill heuristic modules that Fermin and I developed
1753          (in the new library RA.cm). These are unused in SML/NJ but maybe
1754          useful for others (Moby?)
1755    
1756    2. X86
1757    
1758       a. Various fixes in the backend provided by Fermin [C--] and Lal.
1759    
1760    3. Alpha
1761    
1762       a. Added the BSR instruction and code generation that goes with it [C--]
1763       b. Other fixes too numerous to recount provided by Fermin [C--]
1764    
1765    4. Regmaps
1766    
1767       a. The regmaps are not initialized with the identity physical bindings
1768          at creation time.  This is unneeded.
1769    
1770    5. MLRISC Optimizations
1771    
1772       a. The DJ-Graph module can now compute the iterated dominance frontiers
1773          intersects with liveness incrementally in linear time! Woohoo!
1774          This is now used in my new SSA construction algorithm.
1775    
1776       b. THe branch reorganization module is now smarter about linear chains of
1777          basic blocks.
1778    
1779    
1780    ----------------------------------------------------------------------
1781    Name: Matthias Blume
1782    Date: 2000/04/12 13:52:00 JST
1783    Tag: blume_main_v110p27_1
1784    Description:
1785    
1786    Changed install.sh script to handle archive files without version number
1787    and to use "boot.<arch>-<os>" instead of "sml.boot.<arch>-<os>" for the
1788    name of the boot file archive.
1789    
1790    ----------------------------------------------------------------------
1791    Name: Dave MacQueen
1792    Date: 2000/04/09 14:00 EDT
1793    Tag: dbm-20000410-Version_110_27
1794    Description:
1795    
1796    1. Updated src/compiler/TopLevel/main/version.sml to version 110.27
1797    
1798    2. Updated src/config/version to 110.27
1799    
1800    3. New boot files!
1801    
1802    ----------------------------------------------------------------------
1803    Name: Allen Leung
1804    Date: 2000/04/09 19:09:00 EST
1805    Tag: leunga-20000409-misc
1806    Description:
1807    
1808    1.  Yet another fix for x86 assembly for idivl, imull, mull and friends.
1809    
1810    2.  Miscellaneous improvements to MLRISC (unused in sml/nj)
1811    
1812    ----------------------------------------------------------------------
1813    Name: Stefan
1814    Date: 2000/04/07 10:00:00 EDT
1815    Tag: monnier-20000406-branch-handling
1816    Description:
1817    
1818    Improved handling of branches (mostly those generated from
1819    polymorphic equality), removed switchoff and changed the
1820    default optimization settings (more cpsopt and less flintopt).
1821    
1822    ----------------------------------------------------------------------
1823    Name: Allen Leung
1824    Date: 2000/04/06 01:30:00 EST
1825    Tag: leunga-20000406-peephole-x86-SSA-2
1826    Description:
1827    
1828       Forgot a few files.
1829    
1830    ----------------------------------------------------------------------
1831    Name: Allen Leung
1832    Date: 2000/04/06 00:36:00 EST
1833    Tag: leunga-20000406-peephole-x86-SSA
1834    Description:
1835    
1836    1.  New Peephole code
1837    
1838    2.  Minor improvement to X86 instruction selection
1839    
1840    3.  Various fixes to SSA and machine description -> code translator
1841    
1842    ----------------------------------------------------------------------
1843    Name: Matthias Blume
1844    Date: 2000/04/05 12:30:00 JST
1845    Tag: blume_main_v110p26p2_3
1846    Description:
1847    
1848    This update just merges three minor cosmetic updates to CM's sources
1849    to get ready for the 110.27 code freeze on Friday.  No functionality
1850    has changed.
1851    
1852    ----------------------------------------------------------------------
1853    Name: Allen Leung
1854    Date: 2000/04/04 19:39:00 EST
1855    Tag: leunga-20000404-x86-asm
1856    Description:
1857    
1858    1.  Fixed a problem in X86 assembly.
