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 593, Tue Apr 4 07:26:33 2000 UTC revision 801, Mon Mar 19 22:53:00 2001 UTC
# Line 11  Line 11 
11  Date:  Date:
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    
15    ----------------------------------------------------------------------
16    Name: Matthias Blume
17    Date: 2001/03/19 17:53:00 EST
18    Tag: blume-20010319-parallel
19    Description:
20    
21    Parallel make (using compile servers) now works again.
22    
23    To this end, CM.stabilize and CMB.make have been modified to work in
24    two passes when compile servers are attached:
25       1. Compile everything, do not perform stabilization; this pass
26          uses compile servers
27       2. Stabilize everything; this pass does not use compile servers
28    If there are no compile servers, the two passes are combined into one
29    (as before).  Splitting the passes increases the inherent parallelism
30    in the dependency graph because the entire graph including all
31    libraries is available at the same time.  This, in turn, improves
32    server utilization.  The downside is that the master process will
33    have to do some extra work after compilation is done (because for
34    technical reasons it must re-read all the binfiles during stabilization).
35    
36    ----------------------------------------------------------------------
37    Name: Matthias Blume
38    Date: 2001/03/16 12:22:00 EST
39    Tag: blume-20010316-bootfiles
40    Description:
41    
42    Created a new set of bootfiles (for your automatic installation convenience).
43    
44    ----------------------------------------------------------------------
45    Name: Matthias Blume
46    Date: 2001/03/16 11:00:00 EST
47    Tag: blume-20010316-MLTREE-fixup
48    Description:
49    
50    This is a minor fixup for an (untagged) earlier commit by Allen.
51    (A file was missing).
52    
53    ----------------------------------------------------------------------
54    Name: Allen Leung
55    Date: Mon Mar  5 18:54:57 EST 2001
56    Tag: leunga-20010305-cut-support
57    
58    1. New support for alternative control-flow in MLTREE.
59       Currently we support
60    
61          FLOW_TO(CALL ...., [k1,...,kn])
62    
63       This is needed for 'cuts to' in C-- and try/handle-like constructs
64       in Moby
65    
66       New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
67    
68    2. Register Allocator
69    
70       Changes in interface [from Fermin, John]
71    
72    3. Alpha 8-bit SLL support [Fermin]
73    
74    4. All architectures
75    
76       A new module (ClusterExpandCopies) for expanding parallel copies.
77    
78    ----------------------------------------------------------------------
79    Name: Allen Leung
80    Date: 2001/02/27 23:07:00 EST
81    Tag: leunga-20010227-minor-stuff
82    
83    1. Alpha bug fix for CMOVNE
84    2. Handle mltree COND(..,FCMP ...,...)
85    3. Bug fix in simplifier
86    
87    ----------------------------------------------------------------------
88    Name: Matthias Blume
89    Date: 2001/01/30 17:50:00 EST
90    Tag: blume-20010130-sync
91    Description:
92    
93    This is just a minor update to sync my devel branch with the main brach.
94    The only visible change is the addition of some README files.
95    
96    ----------------------------------------------------------------------
97    Name: Matthias Blume
98    Date: 2001/01/12 23:30:00 JST
99    Tag: blume-20010112-bootfiles
100    Description:
101    
102    Made a new set of bootfiles that goes with the current state of the
103    repository.
104    
105    ----------------------------------------------------------------------
106    Name: Matthias Blume
107    Date: 2001/01/12 21:20:00 JST
108    Tag: blume-20010112-sync
109    Description:
110    
111    I am just flushing out some minor changes that had accumulated in
112    my private branch in order to sync with the main tree.  (This is
113    mainly because I had CVS trouble when trying to merge _into_ my
114    private branch.)
115    
116    Most people should be completely unaffected by this.
117    
118    ----------------------------------------------------------------------
119    Name: Allen Leung
120    Date: Thu Jan 11 21:03:00 EST 2001
121    Tag: leunga-20010111-labexp=mltree
122    Description:
123    
124    1.  Removed the type LabelExp and replace it by MLTree.
125    2.  Rewritten mltree-simplify with the pattern matcher tool.
126    3.  There were some bugs in alpha code generator which would break
127        64-bit code generation.
128    4.  Redo the tools to generate code with the
129    5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
130        are now generated by perl scripts.
131    
132    ----------------------------------------------------------------------
133    Name: Matthias Blume
134    Date: 2001/01/10 21:55:00 JST
135    Tag: blume-20010110-rcc
136    Description:
137    
138    The RCC stuff now seems to work (but only on the x86).
139    This required hacking of the c-calls interface (and -implementation) in
140    MLRISC.
141    
142    Normal compiler users should be unaffected.
143    
144    ----------------------------------------------------------------------
145    Name: Matthias Blume
146    Date: 2001/01/09 01:20:00 JST
147    Tag: blume-20010109-rcc
148    Description:
149    
150    This is a fairly big patch, flushing out a large number of pending
151    changes that I made to my development copy over the last couple of days.
152    
153    Of practical relevance at this moment is a workaround for a pickling
154    bug that Allen ran into the other day.  The cause of the bug itself is
155    still unknown and it might be hard to fix it properly, but the
156    workaround has some merits of its own (namely somewhat reducing pickling
157    overhead for certain libraries).  Therefore, I think this solution should
158    be satisfactory at this time.
159    
160    The rest of the changes (i.e., the vast majority) has to do with my
161    ongoing efforts of providing direct support for C function calls from
162    ML.  At the moment there is a new primop "RAW_CCALL", typing magic
163    in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
164    case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
165    translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
166    handle RCC.
167    
168    The last part (the changes to mlriscGen.sml) are still known to be
169    wrong on the x86 and not implemented on all other architectures.  But
170    the infrastructure is in place. I had to change a few functor
171    signatures in the backend to be able to route the CCalls interface
172    from MLRISC there, and I had to specialize the mltree type (on the
173    x86) to include the necessary extensions. (The extensions themselves
174    were already there and redy to go in MLRISC/x86).
175    
176    Everything should be very happy as soon as someone helps me with
177    mlriscGen.sml...
178    
179    In any case, nothing of this should matter to anyone as long as the
180    new primop is not being used (which is going to be the case unless you
181    find it where I hid it :). The rest of the compiler is completely
182    unaffected.
183    
184    ----------------------------------------------------------------------
185    Name: Matthias Blume
186    Date: 2001/01/05 00:30:00 JST
187    Tag: blume-20010105-primops
188    Description:
189    
190    Added some experimental support for work that I am doing right now.
191    These changes mostly concern added primops, but there is also a new
192    experimental C library in the runtime system (but currently not enabled
193    anywhere except on Linux/X86).
194    
195    In the course of adding primops (and playing with them), I discovered that
196    Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
197    broken.  (Zhong was very right he labeled this stuff as "major gross hack".)
198    To recover, I made type information in INL_PRIM mandatory and changed
199    prim.sml as well as built-in.sml accordingly.  The InLine structure now
200    has complete, correct type information (i.e., no bottom types).
201    
202    Since all these changes mean that we need new binfiles, I also bumped the
203    version number to 110.32.1.
204    
205    ----------------------------------------------------------------------
206    Name: Matthias Blume
207    Date: 2000/12/30 22:10:00 JST
208    Tag: blume-20001230-various
209    Description:
210    
211    Added proxy libraries for MLRISC and let MLRISC libraries refer
212    to each other using path anchors.  (See CM manual for explanation.)
