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 647, Tue May 16 02:53:02 2000 UTC revision 771, Sat Dec 30 13:06:09 2000 UTC
# Line 12  Line 12 
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14  ----------------------------------------------------------------------  ----------------------------------------------------------------------
15    Name: Matthias Blume
16    Date: 2000/12/30 22:10:00 JST
17    Tag: blume-20001230-various
18    Description:
19    
20    Added proxy libraries for MLRISC and let MLRISC libraries refer
21    to each other using path anchors.  (See CM manual for explanation.)
22    
23    Updated CM documentation.
24    
25    Fixed some bugs in CM.
26    
27    Implemented "proxy" libraries (= syntactic sugar for CM).
28    
29    Added "-quiet" option to makeml and changed runtime system accordingly.
30    
31    Added cleanup handler for exportML to reset timers and compiler stats.
32    
33    ----------------------------------------------------------------------
34    Name: Lal George
35    Date: 2000/12/22 22:22:58 EST 2000
36    Tag: Release_110_32
37    Description:
38    
39            Infinite precision used throughout MLRISC.
40            see MLRISC/mltree/machine-int.sig
41    
42    ----------------------------------------------------------------------
43    Name: Matthias Blume
44    Date: 2000/12/22 23:16:00 JST
45    Tag: blume-20001222-warn
46    Description:
47    
48    Corrected wording and formatting of some CM warning message which I
49    broke in my previous patch.
50    
51    ----------------------------------------------------------------------
52    Name: Matthias Blume
53    Date: 2000/12/22 21:20:00 JST
54    Tag: blume-20001222-anchorenv
55    Description:
56    
57    Fixed CM's handling of anchor environments in connection with CMB.make.
58    
59    ----------------------------------------------------------------------
60    Name: Matthias Blume
61    Date: 2000/12/22 13:15:00 JST
62    Tag: blume-20001222-cleanup
63    Description:
64    
65    Removed src/cm/ffi which does not (and did not) belong here.
66    
67    ----------------------------------------------------------------------
68    Name: Matthias Blume
69    Date: 2000/12/21 23:55:00 JST
70    Tag: blume-20001221-exn
71    Description:
72    
73    Probably most important: CM no longer silently swallows all exceptions
74    in the compiler.
75    Plus: some other minor CM changes.  For example, CM now reports some
76    sizes for generated binfiles (code, data, envpickle, lambdapickle).
77    
78    ----------------------------------------------------------------------
79    Name: Matthias Blume
80    Date: 2000/12/15 00:01:05 JST
81    Tag: blume-20001215-dirtool
82    Description:
83    
84    - "dir" tool added.
85    - improvements and cleanup to Tools structure
86    - documentation updates
87    
88    ----------------------------------------------------------------------
89    Name: Allen Leung
90    Date: Thu Dec 14 03:45:24 EST 2000
91    Description:
92    Tag:  leunga-20001214-int-inf
93    Description:
94    
95       In IntInf, added these standard functions, which are missing from our
96    implementation:
97    
98        andb : int * int -> int
99        xorb : int * int -> int
100        orb  : int * int -> int
101        notb : int -> int
102         <<   : int * word -> int
103        ~>>  : int * word -> int
104    
105       Not tested, I hope they are correct.
106    
107    ----------------------------------------------------------------------
108    Name: Allen Leung
109    Date: Fri Dec  8 19:23:26 EST 2000
110    Description:
111    Tag:  leunga-20001208-nowhere
112    Description:
113    
114      Slight improvements to the 'nowhere' tool to handle OR-patterns,
115    to generate better error messages etc.  Plus a brief manual.
116    
117    ----------------------------------------------------------------------
118    Name: Lal George
119    Date: 2000/12/08 09:54:02 EST 2000
120    Tag: Release_110_31
121    Description:
122    
123    - Version 110.31
124    ----------------------------------------------------------------------
125    Name: Allen Leung
126    Date: Thu Dec  7 22:01:04 EST 2000
127    Tag:  leunga-20001207-cell-monster-hack
128    Description:
129    
130    Major MLRISC internal changes.  Affect all clients.
131    Summary:
132    
133    1.  Type CELLS.cell = int is now replaced by a datatype.
134        As a result, the old regmap is now gone.  Almost all interfaces
135        in MLRISC change as a consequence.
136    
137    2.  A new brand version of machine description tool (v3.0) that generates
138        modules expecting the new interface.  The old version is removed.
139    
140    3.  The RA interface has been further abstracted into two new functors.
141        RISC_RA and X86RA.  These functors have much simpler interfaces.
142        [See also directory MLRISC/demo.]
143    
144    4.  Some other new source->source code generation tools are available:
145    
146        a. MLRISC/Tools/RewriteGen -- generate rewriters from rules.
147        b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules.
148           I use this tool to generate the peephole optimizers---with the new
149           cell type changes, peephole rules are becoming difficult to write
150           without conditional pattern matching.
151    
152    5.  More Intmap -> IntHashTable change.  Previous changes by Matthias didn't
153        cover the entire MLRISC source tree so many things broke.
154    
155    6.  CM files have been moved to the subdirectory MLRISC/cm.
156        They are moved because there are a lot of them and they clutter up the
157        root dir.
158    
159    7.  More detailed documentation to come...
160    
161        NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do
162        a makeml -rebuild first.  This is because of other other
163        changes that Matthias has made (see below).
164    
165    
166    ----------------------------------------------------------------------
167    Name: Matthias Blume
168    Date: 2000/11/30 23:12:00 JST
169    Tag: blume-20001130-filereorg
170    Description:
171    
172    Some manual updates and some file reorganizations in CM.
173    
174    ----------------------------------------------------------------------
175    Name: Matthias Blume
176    Date: 2000/11/24 17:45:00 JST
177    Tag: blume-20001124-link
178    Description:
179    
180    Drastically improved link traversal code for the case that the dynamic
181    value was already loaded at bootstrap time.  As a result, CM and CMB
182    now both load blazingly fast -- even on a very slow machine.  Also,
183    memory consumption has been further reduced by this.