1859    
1860        Things like
1861    
1862           jmp %eax
1863           jmp (%eax)
1864    
1865        should be output as
1866    
1867           jmp *%eax
1868           jmp *(%eax)
1869    
1870    2.  Assembly output
1871    
1872          Added a new flag
1873    
1874              "asm-indent-copies" (default to false)
1875    
1876          When this flag is on, parallel copies will be indented an extra level.
1877    
1878    ----------------------------------------------------------------------
1879    Name: Allen Leung
1880    Date: 2000/04/04 03:18:00 EST
1881    Tag: leunga-20000404-C--Moby
1882    Description:
1883    
1884        All of these fixes are related to C--, Moby, and my own optimization
1885        stuff; so they shouldn't affect SML/NJ.
1886    
1887    1.  X86
1888    
1889        Various fixes related floating point, and extensions.
1890    
1891    2.  Alpha
1892    
1893        Some extra patterns related to loads with signed/zero extension
1894        provided by Fermin.
1895    
1896    3.  Assembly
1897    
1898        When generating assembly, resolve the value of client defined constants,
1899        instead of generating symbolic values.  This is controlled by the
1900        new flag "asm-resolve-constants", which is default to true.
1901    
1902    4.  Machine Descriptions
1903    
1904        a. The precedence parser was slightly broken when parsing infixr symbols.
1905        b. The type generalizing code had the bound variables reversed, resulting
1906           in a problem during arity raising.
1907        c. Various fixes in machine descriptions.
1908    
1909    ----------------------------------------------------------------------
1910    Name: Matthias Blume
1911    Date: 2000/04/03 16:05:00 JST
1912    Tag: blume_main_v110p26p2_2
1913    Description:
1914    
1915    I eliminated coreEnv from compInfo.  Access to the "Core" structure is
1916    now done via the ordinary static environment that is context to each
1917    compilation unit.
1918    
1919    To this end, I arranged that instead of "structure Core" as "structure
1920    _Core" is bound in the pervasive environment.  Core access is done via
1921    _Core (which can never be accidentally rebound because _Core is not a
1922    legal surface-syntax symbol).
1923    
1924    The current solution is much cleaner because the core environment is
1925    now simply part of the pervasive environment which is part of every
1926    compilation unit's context anyway.  In particular, this eliminates all
1927    special-case handling that was necessary until now in order to deal
1928    with dynamic and symbolic parts of the core environment.
1929    
1930    Remaining hackery (to bind the "magic" symbol _Core) is localized in the
1931    compilation manager's bootstrap compiler (actually: in the "init group"
1932    handling).  See the comments in src/system/smlnj/init/init.cmi for
1933    more details.
1934    
1935    I also tried to track down all mentions of "Core" (as string argument
1936    to Symbol.strSymbol) in the compiler and replaced them with a
1937    reference to the new CoreSym.coreSym.  Seems cleaner since the actual
1938    name appears in one place only.
1939    
1940    Binfile and bootfile format have not changed, but the switchover from
1941    the old "init.cmi" to the new one is a bit tricky, so I supplied new
1942    bootfiles anyway.
1943    
1944    ----------------------------------------------------------------------
1945    Name: Allen Leung
1946    Date: 2000/04/02 21:17:00 EST
1947    Tag: leunga-20000402-mltree
1948    Description:
1949    
1950       1. Renamed the constructor CALL in MLTREE by popular demand.
1951       2. Added a bunch of files from my repository.  These are currently
1952          used by other non-SMLNJ backends.
1953    
1954    ----------------------------------------------------------------------
1955    Name: Allen Leung
1956    Date: 2000/03/31 21:15:00 EST
1957    Tag: leunga-20000331-aliasing
1958    Description:
1959    
1960    This update contains a rewritten (and hopefully more correct) module
1961    for extracting aliasing information from CPS.