213    
214    Updated CM documentation.
215    
216    Fixed some bugs in CM.
217    
218    Implemented "proxy" libraries (= syntactic sugar for CM).
219    
220    Added "-quiet" option to makeml and changed runtime system accordingly.
221    
222    Added cleanup handler for exportML to reset timers and compiler stats.
223    
224    ----------------------------------------------------------------------
225    Name: Lal George
226    Date: 2000/12/22 22:22:58 EST 2000
227    Tag: Release_110_32
228    Description:
229    
230            Infinite precision used throughout MLRISC.
231            see MLRISC/mltree/machine-int.sig
232    
233    ----------------------------------------------------------------------
234    Name: Matthias Blume
235    Date: 2000/12/22 23:16:00 JST
236    Tag: blume-20001222-warn
237    Description:
238    
239    Corrected wording and formatting of some CM warning message which I
240    broke in my previous patch.
241    
242    ----------------------------------------------------------------------
243    Name: Matthias Blume
244    Date: 2000/12/22 21:20:00 JST
245    Tag: blume-20001222-anchorenv
246    Description:
247    
248    Fixed CM's handling of anchor environments in connection with CMB.make.
249    
250    ----------------------------------------------------------------------
251    Name: Matthias Blume
252    Date: 2000/12/22 13:15:00 JST
253    Tag: blume-20001222-cleanup
254    Description:
255    
256    Removed src/cm/ffi which does not (and did not) belong here.
257    
258    ----------------------------------------------------------------------
259    Name: Matthias Blume
260    Date: 2000/12/21 23:55:00 JST
261    Tag: blume-20001221-exn
262    Description:
263    
264    Probably most important: CM no longer silently swallows all exceptions
265    in the compiler.
266    Plus: some other minor CM changes.  For example, CM now reports some
267    sizes for generated binfiles (code, data, envpickle, lambdapickle).
268    
269    ----------------------------------------------------------------------
270    Name: Matthias Blume
271    Date: 2000/12/15 00:01:05 JST
272    Tag: blume-20001215-dirtool
273    Description:
274    
275    - "dir" tool added.
276    - improvements and cleanup to Tools structure
277    - documentation updates
278    
279    ----------------------------------------------------------------------
280    Name: Allen Leung
281    Date: Thu Dec 14 03:45:24 EST 2000
282    Description:
283    Tag:  leunga-20001214-int-inf
284    Description:
285    
286       In IntInf, added these standard functions, which are missing from our
287    implementation:
288    
289        andb : int * int -> int
290        xorb : int * int -> int
291        orb  : int * int -> int
292        notb : int -> int
293         <<   : int * word -> int
294        ~>>  : int * word -> int
295    
296       Not tested, I hope they are correct.
297    
298    ----------------------------------------------------------------------
299    Name: Allen Leung
300    Date: Fri Dec  8 19:23:26 EST 2000
301    Description:
302    Tag:  leunga-20001208-nowhere
303    Description:
304    
305      Slight improvements to the 'nowhere' tool to handle OR-patterns,
306    to generate better error messages etc.  Plus a brief manual.
307    
308    ----------------------------------------------------------------------
309    Name: Lal George
310    Date: 2000/12/08 09:54:02 EST 2000
311    Tag: Release_110_31
312    Description:
313    
314    - Version 110.31
315    ----------------------------------------------------------------------
316    Name: Allen Leung
317    Date: Thu Dec  7 22:01:04 EST 2000
318    Tag:  leunga-20001207-cell-monster-hack
319    Description:
320    
321    Major MLRISC internal changes.  Affect all clients.
322    Summary:
323    
324    1.  Type CELLS.cell = int is now replaced by a datatype.
325        As a result, the old regmap is now gone.  Almost all interfaces
326        in MLRISC change as a consequence.
327    
328    2.  A new brand version of machine description tool (v3.0) that generates
329        modules expecting the new interface.  The old version is removed.
330    
331    3.  The RA interface has been further abstracted into two new functors.
332        RISC_RA and X86RA.  These functors have much simpler interfaces.
333        [See also directory MLRISC/demo.]
334    
335    4.  Some other new source->source code generation tools are available:
336    
337        a. MLRISC/Tools/RewriteGen -- generate rewriters from rules.
338        b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules.
339           I use this tool to generate the peephole optimizers---with the new
340           cell type changes, peephole rules are becoming difficult to write
341           without conditional pattern matching.
342    
343    5.  More Intmap -> IntHashTable change.  Previous changes by Matthias didn't
344        cover the entire MLRISC source tree so many things broke.
345    
346    6.  CM files have been moved to the subdirectory MLRISC/cm.
347        They are moved because there are a lot of them and they clutter up the
348        root dir.
349    
350    7.  More detailed documentation to come...
351    
352        NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do
353        a makeml -rebuild first.  This is because of other other
354        changes that Matthias has made (see below).
355    
356    
357    ----------------------------------------------------------------------
358    Name: Matthias Blume
359    Date: 2000/11/30 23:12:00 JST
360    Tag: blume-20001130-filereorg
361    Description:
362    
363    Some manual updates and some file reorganizations in CM.
364    
365    ----------------------------------------------------------------------
366    Name: Matthias Blume
367    Date: 2000/11/24 17:45:00 JST
368    Tag: blume-20001124-link
369    Description:
370    
371    Drastically improved link traversal code for the case that the dynamic
372    value was already loaded at bootstrap time.  As a result, CM and CMB
373    now both load blazingly fast -- even on a very slow machine.  Also,
374    memory consumption has been further reduced by this.
375    
376    Warning: The format of the PIDMAP file has changed.  THerefore, to
377    bootstrap you have to do this:
378    
379    1. Run CMB.make
380    2. Make a symbolic link for the boot directory:
381         ln -s sml.boot.ARCH-OS xxx
382    3. "Rebuild" the boot directory:
383         ./makeml -boot xxx -rebuild sml ; rm xxx
384    4. Boot normally:
385          ./makeml
386    
387    ----------------------------------------------------------------------
388    Name: Matthias Blume
389    Date: 2000/11/21 21:20:00 JST
390    Tag: blume-20001121-tools
391    Description:
392    
393    Continued hacking on autoloading problem -- with success this time.
394    Also changed tool-plugin mechanism.  See new CM manual.
395    
396    ----------------------------------------------------------------------
397    Name: Matthias Blume
398    Date: 2000/11/19 14:30:00 JST
399    Tag:  blume-20001119-autoload
400    Description:
401    
402    Some hacking to make autoloading faster.  Success for CMB, no success
403    so far for CM.  There is a reduced structure CM' that autoloads faster.
404    (This is a temporary, non-documented hack to be eliminated again when
405    the general problem is solved.)
406    
407    ----------------------------------------------------------------------
408    Name: Matthias Blume
409    Date: 2000/11/17 14:10:00 JST
410    Tag: blume-20001117-pickle-lib
411    Description:
412    
413    1. Eliminated comp-lib.cm
414    2. Made pickle-lib.cm
415    3. Eliminated all uses of intset.sml (from comp-lib.cm)
416    4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
417       equivalent constructs from smlnj-lib.cm (INtHashTable).
418    5. Point 4. also goes for those uses of intmap.* in MLRISC.
419       Duplicated intmap modules thrown out.
420    6. Hunted down all duplicated SCC code and replaced it with
421       equivalent stuff (GraphSCCFn from smlnj-lib.cm).