184    
185    Warning: The format of the PIDMAP file has changed.  THerefore, to
186    bootstrap you have to do this:
187    
188    1. Run CMB.make
189    2. Make a symbolic link for the boot directory:
190         ln -s sml.boot.ARCH-OS xxx
191    3. "Rebuild" the boot directory:
192         ./makeml -boot xxx -rebuild sml ; rm xxx
193    4. Boot normally:
194          ./makeml
195    
196    ----------------------------------------------------------------------
197    Name: Matthias Blume
198    Date: 2000/11/21 21:20:00 JST
199    Tag: blume-20001121-tools
200    Description:
201    
202    Continued hacking on autoloading problem -- with success this time.
203    Also changed tool-plugin mechanism.  See new CM manual.
204    
205    ----------------------------------------------------------------------
206    Name: Matthias Blume
207    Date: 2000/11/19 14:30:00 JST
208    Tag:  blume-20001119-autoload
209    Description:
210    
211    Some hacking to make autoloading faster.  Success for CMB, no success
212    so far for CM.  There is a reduced structure CM' that autoloads faster.
213    (This is a temporary, non-documented hack to be eliminated again when
214    the general problem is solved.)
215    
216    ----------------------------------------------------------------------
217    Name: Matthias Blume
218    Date: 2000/11/17 14:10:00 JST
219    Tag: blume-20001117-pickle-lib
220    Description:
221    
222    1. Eliminated comp-lib.cm
223    2. Made pickle-lib.cm
224    3. Eliminated all uses of intset.sml (from comp-lib.cm)
225    4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
226       equivalent constructs from smlnj-lib.cm (INtHashTable).
227    5. Point 4. also goes for those uses of intmap.* in MLRISC.
228       Duplicated intmap modules thrown out.
229    6. Hunted down all duplicated SCC code and replaced it with
230       equivalent stuff (GraphSCCFn from smlnj-lib.cm).
231    7. Rewrote Feedback module.
232    8. Moved sortedlist.sml into viscomp-lib.cm.  Eventually it
233       should be thrown out and equivalent modules from smlnj-lib.cm
234       should be used (IntRedBlackSet, IntListSet, ...).
235    
236    Confirmed that compiler compiles to fixpoint.
237    
238    ----------------------------------------------------------------------
239    Name: Allen Leung
240    Date: 2000/11/10 18:00:00
241    Tag: leunga-20001110-new-x86-fp
242    
243    A new x86 floating point code generator has been added.
244    By default this is turned off.  To turn this on, do:
245    
246        CM.autoload "$smlnj/compiler.cm";
247        Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
248    
249    Changes:
250    
251    1.  Changed FTAN to FPTAN so that the assembly output is correct.
252    2.  Changed the extension callback for FTANGENT to generate:
253    
254              fptan
255              fstp  %st(0)
256        instead of
257              fptan
258              fstpl ftempmem
259    
260    3.  Numerous assembly fixes for x86.
261    
262    5.  Cleaned up the machine code output module x86/x86MC.sml and added
263        support for a whole bunch of instructions and addressing modes:
264    
265          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st, %st(n)
266          faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp  %st, %st(n)
267          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st(n), %st
268          fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
269          fxch %st(n)
270          fld %st(n)
271          fst %st(n)
272          fst mem
273          fstp %st(n)
274          fucom %st(n)
275          fucomp %st(n)
276    
277        All these are now generated when the fast fp mode is turned on.
278    
279    6.  Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
280    
281    ----------------------------------------------------------------------
282    Name: Matthias Blume
283    Date: 2000/11/09 11:20:00 JST
284    Tag: blume-20001109-scc
285    Description:
286    
287    Eliminated some code duplication:
288    
289    1. Added "where" clause to GraphSCCFn in SML/NJ Library.
290       (Otherwise the functor is useless.)
291    2. Used GraphSCCFn where SCCUtilFun was used previously.
292    3. Got rid of SCCUtilFun (in comp-lib.cm).
293    
294    ----------------------------------------------------------------------
295    Name: Lal George
296    Date: 2000/11/06 09:02:21 EST 2000
297    Tag: Release_110_30
298    Description:
299    
300    - Version 110.30
301    ----------------------------------------------------------------------
302    Name: Matthias Blume
303    Date: 2000/11/04 14:45:00
304    Tag: blume-20001104-mlbuild
305    Description:
306    
307    - Made ml-build faster on startup.
308    - Documentation fixes.
309    
310    ----------------------------------------------------------------------
311    Name: Matthias Blume
312    Date: 2000/11/02 17:00:00 JST
313    Tag: blume-20001102-condcomp
314    Description:
315    
316    - Small tweaks to pickler -- new BOOTFILES!
317    - Version bumped to 110.29.2.
318    - Added conditional compilation facility to init.cmi (see comment there).
319    ----------------------------------------------------------------------
320    Name: Allen Leung
321    Date: 2000/10/23 19:31:00
322    Tag: leunga-20001023-demo-ra
323    
324    1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
325    2. Test programs for the graph library updated
326    3. Some new MLRISC demo programs added
327    
328    ----------------------------------------------------------------------
329    Name: Matthias Blume
330    Date: 2000/08/31 22:15:00 JST
331    Tag: blume-20001017-errmsg
332    Description:
333    
334    More error message grief: Where there used to be no messages, there
335    now were some that had bogus error regions.  Fixed.