1962    
1963       To turn on this feature:
1964    
1965            Compiler.Control.CG.memDisambiguate := true
1966    
1967       To pretty print the region information with assembly
1968    
1969           Compiler.Control.MLRISC.getFlag "asm-show-region" := true;
1970    
1971       To control how many levels of aliasing information are printed, use:
1972    
1973           Compiler.Control.MLRISC.getInt "points-to-show-level" := n
1974    
1975       The default of n is 3.
1976    
1977    ----------------------------------------------------------------------
1978    Name: David MacQueen
1979    Date: 2000/03/31 11:15:00 EST
1980    Tag: dbm-20000331-runtime_fix
1981    Description:
1982    
1983    This update contains:
1984    
1985    1. runtime/c-lib/c-libraries.c
1986       includes added in revision 1.2 caused compilation errors on hppa-hpux
1987    
1988    2. fix for bug 1556
1989       system/Basis/Implementation/NJ/internal-signals.sml
1990    
1991    ----------------------------------------------------------------------
1992    Name: Matthias Blume
1993    Date: 2000/03/31 18:00:00 JST
1994    Tag: blume_main_v110p26p2_1
1995    Description:
1996    
1997    This update contains:
1998    
1999    1. A small change to CM's handling of stable libraries:
2000       CM now maintains one "global" modmap that is used for all stable
2001       libraries.  The use of such a global modmap maximizes sharing and
2002       minimizes the need for re-traversing parts of environments during
2003       modmap construction.  (However, this has minor impact since modmap
2004       construction seems to account for just one percent or less of total
2005       compile time.)
2006    
2007    2. I added a "genmap" phase to the statistics.  This is where I got the
2008       "one percent" number (see above).
2009    
2010    3. CM's new tool parameter mechanism just became _even_ better. :)
2011       - The parser understands named parameters and recursive options.
2012       - The "make" and "shell" tools use these new features.
2013         (This makes it a lot easier to cascade these tools.)
2014       - There is a small syntax change: named parameters use a
2015    
2016           <name> : ( <option> ... )            or
2017           <name> : <string>
2018    
2019         syntax.  Previously, named parameters were implemented in an
2020         ad-hoc fashion by each tool individually (by parsing strings)
2021         and had the form
2022    
2023           <name>=<string>
2024    
2025       See the CM manual for a full description of these issues.
2026    
2027    ----------------------------------------------------------------------
2028    Name: Matthias Blume
2029    Date: 2000/03/30 18:00:00 JST
2030    Tag: blume_main_v110p26p2_0
2031    Description:
2032    
2033    !!!!! WARNING !!!!!!
2034    !!  New binfiles  !!
2035    !!!!!!!!!!!!!!!!!!!!
2036    
2037    This update contains:
2038    
2039    1. Moderate changes to CM:
2040    
2041       - Changes to CM's tools mechanism.  In particular, it is now possible
2042       to have tools that accept additional "command line" parameters
2043       (specified in the .cm file at each instance where the tool's class is
2044       used).
2045    
2046       This was done to accommodate the new "make" and "shell" tools which
2047       facilitate fairly seamless hookup to portions of code managed using
2048       Makefiles or Shell scripts.
2049    
2050       There are no classes "shared" or "private" anymore.  Instead, the
2051       sharing annotation is now a parameter to the "sml" class.
2052    
2053       There is a bit of generic machinery for implementing one's own
2054       tools that accept command-line parameters.  However, I am not yet fully
2055       satisfied with that part, so expect changes here in the future.
2056    
2057       All existing tools are described in the CM manual.
2058    
2059       - Slightly better error handling.  (CM now suppresses many followup
2060       error messages that tended to be more annoying than helpful.)
2061    
2062    2. Major changes to the compiler's static environment data structures.
2063    
2064       - no CMStaticEnv anymore.
2065            - no CMEnv, no "BareEnvironment" (actually, _only_ BareEnvironment,
2066              but it is called Environment), no conversions between different
2067              kinds of static environments
2068    
2069       - There is still a notion of a "modmap", but such modmaps are generated
2070         on demand at the time when they are needed.  This sounds slow, but I
2071         sped up the code that generates modmaps enough for this not to lead to
2072         a slowdown of the compiler (at least I didn't detect any).