422    7. Rewrote Feedback module.
423    8. Moved sortedlist.sml into viscomp-lib.cm.  Eventually it
424       should be thrown out and equivalent modules from smlnj-lib.cm
425       should be used (IntRedBlackSet, IntListSet, ...).
426    
427    Confirmed that compiler compiles to fixpoint.
428    
429    ----------------------------------------------------------------------
430    Name: Allen Leung
431    Date: 2000/11/10 18:00:00
432    Tag: leunga-20001110-new-x86-fp
433    
434    A new x86 floating point code generator has been added.
435    By default this is turned off.  To turn this on, do:
436    
437        CM.autoload "$smlnj/compiler.cm";
438        Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
439    
440    Changes:
441    
442    1.  Changed FTAN to FPTAN so that the assembly output is correct.
443    2.  Changed the extension callback for FTANGENT to generate:
444    
445              fptan
446              fstp  %st(0)
447        instead of
448              fptan
449              fstpl ftempmem
450    
451    3.  Numerous assembly fixes for x86.
452    
453    5.  Cleaned up the machine code output module x86/x86MC.sml and added
454        support for a whole bunch of instructions and addressing modes:
455    
456          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st, %st(n)
457          faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp  %st, %st(n)
458          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st(n), %st
459          fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
460          fxch %st(n)
461          fld %st(n)
462          fst %st(n)
463          fst mem
464          fstp %st(n)
465          fucom %st(n)
466          fucomp %st(n)
467    
468        All these are now generated when the fast fp mode is turned on.
469    
470    6.  Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
471    
472    ----------------------------------------------------------------------
473    Name: Matthias Blume
474    Date: 2000/11/09 11:20:00 JST
475    Tag: blume-20001109-scc
476    Description:
477    
478    Eliminated some code duplication:
479    
480    1. Added "where" clause to GraphSCCFn in SML/NJ Library.
481       (Otherwise the functor is useless.)
482    2. Used GraphSCCFn where SCCUtilFun was used previously.
483    3. Got rid of SCCUtilFun (in comp-lib.cm).
484    
485    ----------------------------------------------------------------------
486    Name: Lal George
487    Date: 2000/11/06 09:02:21 EST 2000
488    Tag: Release_110_30
489    Description:
490    
491    - Version 110.30
492    ----------------------------------------------------------------------
493    Name: Matthias Blume
494    Date: 2000/11/04 14:45:00
495    Tag: blume-20001104-mlbuild
496    Description:
497    
498    - Made ml-build faster on startup.
499    - Documentation fixes.
500    
501    ----------------------------------------------------------------------
502    Name: Matthias Blume
503    Date: 2000/11/02 17:00:00 JST
504    Tag: blume-20001102-condcomp
505    Description:
506    
507    - Small tweaks to pickler -- new BOOTFILES!
508    - Version bumped to 110.29.2.
509    - Added conditional compilation facility to init.cmi (see comment there).
510    ----------------------------------------------------------------------
511    Name: Allen Leung
512    Date: 2000/10/23 19:31:00
513    Tag: leunga-20001023-demo-ra
514    
515    1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
516    2. Test programs for the graph library updated
517    3. Some new MLRISC demo programs added
518    
519    ----------------------------------------------------------------------
520    Name: Matthias Blume
521    Date: 2000/08/31 22:15:00 JST
522    Tag: blume-20001017-errmsg
523    Description:
524    
525    More error message grief: Where there used to be no messages, there
526    now were some that had bogus error regions.  Fixed.
527    
528    ----------------------------------------------------------------------
529    Name: Matthias Blume
530    Date: 2000/08/31 17:30:00 JST
531    Tag: blume-20001017-v110p29p1
532    Description:
533    
534    I made a version 110.29.1 with new bootfiles.
535    
536    Changes:  Modified pickler/unpickler for faster and leaner unpickling.
537              CM documentation changes and a small bugfix in CM's error reporting.
538    
539    ----------------------------------------------------------------------
540    Name: Lal George
541    Date: 2000/09/27 14:42:35 EDT
542    Tag: george-20000927-nodestatus
543    Description:
544    
545    Changed the type of the nodestatus, so that:
546    
547            SPILLED(~1)             is now SPILLED
548            SPILLED(m) where m>=0   is now MEMREG(m)
549            SPILLED(s) where s<~1   is now SPILL_LOC(~s)
550    
551    ----------------------------------------------------------------------
552    Name: Matthias Blume
553    Date: 2000/09/07 14:45:00 JST
554    Tag: blume-20000907-cmerrmsg
555    Description:
556    
557    Small tweak to CM to avoid getting ML syntax error messages twice.
558    
559    ----------------------------------------------------------------------
560    Name: Matthias Blume
561    Date: 2000/08/31 18:00:00 JST
562    Tag: blume-20000831-cvsbootfiles
563    Description:
564    
565    New URL for boot files (because the 110.29 files on the BL server do
566    now work correctly with my updated install scripts for yacc and lex).
567    
568    ----------------------------------------------------------------------
569    Name: Matthias Blume
570    Date: 2000/08/08 12:33:00 JST
571    Tag: blume-20000808-manual
572    Description:
573    
574    Tiny update to CM manual.
575    
576    ----------------------------------------------------------------------
577    Name: Allen Leung
578    Date: 2000/08/7 19:31:00
579    Tag: leunga-20000807-a-whole-bunch-of-stuff
580    
581      Moby, C--, SSA, x86, machine descriptions etc.  Should only affect C--
582    and Mobdy.
583    
584    1.  x86
585    
586       a.  Fixes to peephole module by John and Dan.
587       b.  Assembly fix to SETcc by Allen.
588       c.  Fix to c-call by John.
589       d.  Fix to spilling by John.  (This one deals with the missing FSTPT case)
590       e.  Instruction selection optimization to SETcc as suggested by John.
591    
592           For example,
593    
594            MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
595    
596           should generate:
597    
598            MOVL a, x
599            SUBL b, x
600            SHRL 31, x
601    
602    2.  IR stuff
603    
604         A bunch of new DJ-graph related algorithms added.  These
605         speed up SSA construction.
606    
607    3.  SSA + Scheduling
608    
609         Added code for SSA and scheduling to the repository
610    
611    ----------------------------------------------------------------------
612    Name: Lal George
613    Date: 2000/07/27 11:53:14 EDT
614    
615    Tag: lal-20000727-linux-ppc
616    Description:
617    
618     Made changes to support Linux PPC.
619     p.s. I have confirmation that the 110.29 boot files work fine.
620    
621    ----------------------------------------------------------------------
622    Name: Matthias Blume
623    Date: 2000/07/27 17:40:00 JST
624    Tag: blume-20000727-scripts
625    Description:
626    
627    !!!! WARNING !!!!
628    You must recompile the runtime system!
629    !!!! WARNING !!!!
630    
631    This is basically another round of script-enhancements:
632    
633    1. sml, ml-build, and ml-makedepend accept options -D and -U to define
634       and undefine CM preprocessor symbols.
635    
636    2. ml-build avoids generating a new heap image if it finds that the
637       existing one is still ok.  (The condition is that no ML file had to
638       be recompiled and all ML files are found to be older that the heap
639       file.)
640    
641       To make this work smoothly, I also hacked the runtime system as
642       well as SMLofNJ.SysInfo to get access to the heap image suffix
643       (.sparc-solaris, ...) that is currently being used.