336    
337    ----------------------------------------------------------------------
338    Name: Matthias Blume
339    Date: 2000/08/31 17:30:00 JST
340    Tag: blume-20001017-v110p29p1
341    Description:
342    
343    I made a version 110.29.1 with new bootfiles.
344    
345    Changes:  Modified pickler/unpickler for faster and leaner unpickling.
346              CM documentation changes and a small bugfix in CM's error reporting.
347    
348    ----------------------------------------------------------------------
349    Name: Lal George
350    Date: 2000/09/27 14:42:35 EDT
351    Tag: george-20000927-nodestatus
352    Description:
353    
354    Changed the type of the nodestatus, so that:
355    
356            SPILLED(~1)             is now SPILLED
357            SPILLED(m) where m>=0   is now MEMREG(m)
358            SPILLED(s) where s<~1   is now SPILL_LOC(~s)
359    
360    ----------------------------------------------------------------------
361    Name: Matthias Blume
362    Date: 2000/09/07 14:45:00 JST
363    Tag: blume-20000907-cmerrmsg
364    Description:
365    
366    Small tweak to CM to avoid getting ML syntax error messages twice.
367    
368    ----------------------------------------------------------------------
369    Name: Matthias Blume
370    Date: 2000/08/31 18:00:00 JST
371    Tag: blume-20000831-cvsbootfiles
372    Description:
373    
374    New URL for boot files (because the 110.29 files on the BL server do
375    now work correctly with my updated install scripts for yacc and lex).
376    
377    ----------------------------------------------------------------------
378    Name: Matthias Blume
379    Date: 2000/08/08 12:33:00 JST
380    Tag: blume-20000808-manual
381    Description:
382    
383    Tiny update to CM manual.
384    
385    ----------------------------------------------------------------------
386    Name: Allen Leung
387    Date: 2000/08/7 19:31:00
388    Tag: leunga-20000807-a-whole-bunch-of-stuff
389    
390      Moby, C--, SSA, x86, machine descriptions etc.  Should only affect C--
391    and Mobdy.
392    
393    1.  x86
394    
395       a.  Fixes to peephole module by John and Dan.
396       b.  Assembly fix to SETcc by Allen.
397       c.  Fix to c-call by John.
398       d.  Fix to spilling by John.  (This one deals with the missing FSTPT case)
399       e.  Instruction selection optimization to SETcc as suggested by John.
400    
401           For example,
402    
403            MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
404    
405           should generate:
406    
407            MOVL a, x
408            SUBL b, x
409            SHRL 31, x
410    
411    2.  IR stuff
412    
413         A bunch of new DJ-graph related algorithms added.  These
414         speed up SSA construction.
415    
416    3.  SSA + Scheduling
417    
418         Added code for SSA and scheduling to the repository
419    
420    ----------------------------------------------------------------------
421    Name: Lal George
422    Date: 2000/07/27 11:53:14 EDT
423    
424    Tag: lal-20000727-linux-ppc
425    Description:
426    
427     Made changes to support Linux PPC.
428     p.s. I have confirmation that the 110.29 boot files work fine.
429    
430    ----------------------------------------------------------------------
431    Name: Matthias Blume
432    Date: 2000/07/27 17:40:00 JST
433    Tag: blume-20000727-scripts
434    Description:
435    
436    !!!! WARNING !!!!
437    You must recompile the runtime system!
438    !!!! WARNING !!!!
439    
440    This is basically another round of script-enhancements:
441    
442    1. sml, ml-build, and ml-makedepend accept options -D and -U to define
443       and undefine CM preprocessor symbols.
444    
445    2. ml-build avoids generating a new heap image if it finds that the
446       existing one is still ok.  (The condition is that no ML file had to
447       be recompiled and all ML files are found to be older that the heap
448       file.)
449    
450       To make this work smoothly, I also hacked the runtime system as
451       well as SMLofNJ.SysInfo to get access to the heap image suffix
452       (.sparc-solaris, ...) that is currently being used.
453    
454       Moreover, the signature of CM.mk_standalone has changed.  See the
455       CM manual.
456    
457    3. ml-makedepend accepts additional options -n, -a, and -o.  (See the
458       CM manual for details.)
459    
460    4. More CM manual updates:
461        - all of the above has been documented.
462        - there is now a section describing the (CM-related) command line
463          arguments that are accepted by the "sml" command
464    
465    ----------------------------------------------------------------------
466    Name: Matthias Blume
467    Date: 2000/07/25 16:20:00 JST
468    Tag: blume-20000725-makedepend
469    Description:
470    
471    Added a script called ml-makedepend.  This can be used in makefiles
472    for Unix' make in a way very similar to the "makedepend" command for
473    C.
474    
475    The script internally uses function CM.sources.
476    
477    Synopsis:
478    
479        ml-makedepend [-f makefile] cmfile targetname
480    
481    The default for the makefile is "makefile" (or "Makefile" should
482    "makefile" not exist).
483    
484    ml-makedepend adds a cmfile/targetname-specific section to this
485    makefile (after removing the previous version of this section).  The
486    section contains a single dependency specification with targetname on
487    the LHS (targetname is an arbitrary name), and a list of files derived
488    from the cmfile on the RHS.  Some of the files on the RHS are
489    ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
490    "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
491    path names.  The makefile writer is responsible for making sure that
492    these variables have correct at the time "make" is invoked.
493    
494    ----------------------------------------------------------------------
495    Name: Matthias Blume
496    Date: 2000/07/22 23:30:00 JST
497    Tag: blume-20000722-urlupdate
498    Description:
499    
500    Changed BOOT and config/srcarchiveurl to point to BL server:
501    
502        ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
503    
504    ----------------------------------------------------------------------
505    Name: Matthias Blume
506    Date: 2000/07/18 18:00:00 JST
507    Tag: blume-20000718-Version_110_29
508    Description:
509    
510    1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
511    
512    2. Updated config/version to 110.29
513    
514    3. Updated config/srcarchiveurl
515    
516    3. New boot files!
517       ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
518    
519    ----------------------------------------------------------------------
520    Name: Matthias Blume
521    Date: 2000/07/11 13:58:00 JST
522    Tag: blume-20000711-doctypo
523    Description:
524    
525    Fixed a few typos in CM manual.
526    
527    ----------------------------------------------------------------------
528    Name: Allen Leung
529    Date: 2000/06/15 00:38:00
530    Tag: leunga-20000704-sparc-x86
531    
532    1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
533    2. fix to x86 RET bug [found by Dan Grossman]
534    3. sparc assembly bug fix for ticc instructions [found by Fermin]
535    
536       Affects c-- and moby only
537    
538    ----------------------------------------------------------------------
539    Name: Matthias Blume
540    Date: 2000/07/04 15:26:00
541    Tag: blume-20000704-trigger
542    Description:
543    
544    1. Improvements to CM manual.