2073    
2074       - To facilitate rapid modmap generation, static environments now
2075         contain an (optional) "modtree" structure.  Modtree annotations are
2076         constructed by the unpickler during unpickling.  (This means that
2077         the elaborator does not have to worry about modtrees at all.)
2078         Modtrees have the advantage that they are compositional in the same
2079         way as the environment data structure itself is compositional.
2080         As a result, modtrees never hang on to parts of an environment that
2081         has already been rendered "stale" by filtering or rebinding.
2082    
2083       - I went through many, many trials and errors before arriving at the
2084         current solution.  (The initial idea of "linkpaths" did not work.)
2085         But the result of all this is that I have touched a lot of files that
2086         depend on the "modules" and "types" data structures (most of the
2087         elaborator). There were a lot of changes during my "linkpath" trials
2088         that could have been reverted to their original state but weren't.
2089         Please, don't be too harsh on me for messing with this code a bit more
2090         than what was strictly necessary...  (I _did_ resist the tempation
2091         of doing any "global reformatting" to avoid an untimely death at
2092         Dave's hands. :)
2093    
2094       - One positive aspect of the previous point:  At least I made sure that
2095         all files that I touched now compile without warnings (other than
2096         "polyEqual").
2097    
2098       - compiler now tends to run "leaner" (i.e., ties up less memory in
2099         redundant modmaps)
2100    
2101    ----------------------------------------------------------------------
2102  Name: Allen Leung  Name: Allen Leung
2103  Date: 2000/03/29 18:00:00  Date: 2000/03/29 18:00:00
2104  Tag: leunga-20000327-mlriscGen_hppa_alpha_x86  Tag: leunga-20000327-mlriscGen_hppa_alpha_x86
2105    Boot files (optional): ftp://react-ilp.cs.nyu.edu/leunga/110.26.1-sml.boot.x86-unix-20000330.tar.gz
2106  Description:  Description:
2107    
2108     This update contains *MAJOR* changes to the way code is generated from CPS     This update contains *MAJOR* changes to the way code is generated from CPS
# Line 103  Line 2190 
2190    
2191       I've changed andl to testl in the floating point test sequence       I've changed andl to testl in the floating point test sequence
2192       whenever appropriate.  The Intel optimization guide states that       whenever appropriate.  The Intel optimization guide states that
2193       testl is perferable to andl.       testl is preferable to andl.
2194    
2195  7. RA (x86 only)  7. RA (x86 only)
2196    
# Line 285  Line 2372 
2372    
2373  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
2374  additional argument called "template" which is an optional string that  additional argument called "template" which is an optional string that
2375  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
2376  explanation.  explanation.
2377    
2378  2. A special-purpose tool can be "regisitered" by simply dropping the  2. A special-purpose tool can be "registered" by simply dropping the
2379  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
2380  directory where the .cm file lives that uses this tool.  (The  directory where the .cm file lives that uses this tool.  (The
2381  behavior/misfeature until now was to look for the tool description  behavior/misfeature until now was to look for the tool description
# Line 332  Line 2419 
2419  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
2420  <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,
2421  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
2422  perhaps got accidentially killed).  In this case, fixpt will quickly  perhaps got accidentally killed).  In this case, fixpt will quickly
2423  move through what exists before continuing where it left off earlier,  move through what exists before continuing where it left off earlier,
2424  and, thus, saves a lot of time.  and, thus, saves a lot of time.
2425    
# Line 382  Line 2469 
2469    it from that remote directory.    it from that remote directory.
2470    This should simplify installation further:  For machines that have    This should simplify installation further:  For machines that have
2471    access to the internet, just fetch <version>-config.tgz, unpack it,    access to the internet, just fetch <version>-config.tgz, unpack it,
2472    edit config/targets, and go (run config/install.sh).  The scipt will    edit config/targets, and go (run config/install.sh).  The script will
2473    fetch everything else that it might need all by itself.    fetch everything else that it might need all by itself.
2474    
2475    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.585  
changed lines
  Added in v.813

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