644    
645       Moreover, the signature of CM.mk_standalone has changed.  See the
646       CM manual.
647    
648    3. ml-makedepend accepts additional options -n, -a, and -o.  (See the
649       CM manual for details.)
650    
651    4. More CM manual updates:
652        - all of the above has been documented.
653        - there is now a section describing the (CM-related) command line
654          arguments that are accepted by the "sml" command
655    
656    ----------------------------------------------------------------------
657    Name: Matthias Blume
658    Date: 2000/07/25 16:20:00 JST
659    Tag: blume-20000725-makedepend
660    Description:
661    
662    Added a script called ml-makedepend.  This can be used in makefiles
663    for Unix' make in a way very similar to the "makedepend" command for
664    C.
665    
666    The script internally uses function CM.sources.
667    
668    Synopsis:
669    
670        ml-makedepend [-f makefile] cmfile targetname
671    
672    The default for the makefile is "makefile" (or "Makefile" should
673    "makefile" not exist).
674    
675    ml-makedepend adds a cmfile/targetname-specific section to this
676    makefile (after removing the previous version of this section).  The
677    section contains a single dependency specification with targetname on
678    the LHS (targetname is an arbitrary name), and a list of files derived
679    from the cmfile on the RHS.  Some of the files on the RHS are
680    ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
681    "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
682    path names.  The makefile writer is responsible for making sure that
683    these variables have correct at the time "make" is invoked.
684    
685    ----------------------------------------------------------------------
686    Name: Matthias Blume
687    Date: 2000/07/22 23:30:00 JST
688    Tag: blume-20000722-urlupdate
689    Description:
690    
691    Changed BOOT and config/srcarchiveurl to point to BL server:
692    
693        ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
694    
695    ----------------------------------------------------------------------
696    Name: Matthias Blume
697    Date: 2000/07/18 18:00:00 JST
698    Tag: blume-20000718-Version_110_29
699    Description:
700    
701    1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
702    
703    2. Updated config/version to 110.29
704    
705    3. Updated config/srcarchiveurl
706    
707    3. New boot files!
708       ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
709    
710    ----------------------------------------------------------------------
711    Name: Matthias Blume
712    Date: 2000/07/11 13:58:00 JST
713    Tag: blume-20000711-doctypo
714    Description:
715    
716    Fixed a few typos in CM manual.
717    
718    ----------------------------------------------------------------------
719    Name: Allen Leung
720    Date: 2000/06/15 00:38:00
721    Tag: leunga-20000704-sparc-x86
722    
723    1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
724    2. fix to x86 RET bug [found by Dan Grossman]
725    3. sparc assembly bug fix for ticc instructions [found by Fermin]
726    
727       Affects c-- and moby only
728    
729    ----------------------------------------------------------------------
730    Name: Matthias Blume
731    Date: 2000/07/04 15:26:00
732    Tag: blume-20000704-trigger
733    Description:
734    
735    1. Improvements to CM manual.
736    2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
737       of getting a back-trace.  The function, when called, raises an
738       internal exception which explicitly carries the full back-trace history,
739       so it is unaffected by any intervening handle-raise pairs ("trivial"
740       or not).  The interactive loop will print that history once it arrives
741       at top level.
742       Short of having all exceptions implicitly carry the full history, the
743       recommended way of using this facility is:
744         - compile your program with instrumentation "on"
745         - run it, when it raises an exception, look at the history
746         - if the history is "cut off" because of some handler, go and modify
747           your program so that it explicitly calls BTrace.trigger
748         - recompile (still instrumented), and rerun; look at the full history
749    
750    ----------------------------------------------------------------------
751    Name: Matthias Blume
752    Date: 2000/07/03 15:36:00 JST
753    Tag: blume-20000702-manual
754    Description:
755    
756    Small corrections and updates to CM manual.
757    
758    ----------------------------------------------------------------------
759    Name: Matthias Blume
760    Date: 2000/06/29 16:04:00 JST
761    Tag: blume-20000629-yacctool
762    Description:
763    
764    Changes:
765    
766    1. Class "mlyacc" now takes separate arguments to pass options to
767       generated .sml- and .sig-files independently.
768    2. Corresponding CM manual updates.
769    3. BTrace module now also reports call sites.  (However, for loop clusters
770       it only shows from where the cluster was entered.)  There are associated
771       modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
772    
773    ----------------------------------------------------------------------
774    Name: Matthias Blume
775    Date: 2000/06/27 16:51:00 JST
776    Tag: blume-20000627-noweb
777    Description:
778    
779    Changes:
780    
781     1. Implemented "subdir" and "witness" options for noweb tool.
782        This caused some slight internal changes in CM's tool implementation.
783     2. Fixed bug in "tool plugin" mechanism.  This is essentially cleaning
784        some remaining issues from earlier path anchor changes.
785     3. Updated CM manual accordingly.
786    
787     4. Changed implementation of back-tracing so that I now consider it
788        ready for prime-time.
789    
790        In particular, you don't have to explicitly trigger the back-trace
791        anymore.  Instead, if you are running BTrace-instrumented code and
792        there is an uncaught exception (regardless of whether or not it was
793        raised in instrumented code), the top-level evalloop will print
794        the back-trace.
795    
796        Features:
797    
798          - Instrumented and uninstrumented code work together seemlessly.
799            (Of course, uninstrumented code is never mentioned in actual
800             back-traces.)
801    
802          - Asymptotic time- and space-complexity of instrumented code is
803            equal to that of uninstrumented code.  (This means that
804            tail-recursion is preserved by the instrumentation phase.)
805    
806          - Modules whose code has been instrumented in different sessions
807            work together without problem.
808    
809          - There is no penalty whatsoever on uninstrumented code.
810    
811          - There is no penalty on "raise" expressions, even in
812            instrumented code.
813    
814        A potential bug (or perhaps it is a feature, too):
815    
816          A back-trace reaches no further than the outermost instrumented
817          non-trivial "raise".  Here, a "trivial" raise is one that is the
818          sole RHS of a "handle" rule.  Thus, back-traces reach trough
819    
820               <exp> handle e => raise e
821    
822          and even
823    
824               <exp> handle Foo => raise Bar
825    
826          and, of course, through
827    
828               <exp> handle Foo => ...
829    
830         if the exception was not Foo.
831    
832         Back-traces always reach right through any un-instrumented code
833         including any of its "handle" expressions, trivial or not.
834    
835       To try this out, do the following:
836    
837         - Erase all existing binfiles for your program.
838           (You may keep binfiles for those modules where you think you
839            definitely don't need back-tracing.)
840         - Turn on back-trace instrumentation:
841              SMLofNJ.Internals.BTrace.mode (SOME true);
842         - Recompile your program.  (I.e., run "CM.make" or "use".)
843         - You may now turn instrumentation off again (if you want):
844              SMLofNJ.Internals.BTrace.mode (SOME false);
845         - Run your program as usual.  If it raises an exception that
846           reaches the interactive toplevel, then a back-trace will
847           automatically be printed.  After that, the toplevel loop
848           will print the exception history as usual.
849    
850    ----------------------------------------------------------------------
851    Name: Matthias Blume
852    Date: 2000/06/26 09:56:46 JST
853    Tag: blume-20000626-setup
854    Description:
855    
856    CM: - setup-parameter to "sml" added; this can be used to run arbitrary
857          ML code before and after compiling a file (e.g., to set compiler
858          flags)
859    
860    Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
861              - associated changes to btrace.sml (BTrace instrumentation pass)
862              - cleaner implementation of btimp.sml (BTrace tracing and report
863                module)
864    
865    CM manual: * new path encoding documented
866               * description of setup-parameter to "sml" added
867    
868    The biggest user-visible change to back-tracing is that it is no
869    longer necessary to compile all traced modules within the same
870    session.  (This was a real limitation.)