545    2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
546       of getting a back-trace.  The function, when called, raises an
547       internal exception which explicitly carries the full back-trace history,
548       so it is unaffected by any intervening handle-raise pairs ("trivial"
549       or not).  The interactive loop will print that history once it arrives
550       at top level.
551       Short of having all exceptions implicitly carry the full history, the
552       recommended way of using this facility is:
553         - compile your program with instrumentation "on"
554         - run it, when it raises an exception, look at the history
555         - if the history is "cut off" because of some handler, go and modify
556           your program so that it explicitly calls BTrace.trigger
557         - recompile (still instrumented), and rerun; look at the full history
558    
559    ----------------------------------------------------------------------
560    Name: Matthias Blume
561    Date: 2000/07/03 15:36:00 JST
562    Tag: blume-20000702-manual
563    Description:
564    
565    Small corrections and updates to CM manual.
566    
567    ----------------------------------------------------------------------
568    Name: Matthias Blume
569    Date: 2000/06/29 16:04:00 JST
570    Tag: blume-20000629-yacctool
571    Description:
572    
573    Changes:
574    
575    1. Class "mlyacc" now takes separate arguments to pass options to
576       generated .sml- and .sig-files independently.
577    2. Corresponding CM manual updates.
578    3. BTrace module now also reports call sites.  (However, for loop clusters
579       it only shows from where the cluster was entered.)  There are associated
580       modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
581    
582    ----------------------------------------------------------------------
583    Name: Matthias Blume
584    Date: 2000/06/27 16:51:00 JST
585    Tag: blume-20000627-noweb
586    Description:
587    
588    Changes:
589    
590     1. Implemented "subdir" and "witness" options for noweb tool.
591        This caused some slight internal changes in CM's tool implementation.
592     2. Fixed bug in "tool plugin" mechanism.  This is essentially cleaning
593        some remaining issues from earlier path anchor changes.
594     3. Updated CM manual accordingly.
595    
596     4. Changed implementation of back-tracing so that I now consider it
597        ready for prime-time.
598    
599        In particular, you don't have to explicitly trigger the back-trace
600        anymore.  Instead, if you are running BTrace-instrumented code and
601        there is an uncaught exception (regardless of whether or not it was
602        raised in instrumented code), the top-level evalloop will print
603        the back-trace.
604    
605        Features:
606    
607          - Instrumented and uninstrumented code work together seemlessly.
608            (Of course, uninstrumented code is never mentioned in actual
609             back-traces.)
610    
611          - Asymptotic time- and space-complexity of instrumented code is
612            equal to that of uninstrumented code.  (This means that
613            tail-recursion is preserved by the instrumentation phase.)
614    
615          - Modules whose code has been instrumented in different sessions
616            work together without problem.
617    
618          - There is no penalty whatsoever on uninstrumented code.
619    
620          - There is no penalty on "raise" expressions, even in
621            instrumented code.
622    
623        A potential bug (or perhaps it is a feature, too):
624    
625          A back-trace reaches no further than the outermost instrumented
626          non-trivial "raise".  Here, a "trivial" raise is one that is the
627          sole RHS of a "handle" rule.  Thus, back-traces reach trough
628    
629               <exp> handle e => raise e
630    
631          and even
632    
633               <exp> handle Foo => raise Bar
634    
635          and, of course, through
636    
637               <exp> handle Foo => ...
638    
639         if the exception was not Foo.
640    
641         Back-traces always reach right through any un-instrumented code
642         including any of its "handle" expressions, trivial or not.
643    
644       To try this out, do the following:
645    
646         - Erase all existing binfiles for your program.
647           (You may keep binfiles for those modules where you think you
648            definitely don't need back-tracing.)
649         - Turn on back-trace instrumentation:
650              SMLofNJ.Internals.BTrace.mode (SOME true);
651         - Recompile your program.  (I.e., run "CM.make" or "use".)
652         - You may now turn instrumentation off again (if you want):
653              SMLofNJ.Internals.BTrace.mode (SOME false);
654         - Run your program as usual.  If it raises an exception that
655           reaches the interactive toplevel, then a back-trace will
656           automatically be printed.  After that, the toplevel loop
657           will print the exception history as usual.
658    
659    ----------------------------------------------------------------------
660    Name: Matthias Blume
661    Date: 2000/06/26 09:56:46 JST
662    Tag: blume-20000626-setup
663    Description:
664    
665    CM: - setup-parameter to "sml" added; this can be used to run arbitrary
666          ML code before and after compiling a file (e.g., to set compiler
667          flags)
668    
669    Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
670              - associated changes to btrace.sml (BTrace instrumentation pass)
671              - cleaner implementation of btimp.sml (BTrace tracing and report
672                module)
673    
674    CM manual: * new path encoding documented
675               * description of setup-parameter to "sml" added
676    
677    The biggest user-visible change to back-tracing is that it is no
678    longer necessary to compile all traced modules within the same
679    session.  (This was a real limitation.)
680    
681    ----------------------------------------------------------------------
682    Name: Matthias Blume
683    Date: 2000/06/24 12:40:00 JST
684    Tag: blume-20000624-startup
685    Description:
686    
687    Fixes startup slowdown problem.  (I was calling SrcPath.sync a _tad_
688    bit too often -- to put it mildly. :)
689    
690    ----------------------------------------------------------------------
691    Name: Matthias Blume
692    Date: 2000/06/23 18:20:00 JST
693    Tag: blume-20000623-btrace
694    Description:
695    
696    This updates adds a backtrace facility to aid programmers in debugging
697    their programs.  This involves the following changes:
698    
699    1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
700       keeping track of the current call stack.  When programs are compiled
701       in a special mode, the compiler will insert calls to these hooks
702       into the user program.
703       "Hook" means that it is possible for different implementations of
704       back-tracing to register themselves (at different times).