871    
872    ----------------------------------------------------------------------
873    Name: Matthias Blume
874    Date: 2000/06/24 12:40:00 JST
875    Tag: blume-20000624-startup
876    Description:
877    
878    Fixes startup slowdown problem.  (I was calling SrcPath.sync a _tad_
879    bit too often -- to put it mildly. :)
880    
881    ----------------------------------------------------------------------
882    Name: Matthias Blume
883    Date: 2000/06/23 18:20:00 JST
884    Tag: blume-20000623-btrace
885    Description:
886    
887    This updates adds a backtrace facility to aid programmers in debugging
888    their programs.  This involves the following changes:
889    
890    1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
891       keeping track of the current call stack.  When programs are compiled
892       in a special mode, the compiler will insert calls to these hooks
893       into the user program.
894       "Hook" means that it is possible for different implementations of
895       back-tracing to register themselves (at different times).
896    
897    2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
898       as an Absyn.dec->Absyn.dec rewrite.  Normally this phase is turned off.
899       It can be turned on using this call:
900         SMLofNJ.Internals.BTrace.mode (SOME true);
901       Turning it off again:
902         SMLofNJ.Internals.BTrace.mode (SOME false);
903       Querying the current status:
904         SMLofNJ.Internals.BTrace.mode NONE;
905       Annotated programs are about twice as big as normal ones, and they
906       run a factor of 2 to 4 slower with a dummy back-trace plugin (one
907       where all hooks do nothing).  The slowdown with a plugin that is
908       actually useful (such as the one supplied by default) is even greater,
909       but in the case of the default plugin it is still only an constant
910       factor (amortized).
911    
912    3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
913       with a sub-structure BTrace for controlling back-tracing.  In particular,
914       the above-mentioned function "mode" controls whether the annotation
915       phase is invoked by the compiler.  Another important function is
916       "trigger": when called it aborts the current execution and causes
917       the top-level loop to print a full back-trace.
918    
919    4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
920       for back-tracing.  It keeps track of the dynamic call stack and in
921       addition to that it keeps a partial history at each "level" of that
922       stack.  For example, if a tail-calls b, b tail-calls c, and c tail-calls
923       d and b (at separate times, dynamically), then the report will show:
924    
925       GOTO   d
926             /c
927       GOTO  \b
928       CALL   a
929    
930       This shows that there was an initial non-tail call of a, then a
931       tail-call to b or c, looping behavior in a cluster of functions that
932       consist of b and c, and then a goto from that cluster (i.e., either from
933       b or from c) to d.
934    
935       Note that (depending on the user program) the amount of information
936       that the back-trace module has to keep track of at each level is bounded
937       by a constant.  Thus, the whole implementation has the same asymptotical
938       complexity as the original program (both in space and in time).
939    
940    5. compiler/TopLevel/interact/evalloop.sml has been modified to
941       handle the special exception SMLofNJ.Internals.BTrace.BTrace
942       which is raised by the "trigger" function mentioned above.
943    
944    Notes on usage:
945    
946    - Annotated code works well together with unannotated code:
947    Unannotated calls simply do not show up at all in the backtrace.
948    
949    - It is not a good idea to let modules that were annotated during
950    different sessions run at the same time.  This is because the compiler
951    chooses small integers to identify individual functions, and there
952    will be clashes if different modules were compiled in separate sessions.
953    (Nothing will crash, and you will even be told about the clashes, but
954    back-trace information will in general not be useful.)
955    
956    - Back-tracing can be confused by callcc and capture.
957    
958    - The only way of getting a back-trace right now is to explicitly
959    invoke the "trigger" function from your user program.  Eventually, we
960    should make every exception carry back-trace information (if
961    available).  But since this creates more overhead at "raise"-time
962    (similar to the current exnHistory overhead), I have not yet
963    implemented this.  (The implementation will be rather easy.)  With
964    exceptions carrying back-trace information, this facility will be even
965    more useful because users don't need to modify their programs...
966    
967    - While it is possible to compile the compiler with back-trace
968    annotations turned on (I did it to get some confidence in
969    correctness), you must make absolutely sure that core.sml and
970    btimp.sml are compiled WITHOUT annotation!  (core.sml cannot actually
971    be compiled with annotation because there is no core access yet, but
972    if you compile btimp.sml with annotation, then the system will go into
973    an infinite recursion and crash.)
974    Since CM currently does not know about BTrace, the only way to turn
975    annotations on and off for different modules of the compiler is to
976    interrupt CMB.make, change the settings, and re-invoke it.  Of course,
977    this is awkward and clumsy.
978    
979    Sample sessions:
980    
981    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
982    - SMLofNJ.Internals.BTrace.mode (SOME true);
983    [autoloading]
984    [autoloading done]
985    val it = false : bool
986    - structure X = struct
987    -     fun main n = let
988    -         fun a (x, 0) = d x
989    -           | a (x, n) = b (x, n - 1)
990    -         and b (x, n) = c (x, n)
991    -         and c (x, n) = a (x, n)
992    -         and d x = e (x, 3)
993    -         and e (x, 0) = f x
994    -           | e (x, n) = e (x, n - 1)
995    -         and f 0 = SMLofNJ.Internals.BTrace.trigger ()
996    -           | f n = n * g (n - 1)
997    -         and g n = a (n, 3)
998    -     in
999    -         f n
1000    -     end
1001    - end;
1002    structure X : sig val main : int -> int end
1003    - X.main 3;
1004    *** BACK-TRACE ***
1005    GOTO   stdIn:4.2-13.20: X.main[2].f
1006    GOTO-( stdIn:4.2-13.20: X.main[2].e
1007    GOTO   stdIn:4.2-13.20: X.main[2].d
1008         / stdIn:4.2-13.20: X.main[2].a
1009         | stdIn:4.2-13.20: X.main[2].b
1010    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1011    CALL   stdIn:4.2-13.20: X.main[2].g
1012    GOTO   stdIn:4.2-13.20: X.main[2].f
1013    GOTO-( stdIn:4.2-13.20: X.main[2].e
1014    GOTO   stdIn:4.2-13.20: X.main[2].d
1015         / stdIn:4.2-13.20: X.main[2].a
1016         | stdIn:4.2-13.20: X.main[2].b
1017    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1018    CALL   stdIn:4.2-13.20: X.main[2].g
1019    GOTO   stdIn:4.2-13.20: X.main[2].f
1020    GOTO-( stdIn:4.2-13.20: X.main[2].e
1021    GOTO   stdIn:4.2-13.20: X.main[2].d
1022         / stdIn:4.2-13.20: X.main[2].a
1023         | stdIn:4.2-13.20: X.main[2].b
1024    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1025    CALL   stdIn:4.2-13.20: X.main[2].g
1026    GOTO   stdIn:4.2-13.20: X.main[2].f
1027    CALL   stdIn:2.15-17.4: X.main[2]
1028    -
1029    
1030    (Note that because of a FLINt bug the above code currently does not
1031    compile without BTrace turned on.)
1032    
1033    Here is another example, using my modified Tiger compiler:
1034    
1035    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
1036    - SMLofNJ.Internals.BTrace.mode (SOME true);
1037    [autoloading]
1038    [autoloading done]
1039    val it = false : bool
1040    - CM.make "sources.cm";
1041    [autoloading]
1042    ...
1043    [autoloading done]
1044    [scanning sources.cm]
1045    [parsing (sources.cm):parse.sml]
1046    [creating directory CM/SKEL ...]
1047    [parsing (sources.cm):tiger.lex.sml]
1048    ...
1049    [wrote CM/sparc-unix/semant.sml]
1050    [compiling (sources.cm):main.sml]
1051    [wrote CM/sparc-unix/main.sml]
1052    [New bindings added.]
1053    val it = true : bool
1054    - Main.compile ("../testcases/merge.tig", "foo.out");
1055    *** BACK-TRACE ***
1056    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
1057    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1058    CALL   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1059    GOTO   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1060    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1061    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1062    CALL   lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
1063         / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
1064    CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
1065    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1066    CALL   lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
1067    CALL   main.sml:1.18-118.4: Main.compile[2]
1068    -
1069    
1070    ----------------------------------------------------------------------
1071    Name: Matthias Blumen
1072    Date: 2000/06/21 18:00:00 JST
1073    Tag: blume-20000621-manual
1074    Description:
1075    
1076    CM manual update: Path environments documented.
1077    
1078    ----------------------------------------------------------------------
1079    Name: Matthias Blume
1080    Date: 2000/06/19 13:40:00
1081    Tag: blume-20000619-manual
1082    Description:
1083    
1084    CM manual and system/README update.  This only covers the fact that
1085    there are no more implicit anchors.  (Path environments and the "bind"
1086    option to "cm" have yet to be documented.)
1087    
1088    ----------------------------------------------------------------------
1089    Name: Matthias Blume
1090    Date: 2000/06/19 11:05:00 JST
1091    Tag: blume-20000619-chdir-bugfix
1092    Description:
1093    
1094    Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
1095    
1096    ----------------------------------------------------------------------
1097    Name: Matthias Blume
1098    Date: 2000/06/18 22:00:10 JST
1099    Tag: blume-20000618-implicit-anchors-really-gone
1100    Description:
1101    
1102    I updates the previous HISTORY entry where I forgot to mention that
1103    implicit anchors are no longer with us.
1104    
1105    The current update also gets rid of the (now useless) controller
1106    CM.Control.implicit_anchors.
1107    
1108    ----------------------------------------------------------------------
1109    Name: Matthias Blume
1110    Date: 2000/06/16 17:30:00 JST
1111    Tag: blume-20000616-anchorenv
1112    Description:
1113    
1114    This patch implements the long anticipated (just kidding :) "anchor
1115    environment" mechanism.  In the course of doing this, I also
1116    re-implemented CM's internal "SrcPath" module from scratch.  The new
1117    one should be more robust in certain boundary cases.  In any case, it
1118    is a lot cleaner than its predecessor (IMHO).
1119    
1120    This time, although there is yet another boot file format change, I
1121    kept the unpickler backward-compatible.  As a result, no new bootfiles
1122    are necessary and bootstrapping is straightforward.  (You cannot read
1123    new bootfiles into an old system, but the other way around is no
1124    problem.)
1125    
1126    Visible changes:
1127    
1128    ** 0. Implicit path anchors (without the leading $-symbol) are no
1129    longer recognized at all. This means that such path names are not
1130    illegal either.  For example, the name basis.cm simply refers to a
1131    local file called "basis.cm" (i.e, the name is an ordinary path
1132    relative to .cm-files directory).  Or, to put it differently, only
1133    names that start with $ are anchored paths.
1134    
1135    ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
1136    vanished.
1137    
1138    John (Reppy) had critizised this as soon as I originally proposed and
1139    implemented it, but at that time I did not really deeply believe
1140    him. :) Now I came full-circle because I need the $<singlearc> syntax
1141    in another place where it cannot be seen as an abbreviation for
1142    $/<singlearc>.  To avoid the confusion, $<singlearc> now means what it
1143    seems to mean (i.e., it "expands" into the corresponding anchor
1144    value).
1145    
1146    However, when paths are used as members in CM description files, it
1147    continues to be true that there must be at least another arc after the
1148    anchor.  This is now enforced separately during semantic analysis
1149    (i.e., from a lexical/syntactical point of view, the notation is ok.)
1150    
1151    ** 2. The "cm" class now accepts an option "bind".  The option's value
1152    is a sub-option list of precisely two items -- one labeled "anchor"
1153    and the other one labeled "value".  As you might expect, "anchor" is
1154    used to specify an anchor name to be bound, and "value" specifies what
1155    the anchor is being bound to.
1156    
1157    The value must be a directory name and can be given in either standard
1158    syntax (including the possibility that it is itself an anchored path)
1159    or native syntax.
1160    
1161    Examples:
1162    
1163       foo.cm (bind:(anchor:bar value:$mystuff/bar))
1164       lib.cm (bind:(anchor:a value:"H:\\x\\y\\z"))  (* only works under windows *)
1165    
1166    and so on.
1167    
1168    The meaning of this is that the .cm-file will be processed with an
1169    augmented anchor environment where the given anchor(s) is/are bound to
1170    the given values(s).
1171    
1172    The rationale for having this feature is this: Suppose you are trying
1173    to use two different (already stable) libraries a.cm and b.cm (that
1174    you perhaps didn't write yourself).  Further, suppose each of these
1175    two libraries internally uses its own auxiliary library $aux/lib.cm.
1176    Normally you would now have a problem because the anchor "lib" can not
1177    be bound to more than one value globally.  Therefore, the project that
1178    uses both a.cm and b.cm must locally redirect the anchor to some other
1179    place:
1180    
1181       a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
1182       b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
1183    
1184    This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
1185    /usr/lib/smlnj/b-stuff/aux.cm, respectively.
1186    
1187    Hard-wiring path names is a bit inflexible (and CM will verbosely warn
1188    you when you do so at the time of CM.stabilize).  Therefore, you can
1189    also use an anchored path as the value:
1190    
1191      a.cm (bind:(anchor:lib value:$a-lib))
1192      b.cm (bind:(anchor:lib value:$b-lib))
1193    
1194    Now you can globally configure (using the usual CM.Anchor.anchor or
1195    pathconfig machinery) bindings for "a-lib" and "b-lib".  Since "lib"
1196    itself is always locally bound, setting it globally is no longer
1197    meaningful or necessary (but it does not hurt either).  In fact, "lib"
1198    can still be used as a global anchor for separate purposes.  As a
1199    matter of fact, one can locally define "lib" in terms of a global
1200    "lib":
1201    
1202      a.cm (bind:(anchor:lib value:$lib/a))
1203      b.cm (bind:(anchor:lib value:$lib/b))
1204    
1205    ** 3: The encoding of path names has changed.  This affects the way
1206    path names are shown in CM's progress report and also the internal
1207    protocol encoding used for parallel make.
1208    
1209    The encoding now uses one or more ':'-separated segments.  Each
1210    segments corresponds to a file that has been specified relative to the
1211    file given by its preceding segment.  The first segment is either
1212    relative to the CWD, absolute, or anchored.  Each segment itself is
1213    basically a Unix pathname; all segments but the first are relative.