705    
706    2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
707       as an Absyn.dec->Absyn.dec rewrite.  Normally this phase is turned off.
708       It can be turned on using this call:
709         SMLofNJ.Internals.BTrace.mode (SOME true);
710       Turning it off again:
711         SMLofNJ.Internals.BTrace.mode (SOME false);
712       Querying the current status:
713         SMLofNJ.Internals.BTrace.mode NONE;
714       Annotated programs are about twice as big as normal ones, and they
715       run a factor of 2 to 4 slower with a dummy back-trace plugin (one
716       where all hooks do nothing).  The slowdown with a plugin that is
717       actually useful (such as the one supplied by default) is even greater,
718       but in the case of the default plugin it is still only an constant
719       factor (amortized).
720    
721    3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
722       with a sub-structure BTrace for controlling back-tracing.  In particular,
723       the above-mentioned function "mode" controls whether the annotation
724       phase is invoked by the compiler.  Another important function is
725       "trigger": when called it aborts the current execution and causes
726       the top-level loop to print a full back-trace.
727    
728    4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
729       for back-tracing.  It keeps track of the dynamic call stack and in
730       addition to that it keeps a partial history at each "level" of that
731       stack.  For example, if a tail-calls b, b tail-calls c, and c tail-calls
732       d and b (at separate times, dynamically), then the report will show:
733    
734       GOTO   d
735             /c
736       GOTO  \b
737       CALL   a
738    
739       This shows that there was an initial non-tail call of a, then a
740       tail-call to b or c, looping behavior in a cluster of functions that
741       consist of b and c, and then a goto from that cluster (i.e., either from
742       b or from c) to d.
743    
744       Note that (depending on the user program) the amount of information
745       that the back-trace module has to keep track of at each level is bounded
746       by a constant.  Thus, the whole implementation has the same asymptotical
747       complexity as the original program (both in space and in time).
748    
749    5. compiler/TopLevel/interact/evalloop.sml has been modified to
750       handle the special exception SMLofNJ.Internals.BTrace.BTrace
751       which is raised by the "trigger" function mentioned above.
752    
753    Notes on usage:
754    
755    - Annotated code works well together with unannotated code:
756    Unannotated calls simply do not show up at all in the backtrace.
757    
758    - It is not a good idea to let modules that were annotated during
759    different sessions run at the same time.  This is because the compiler
760    chooses small integers to identify individual functions, and there
761    will be clashes if different modules were compiled in separate sessions.
762    (Nothing will crash, and you will even be told about the clashes, but
763    back-trace information will in general not be useful.)
764    
765    - Back-tracing can be confused by callcc and capture.
766    
767    - The only way of getting a back-trace right now is to explicitly
768    invoke the "trigger" function from your user program.  Eventually, we
769    should make every exception carry back-trace information (if
770    available).  But since this creates more overhead at "raise"-time
771    (similar to the current exnHistory overhead), I have not yet
772    implemented this.  (The implementation will be rather easy.)  With
773    exceptions carrying back-trace information, this facility will be even
774    more useful because users don't need to modify their programs...
775    
776    - While it is possible to compile the compiler with back-trace
777    annotations turned on (I did it to get some confidence in
778    correctness), you must make absolutely sure that core.sml and
779    btimp.sml are compiled WITHOUT annotation!  (core.sml cannot actually
780    be compiled with annotation because there is no core access yet, but
781    if you compile btimp.sml with annotation, then the system will go into
782    an infinite recursion and crash.)
783    Since CM currently does not know about BTrace, the only way to turn
784    annotations on and off for different modules of the compiler is to
785    interrupt CMB.make, change the settings, and re-invoke it.  Of course,
786    this is awkward and clumsy.
787    
788    Sample sessions:
789    
790    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
791    - SMLofNJ.Internals.BTrace.mode (SOME true);
792    [autoloading]
793    [autoloading done]
794    val it = false : bool
795    - structure X = struct
796    -     fun main n = let
797    -         fun a (x, 0) = d x
798    -           | a (x, n) = b (x, n - 1)
799    -         and b (x, n) = c (x, n)
800    -         and c (x, n) = a (x, n)
801    -         and d x = e (x, 3)
802    -         and e (x, 0) = f x
803    -           | e (x, n) = e (x, n - 1)
804    -         and f 0 = SMLofNJ.Internals.BTrace.trigger ()
805    -           | f n = n * g (n - 1)
806    -         and g n = a (n, 3)
807    -     in
808    -         f n
809    -     end
810    - end;
811    structure X : sig val main : int -> int end
812    - X.main 3;
813    *** BACK-TRACE ***
814    GOTO   stdIn:4.2-13.20: X.main[2].f
815    GOTO-( stdIn:4.2-13.20: X.main[2].e
816    GOTO   stdIn:4.2-13.20: X.main[2].d
817         / stdIn:4.2-13.20: X.main[2].a
818         | stdIn:4.2-13.20: X.main[2].b
819    GOTO-\ stdIn:4.2-13.20: X.main[2].c
820    CALL   stdIn:4.2-13.20: X.main[2].g
821    GOTO   stdIn:4.2-13.20: X.main[2].f
822    GOTO-( stdIn:4.2-13.20: X.main[2].e
823    GOTO   stdIn:4.2-13.20: X.main[2].d
824         / stdIn:4.2-13.20: X.main[2].a
825         | stdIn:4.2-13.20: X.main[2].b
826    GOTO-\ stdIn:4.2-13.20: X.main[2].c
827    CALL   stdIn:4.2-13.20: X.main[2].g
828    GOTO   stdIn:4.2-13.20: X.main[2].f
829    GOTO-( stdIn:4.2-13.20: X.main[2].e
830    GOTO   stdIn:4.2-13.20: X.main[2].d
831         / stdIn:4.2-13.20: X.main[2].a
832         | stdIn:4.2-13.20: X.main[2].b
833    GOTO-\ stdIn:4.2-13.20: X.main[2].c
834    CALL   stdIn:4.2-13.20: X.main[2].g
835    GOTO   stdIn:4.2-13.20: X.main[2].f
836    CALL   stdIn:2.15-17.4: X.main[2]
837    -
838    
839    (Note that because of a FLINt bug the above code currently does not
840    compile without BTrace turned on.)