1214    
1215    Example:
1216    
1217       $foo/bar/baz.cm:a/b/c.sml
1218    
1219    This path denotes the file bar/a/b/c.sml relative to the directory
1220    denoted by anchor "foo".  Notice that the encoding also includes
1221    baz.cm which is the .cm-file that listed a/b/c.sml.  As usual, such
1222    paths are resolved relative to the .cm-files directory, so baz.cm must
1223    be ignored to get the "real" pathname.
1224    
1225    To make this fact more obvious, CM puts the names of such "virtual
1226    arcs" into parentheses when they appear in progress reports. (No
1227    parentheses will appear in the internal protocol encoding.)  Thus,
1228    what you really see is:
1229    
1230      $foo/bar/(baz.cm):a/b/c.sml
1231    
1232    I find this notation to be much more informative than before.
1233    
1234    Another new feature of the encoding is that special characters
1235    including parentheses, colons, (back)slashes, and white space are
1236    written as \ddd (where ddd is the decimal encoding of the character).
1237    
1238    *** The CM manual still needs to be updated.
1239    
1240    ----------------------------------------------------------------------
1241    Name: Allen Leung
1242    Date: 2000/06/15 00:38:00
1243    Tag: leunga-20000615-x86-peephole
1244    
1245    x86 Peephole fix by Fermin.  Affects c-- and moby only.
1246    
1247    ----------------------------------------------------------------------
1248    Name: Matthias Blume
1249    Date: 2000/06/12 11:40:00
1250    Tag: blume-20000612-parmakefix
1251    Description:
1252    
1253    More cleanup after changing the file naming scheme: This time I
1254    repaired the parallel make mechanism for CMB.make which I broke earlier.
1255    
1256    ----------------------------------------------------------------------
1257    Name: Allen Leung
1258    Date: 2000/06/09 01:25:00
1259    Tag: leunga-20000609-various
1260    
1261    None of these things should affect normal SML/NJ operations
1262    
1263    1. Peephole improvements provided by Fermin (c--)
1264    2. New annotation DEFUSE for adding extra dependence (moby)
1265    3. New X86 LOCK instructions (moby)
1266    4. New machine description language for reservation tables (scheduling)
1267    5. Fixes to various optimization/analysis modules (branch chaining, dominator
1268       trees etc.)
1269    6. I've changed the CM files so that they can work with versions
1270       110.0.6, 110.25 and 110.28
1271    
1272    ----------------------------------------------------------------------
1273    Name: Matthias Blume
1274    Date: 2000/06/09 12:40:00
1275    Tag: blume-20000609-log
1276    Description:
1277    
1278    - Removed all(?) remaining RCS Log entries from sources.
1279    
1280    - Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
1281      anchored paths).
1282    
1283    ----------------------------------------------------------------------
1284    Name: Matthias Blume
1285    Date: 2000/06/07 17:00:00 JST
1286    Tag: blume-20000607-no-implicit-anchors
1287    Description:
1288    
1289    1. This update changes the default setting for
1290    CM.Control.implicit_anchors from true to false.  This means that
1291    implicit anchors are no longer permitted by default.  I also tried to
1292    make sure that nothing else still relies on implicit anchors.
1293    (This is the next step on the schedule towards a CM that does not even
1294    have the notion of implicit anchors anymore.)
1295    
1296    2. More CM manual updates.
1297    
1298    3. I managed to track down and fix the pickling bug I mentioned last
1299    time.  Because of the previously existing workaround, this entails no
1300    immediate practical changes.
1301    
1302    ----------------------------------------------------------------------
1303    Name: Matthias Blume
1304    Date: 2000/06/06 11:15:00 JST
1305    Tag: blume-20000606-lazierpickle
1306    Description:
1307    
1308    !!!! NEW BOOT FILES !!!!
1309    
1310    * The main purpose of this update is to make library pickles lazier in
1311    order to reduce the initial space penalty for autoloading a library.
1312    As a result, it is now possible to have $smlnj/compiler.cm
1313    pre-registered.  This should take care of the many complaints or
1314    inquiries about missing structure Compiler.  This required changes to
1315    CM's internal data structures and small tweaks to some algorithms.
1316    
1317    As a neat additional effect, it is no longer necessary (for the sake
1318    of lean heap image files) to distinguish between a "minimal" CM and a
1319    "full" CM.  Now, there is only one CM (i.e., the "full" version:
1320    $smlnj/cm.cm aka $smlnj/cm/full.cm), and it is always available at the
1321    interactive top level. ($smlnj/cm/minimal.cm is gone.)
1322    
1323    To make the life of compiler-hackers easier, "makeml" now also
1324    pre-registers $smlnj/cmb.cm (aka $smlnj/cmb/current.cm).  In other
1325    words, after you bootstrap a new sml for the first time, you will not
1326    have to autoload $smlnj/cmb.cm again afterwards.  (The first time
1327    around you will still have to do it, though.)
1328    
1329    * A second change consists of major updates to the CM manual.  There
1330    are now several appendices with summary information and also a full
1331    specification of the CM description file syntax.
1332    
1333    * In directory src/system I added the script "allcross".  This script
1334    invokes sml and cross-compiles the compiler for all supported
1335    architectures.  (Useful when providing a new set of boot files.)
1336    
1337    * There seems to be a latent bug in my "lazy pickles" mechanism.  I
1338    added a small tweak to pickle-util.sml to work around this problem,
1339    but it is not a proper fix yet.  I will investigate further.  (The
1340    effect of the bug was an inflation of library pickle size.)
1341    
1342    * Version number increased to 110.28.1 (to avoid compatibility problems).
1343    
1344    ----------------------------------------------------------------------
1345    Name: Allen Leung
1346    Date: 2000/05/25 17:28 EDT
1347    Tag: leunga-20000525-ra
1348    Description:
1349    
1350      Fixed a bug in freezing phase of the register allocator.
1351    
1352    ----------------------------------------------------------------------
1353    Name: Allen Leung
1354    Date: 2000/05/15 22:53 EDT
1355    Tag: leunga-20000515-alpha-x86-ra
1356    Description:
1357    
1358      1. Alpha
1359    
1360          Slight cleanup.  Removed the instruction SGNXL
1361    
1362      2. X86
1363    
1364          Added the following instructions to the instruction set:
1365    
1366            ROLx, RORx,
1367            BTx, BTSx, BTLx, BTRx,
1368            XCHGx, and variants with the LOCK prefix
1369    
1370      3. Register Allocation
1371    
1372          The module ra-rewrite-with-renaming has been improved.
1373    
1374      These have no effect on SML/NJ.
1375    
1376    ----------------------------------------------------------------------
1377    Name: Matthias Blume
1378    Date: 2000/05/15 16:20:00 JST
1379    Tag: blume-20000515-lightrebuild
1380    Description:
1381    
1382    1. I added an alternative to "-rebuild" to "makeml".  The difference is
1383       that prior to calling CMB.make' the CM-variable "LIGHT" will be
1384       defined.  In effect, the command will not build any cross-compiler
1385       backends and therefore finish more quickly.
1386    
1387       The "fixpt" script also takes a "-light" switch to be able to use
1388       this new facility while compiling for a fixpoint.
1389    
1390    2. I replaced all mentions of anchored paths in group owner specifications
1391       with simple relative paths (usually starting with "..").
1392       The rationale is that a library's internal workings should not be
1393       compromised by the lack of some anchor.  (An anchor is necessary
1394       for someone who wants to refer to the library by an anchored path,
1395       but it should not be necessary to build the same library in the first
1396       place.)