841    
842    Here is another example, using my modified Tiger compiler:
843    
844    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
845    - SMLofNJ.Internals.BTrace.mode (SOME true);
846    [autoloading]
847    [autoloading done]
848    val it = false : bool
849    - CM.make "sources.cm";
850    [autoloading]
851    ...
852    [autoloading done]
853    [scanning sources.cm]
854    [parsing (sources.cm):parse.sml]
855    [creating directory CM/SKEL ...]
856    [parsing (sources.cm):tiger.lex.sml]
857    ...
858    [wrote CM/sparc-unix/semant.sml]
859    [compiling (sources.cm):main.sml]
860    [wrote CM/sparc-unix/main.sml]
861    [New bindings added.]
862    val it = true : bool
863    - Main.compile ("../testcases/merge.tig", "foo.out");
864    *** BACK-TRACE ***
865    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
866    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
867    CALL   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
868    GOTO   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
869    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
870    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
871    CALL   lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
872         / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
873    CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
874    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
875    CALL   lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
876    CALL   main.sml:1.18-118.4: Main.compile[2]
877    -
878    
879    ----------------------------------------------------------------------
880    Name: Matthias Blumen
881    Date: 2000/06/21 18:00:00 JST
882    Tag: blume-20000621-manual
883    Description:
884    
885    CM manual update: Path environments documented.
886    
887    ----------------------------------------------------------------------
888    Name: Matthias Blume
889    Date: 2000/06/19 13:40:00
890    Tag: blume-20000619-manual
891    Description:
892    
893    CM manual and system/README update.  This only covers the fact that
894    there are no more implicit anchors.  (Path environments and the "bind"
895    option to "cm" have yet to be documented.)
896    
897    ----------------------------------------------------------------------
898    Name: Matthias Blume
899    Date: 2000/06/19 11:05:00 JST
900    Tag: blume-20000619-chdir-bugfix
901    Description:
902    
903    Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
904    
905    ----------------------------------------------------------------------
906    Name: Matthias Blume
907    Date: 2000/06/18 22:00:10 JST
908    Tag: blume-20000618-implicit-anchors-really-gone
909    Description:
910    
911    I updates the previous HISTORY entry where I forgot to mention that
912    implicit anchors are no longer with us.
913    
914    The current update also gets rid of the (now useless) controller
915    CM.Control.implicit_anchors.
916    
917    ----------------------------------------------------------------------
918    Name: Matthias Blume
919    Date: 2000/06/16 17:30:00 JST
920    Tag: blume-20000616-anchorenv
921    Description:
922    
923    This patch implements the long anticipated (just kidding :) "anchor
924    environment" mechanism.  In the course of doing this, I also
925    re-implemented CM's internal "SrcPath" module from scratch.  The new
926    one should be more robust in certain boundary cases.  In any case, it
927    is a lot cleaner than its predecessor (IMHO).
928    
929    This time, although there is yet another boot file format change, I
930    kept the unpickler backward-compatible.  As a result, no new bootfiles
931    are necessary and bootstrapping is straightforward.  (You cannot read
932    new bootfiles into an old system, but the other way around is no
933    problem.)
934    
935    Visible changes:
936    
937    ** 0. Implicit path anchors (without the leading $-symbol) are no
938    longer recognized at all. This means that such path names are not
939    illegal either.  For example, the name basis.cm simply refers to a
940    local file called "basis.cm" (i.e, the name is an ordinary path
941    relative to .cm-files directory).  Or, to put it differently, only
942    names that start with $ are anchored paths.
943    
944    ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
945    vanished.
946    
947    John (Reppy) had critizised this as soon as I originally proposed and
948    implemented it, but at that time I did not really deeply believe
949    him. :) Now I came full-circle because I need the $<singlearc> syntax
950    in another place where it cannot be seen as an abbreviation for
951    $/<singlearc>.  To avoid the confusion, $<singlearc> now means what it
952    seems to mean (i.e., it "expands" into the corresponding anchor
953    value).
954    
955    However, when paths are used as members in CM description files, it
956    continues to be true that there must be at least another arc after the
957    anchor.  This is now enforced separately during semantic analysis
958    (i.e., from a lexical/syntactical point of view, the notation is ok.)
959    
960    ** 2. The "cm" class now accepts an option "bind".  The option's value
961    is a sub-option list of precisely two items -- one labeled "anchor"
962    and the other one labeled "value".  As you might expect, "anchor" is
963    used to specify an anchor name to be bound, and "value" specifies what
964    the anchor is being bound to.
965    
966    The value must be a directory name and can be given in either standard
967    syntax (including the possibility that it is itself an anchored path)
968    or native syntax.
969    
970    Examples:
971    
972       foo.cm (bind:(anchor:bar value:$mystuff/bar))
973       lib.cm (bind:(anchor:a value:"H:\\x\\y\\z"))  (* only works under windows *)
974    
975    and so on.
976    
977    The meaning of this is that the .cm-file will be processed with an
978    augmented anchor environment where the given anchor(s) is/are bound to
979    the given values(s).
980    
981    The rationale for having this feature is this: Suppose you are trying
982    to use two different (already stable) libraries a.cm and b.cm (that
983    you perhaps didn't write yourself).  Further, suppose each of these
984    two libraries internally uses its own auxiliary library $aux/lib.cm.
985    Normally you would now have a problem because the anchor "lib" can not
986    be bound to more than one value globally.  Therefore, the project that
987    uses both a.cm and b.cm must locally redirect the anchor to some other
988    place:
989    
990       a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
991       b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
992    
993    This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
994    /usr/lib/smlnj/b-stuff/aux.cm, respectively.
995    
996    Hard-wiring path names is a bit inflexible (and CM will verbosely warn
997    you when you do so at the time of CM.stabilize).  Therefore, you can
998    also use an anchored path as the value:
999    
1000      a.cm (bind:(anchor:lib value:$a-lib))
1001      b.cm (bind:(anchor:lib value:$b-lib))
1002    
1003    Now you can globally configure (using the usual CM.Anchor.anchor or
1004    pathconfig machinery) bindings for "a-lib" and "b-lib".  Since "lib"
1005    itself is always locally bound, setting it globally is no longer
1006    meaningful or necessary (but it does not hurt either).  In fact, "lib"
1007    can still be used as a global anchor for separate purposes.  As a
1008    matter of fact, one can locally define "lib" in terms of a global
1009    "lib":
1010    
1011      a.cm (bind:(anchor:lib value:$lib/a))
1012      b.cm (bind:(anchor:lib value:$lib/b))
1013    
1014    ** 3: The encoding of path names has changed.  This affects the way
1015    path names are shown in CM's progress report and also the internal
1016    protocol encoding used for parallel make.
1017    
1018    The encoding now uses one or more ':'-separated segments.  Each
1019    segments corresponds to a file that has been specified relative to the
1020    file given by its preceding segment.  The first segment is either
1021    relative to the CWD, absolute, or anchored.  Each segment itself is
1022    basically a Unix pathname; all segments but the first are relative.
1023    
1024    Example:
1025    
1026       $foo/bar/baz.cm:a/b/c.sml
1027    
1028    This path denotes the file bar/a/b/c.sml relative to the directory
1029    denoted by anchor "foo".  Notice that the encoding also includes
1030    baz.cm which is the .cm-file that listed a/b/c.sml.  As usual, such
1031    paths are resolved relative to the .cm-files directory, so baz.cm must
1032    be ignored to get the "real" pathname.
1033    
1034    To make this fact more obvious, CM puts the names of such "virtual
1035    arcs" into parentheses when they appear in progress reports. (No
1036    parentheses will appear in the internal protocol encoding.)  Thus,
1037    what you really see is:
1038    
1039      $foo/bar/(baz.cm):a/b/c.sml
1040    
1041    I find this notation to be much more informative than before.
1042    
1043    Another new feature of the encoding is that special characters
1044    including parentheses, colons, (back)slashes, and white space are
1045    written as \ddd (where ddd is the decimal encoding of the character).
1046    
1047    *** The CM manual still needs to be updated.
1048    
1049    ----------------------------------------------------------------------
1050    Name: Allen Leung
1051    Date: 2000/06/15 00:38:00
1052    Tag: leunga-20000615-x86-peephole
1053    
1054    x86 Peephole fix by Fermin.  Affects c-- and moby only.
1055    
1056    ----------------------------------------------------------------------
1057    Name: Matthias Blume
1058    Date: 2000/06/12 11:40:00
1059    Tag: blume-20000612-parmakefix
1060    Description:
1061    
1062    More cleanup after changing the file naming scheme: This time I
1063    repaired the parallel make mechanism for CMB.make which I broke earlier.
1064    
1065    ----------------------------------------------------------------------
1066    Name: Allen Leung
1067    Date: 2000/06/09 01:25:00
1068    Tag: leunga-20000609-various
1069    
1070    None of these things should affect normal SML/NJ operations
1071    
1072    1. Peephole improvements provided by Fermin (c--)
1073    2. New annotation DEFUSE for adding extra dependence (moby)
1074    3. New X86 LOCK instructions (moby)
1075    4. New machine description language for reservation tables (scheduling)
1076    5. Fixes to various optimization/analysis modules (branch chaining, dominator
1077       trees etc.)
1078    6. I've changed the CM files so that they can work with versions
1079       110.0.6, 110.25 and 110.28
1080    
1081    ----------------------------------------------------------------------
1082    Name: Matthias Blume
1083    Date: 2000/06/09 12:40:00
1084    Tag: blume-20000609-log
1085    Description:
1086    
1087    - Removed all(?) remaining RCS Log entries from sources.
1088    
1089    - Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
1090      anchored paths).
1091    
1092    ----------------------------------------------------------------------
1093    Name: Matthias Blume
1094    Date: 2000/06/07 17:00:00 JST
1095    Tag: blume-20000607-no-implicit-anchors
1096    Description:
1097    
1098    1. This update changes the default setting for
1099    CM.Control.implicit_anchors from true to false.  This means that
1100    implicit anchors are no longer permitted by default.  I also tried to
1101    make sure that nothing else still relies on implicit anchors.
1102    (This is the next step on the schedule towards a CM that does not even
1103    have the notion of implicit anchors anymore.)
1104    
1105    2. More CM manual updates.
1106    
1107    3. I managed to track down and fix the pickling bug I mentioned last
1108    time.  Because of the previously existing workaround, this entails no
1109    immediate practical changes.
1110    
1111    ----------------------------------------------------------------------
1112    Name: Matthias Blume
1113    Date: 2000/06/06 11:15:00 JST
1114    Tag: blume-20000606-lazierpickle
1115    Description:
1116    
1117    !!!! NEW BOOT FILES !!!!
1118    
1119    * The main purpose of this update is to make library pickles lazier in
1120    order to reduce the initial space penalty for autoloading a library.
1121    As a result, it is now possible to have $smlnj/compiler.cm
1122    pre-registered.  This should take care of the many complaints or
1123    inquiries about missing structure Compiler.  This required changes to
1124    CM's internal data structures and small tweaks to some algorithms.
1125    
1126    As a neat additional effect, it is no longer necessary (for the sake
1127    of lean heap image files) to distinguish between a "minimal" CM and a
1128    "full" CM.  Now, there is only one CM (i.e., the "full" version:
1129    $smlnj/cm.cm aka $smlnj/cm/full.cm), and it is always available at the
1130    interactive top level. ($smlnj/cm/minimal.cm is gone.)
1131    
1132    To make the life of compiler-hackers easier, "makeml" now also
1133    pre-registers $smlnj/cmb.cm (aka $smlnj/cmb/current.cm).  In other
1134    words, after you bootstrap a new sml for the first time, you will not
1135    have to autoload $smlnj/cmb.cm again afterwards.  (The first time
1136    around you will still have to do it, though.)
1137    
1138    * A second change consists of major updates to the CM manual.  There
1139    are now several appendices with summary information and also a full
1140    specification of the CM description file syntax.
1141    
1142    * In directory src/system I added the script "allcross".  This script
1143    invokes sml and cross-compiles the compiler for all supported
1144    architectures.  (Useful when providing a new set of boot files.)
1145    
1146    * There seems to be a latent bug in my "lazy pickles" mechanism.  I
1147    added a small tweak to pickle-util.sml to work around this problem,
1148    but it is not a proper fix yet.  I will investigate further.  (The
1149    effect of the bug was an inflation of library pickle size.)
1150    
1151    * Version number increased to 110.28.1 (to avoid compatibility problems).
1152    
1153    ----------------------------------------------------------------------
1154    Name: Allen Leung
1155    Date: 2000/05/25 17:28 EDT
1156    Tag: leunga-20000525-ra
1157    Description:
1158    
1159      Fixed a bug in freezing phase of the register allocator.
1160    
1161    ----------------------------------------------------------------------
1162  Name: Allen Leung  Name: Allen Leung
1163  Date: 2000/05/15 22:53 EDT  Date: 2000/05/15 22:53 EDT
1164  Tag: leunga-20000515-alpha-x86-ra  Tag: leunga-20000515-alpha-x86-ra
# Line 494  Line 1641 
1641    
1642  3.  Assembly  3.  Assembly
1643    
1644      When generating assemby, resolve the value of client defined constants,      When generating assembly, resolve the value of client defined constants,
1645      instead of generating symbolic values.  This is controlled by the      instead of generating symbolic values.  This is controlled by the
1646      new flag "asm-resolve-constants", which is default to true.      new flag "asm-resolve-constants", which is default to true.
1647    
# Line 517  Line 1664 
1664    
1665  To this end, I arranged that instead of "structure Core" as "structure  To this end, I arranged that instead of "structure Core" as "structure
1666  _Core" is bound in the pervasive environment.  Core access is done via  _Core" is bound in the pervasive environment.  Core access is done via
1667  _Core (which can never be accidentially rebound because _Core is not a  _Core (which can never be accidentally rebound because _Core is not a
1668  legal surface-syntax symbol).  legal surface-syntax symbol).
1669    
1670  The current solution is much cleaner because the core environment is  The current solution is much cleaner because the core environment is
# Line 527  Line 1674 
1674  with dynamic and symbolic parts of the core environment.  with dynamic and symbolic parts of the core environment.
1675    
1676  Remaining hackery (to bind the "magic" symbol _Core) is localized in the  Remaining hackery (to bind the "magic" symbol _Core) is localized in the
1677  compilation mananger's bootstrap compiler (actually: in the "init group"  compilation manager's bootstrap compiler (actually: in the "init group"
1678  handling).  See the comments in src/system/smlnj/init/init.cmi for  handling).  See the comments in src/system/smlnj/init/init.cmi for
1679  more details.  more details.
1680    
# Line 642  Line 1789 
1789     (specified in the .cm file at each instance where the tool's class is     (specified in the .cm file at each instance where the tool's class is
1790     used).     used).
1791    
1792     This was done to accomodate the new "make" and "shell" tools which     This was done to accommodate the new "make" and "shell" tools which
1793     facilitate fairly seemless hookup to portions of code managed using     facilitate fairly seamless hookup to portions of code managed using
1794     Makefiles or Shell scripts.     Makefiles or Shell scripts.
1795    
1796     There are no classes "shared" or "private" anymore.  Instead, the     There are no classes "shared" or "private" anymore.  Instead, the
# Line 655  Line 1802 
1802    
1803     All existing tools are described in the CM manual.     All existing tools are described in the CM manual.
1804    
1805     - Slightly better error handling.  (CM now surpresses many followup     - Slightly better error handling.  (CM now suppresses many followup
1806     error messages that tended to be more annoying than helpful.)     error messages that tended to be more annoying than helpful.)
1807    
1808  2. Major changes to the compiler's static environment data structures.  2. Major changes to the compiler's static environment data structures.
# Line 789  Line 1936 
1936    
1937       I've changed andl to testl in the floating point test sequence       I've changed andl to testl in the floating point test sequence
1938       whenever appropriate.  The Intel optimization guide states that       whenever appropriate.  The Intel optimization guide states that
1939       testl is perferable to andl.       testl is preferable to andl.
1940    
1941  7. RA (x86 only)  7. RA (x86 only)
1942    
# Line 971  Line 2118 
2118    
2119  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
2120  additional argument called "template" which is an optional string that  additional argument called "template" which is an optional string that
2121  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
2122  explanation.  explanation.
2123    
2124  2. A special-purpose tool can be "regisitered" by simply dropping the  2. A special-purpose tool can be "registered" by simply dropping the
2125  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
2126  directory where the .cm file lives that uses this tool.  (The  directory where the .cm file lives that uses this tool.  (The
2127  behavior/misfeature until now was to look for the tool description  behavior/misfeature until now was to look for the tool description
# Line 1018  Line 2165 
2165  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
2166  <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,
2167  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
2168  perhaps got accidentially killed).  In this case, fixpt will quickly  perhaps got accidentally killed).  In this case, fixpt will quickly
2169  move through what exists before continuing where it left off earlier,  move through what exists before continuing where it left off earlier,
2170  and, thus, saves a lot of time.  and, thus, saves a lot of time.
2171    
# Line 1068  Line 2215 
2215    it from that remote directory.    it from that remote directory.
2216    This should simplify installation further:  For machines that have    This should simplify installation further:  For machines that have
2217    access to the internet, just fetch <version>-config.tgz, unpack it,    access to the internet, just fetch <version>-config.tgz, unpack it,
2218    edit config/targets, and go (run config/install.sh).  The scipt will    edit config/targets, and go (run config/install.sh).  The script will
2219    fetch everything else that it might need all by itself.    fetch everything else that it might need all by itself.
2220    
2221    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.647  
changed lines
  Added in v.771

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