1397    
1398    3. I changed the way CM's tool mechanism determines the shell command
1399       string used for things like ml-yacc etc. so that it does not break
1400       when CM.Control.implicit_anchors is turned off.
1401    
1402    ----------------------------------------------------------------------
1403    Name: Matthias Blume
1404    Date: 2000/05/12 18:20:00 JST
1405    Tag: blume-20000512-ml-build
1406    Description:
1407    
1408    Fixed a bug in config/_ml-build that prevented ml-yacc and ml-lex from
1409    getting installed properly (by config/install.sh).
1410    
1411    ----------------------------------------------------------------------
1412    Name: Matthias Blume
1413    Date: 2000/05/12 17:30:00 JST
1414    Tag: blume-20000512-anchors
1415    Description:
1416    
1417    !!! NEW BOOT FILES !!!
1418    
1419    This change is in preparation of fading out support for "implicitly
1420    anchored path names".  I went through all sources and used the
1421    explicit (and relatively new) $-notation.  See system/README and the
1422    CM manual for more info on this.
1423    
1424    I also modified the anchoring scheme for some things such as "smlnj",
1425    "MLRISC", "cm", etc. to take advantage of the fact that explicit
1426    anchors are more expressive: anchor name and first arc do not have to
1427    coincide.  This entails the following user-visible change:
1428    
1429    You have to write $smlnj/foo/bar instead of smlnj/foo/bar.  In
1430    particular, when you fire up sml with a command-line argument, say,
1431    e.g.:
1432    
1433       sml '$smlnj/cmb.cm'
1434    
1435    At the ML toplevel prompt:
1436    
1437       CM.autoload "$smlnj/cmb.cm";
1438    
1439    There is also a new controller in CM.Control that can be used to turn
1440    off all remaining support for implicit anchors by saying:
1441    
1442        CM.autoload "$smlnj/
1443        #set CM.Control.implicit_anchors false;
1444    
1445    This causes CM to reject implicitly anchored paths.  This is (for the
1446    time being) less permissive than the "final" version where there will
1447    be no more such implicit anchors and relative paths will be just that:
1448    relative.
1449    
1450    The next step (version after next version?) will be to make the
1451    default for CM.Control.implicit_anchors false.  After the dust has
1452    settled, I can then produce the "final" version of this...
1453    
1454    Note: Since bootstrapping is a bit tricky, I provided new boot files.
1455    
1456    ----------------------------------------------------------------------
1457    Name: Matthias Blume
1458    Date: 2000/05/11 16:30:00 JST
1459    Tag: blume-20000511-sources
1460    Description:
1461    
1462    The main change is that I added function CM.sources as a generalized
1463    version of the earlier CM.makedepend.  This entails the following
1464    additional changes:
1465    
1466      - CM.makedepend has been dropped.
1467    
1468      - CM manual has been updated.
1469    
1470      - TOOLS signature and API have been changed.
1471    
1472    ----------------------------------------------------------------------
1473    Name: Allen Leung
1474    Date: 2000/05/10 21:17 EDT
1475    Tag: leunga-20000510-moby-c--ssa
1476    Description:
1477    
1478      Various bug fixes and new features for C--, Moby and MLRISC optimizations.
1479    None of these affect SML/NJ.
1480    
1481    1. Register Allocation
1482    
1483        a. A new ra spilling module (ra/ra-spill-with-renaming) is implemented.
1484           This module tries to remove local (i.e. basic block level) redundancies
1485           during spilling.
1486    
1487        b. A new framework for performing region based register allocation.
1488           Not yet entirely functional.
1489    
1490    2. X86
1491    
1492       a. DefUse for POP was missing the stack pointer [found by Lal]
1493       b. Reload for CALL was incorrect in X86Spill [found by John]
1494       c. Various fixes in X86Spill so that it can be used correctly for
1495          the new spilling module.
1496    
1497    3. SSA/IR
1498    
1499       a. New module ir/dj-dataflow.sml implements elimination based
1500          data flow analysis.
1501    
1502    4. MLRiscGen
1503    
1504       a. Fix for gc type annotation
1505    
1506    5. MDGen
1507    
1508       Various fixes for machine description -> ml code translation.  For ssa
1509       only.
1510    
1511    ----------------------------------------------------------------------
1512    Name: Allen Leung
1513    Date: 2000/05/08 22:17 EDT
1514    Tag: leunga-20000508-labexp
1515    Description:
1516    
1517      Fermin has found a few assembly problems with constant expressions
1518      generated in LabelExp.  Mostly, the problems involve extra parentheses,
1519      which choke on dumb assemblers.  This is his fix.
1520    
1521    ----------------------------------------------------------------------
1522    Name: Dave MacQueen
1523    Date: 2000/04/09 14:00 EDT
1524    Tag: dbm-20000502-Version_110_28
1525    Description:
1526    
1527    1. Updated src/compiler/TopLevel/main/version.sml to version 110.28
1528    
1529    2. Updated config/version to 110.28
1530    
1531    3. Updated config/srcarchiveurl
1532    
1533    3. New boot files!
1534       ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.28/
1535    
1536    ----------------------------------------------------------------------
1537    Name: Matthias Blume
1538    Date: 2000/05/01 19:05:00 JST
1539    Tag: blume-20000501-noweb
1540    Description:
1541    
1542    A new noweb tool has been added.  The existing system is entirely
1543    unaffected by this, but some CM users have asked for renewed noweb
1544    support.  Everything is documented in the CM manual.
1545    
1546    New (plugin) libraries:
1547    
1548       noweb-tool.cm
1549       nw-ext.cm
1550    
1551    ----------------------------------------------------------------------
1552    Name: Dave MacQueen
1553    Date: 2000/04/30 12:40PM EDT
1554    Tag: dbm-20000430-bug_fixes
1555    Description:
1556    
1557    1. Fix for bug 1498
1558       smlnj/src/system/Basis/Implementation/Unsafe/object.sig
1559       smlnj/src/system/Basis/Implementation/Unsafe/object.sml
1560         added toRealArray function
1561       smlnj/src/compiler/MiscUtil/print/ppobj.sml
1562         added check for tag Obj.RealArray to array printing case in ppObj
1563    
1564    2. Fix for bug 1510
1565       smlnj/src/compiler/Semant/types/typesutil.sml
1566         fixed definition of dummyargs (used by equalTycon) so that
1567         dummy args are distinct types
1568    
1569    ----------------------------------------------------------------------
1570    Name: Matthias Blume
1571    Date: 2000/04/30 01:00:00 JST
1572    Tag: blume-20000430-versions
1573    Description:
1574    
1575    1. CM version numbering added.  This is an implementation of Lal's
1576       proposal for adding version numbers and version checking to .cm
1577       files.  Lal said that his proposal was just that -- a proposal.
1578       For the time being I went ahead and implemented it so that people
1579       can comment on it.  Everything is completely backward-compatible
1580       (except for the stable library format, i.e., new bootfiles!).
1581    
1582       As usual, see the CM manual for details.
1583    
1584    2. An alternative syntax for anchored paths has been implemented.
1585       Dave has recently voiced the same concerns that I had when I did
1586       this, so there should be some support.  My take is that eventually
1587       I will let support for the current syntax (where anchors are
1588       "implicit") fade out in favor of the new, explicit syntax.
1589       In order to be backward-compatible, both old and new syntax are
1590       currently supported.
1591    
1592       Again, see the CM manual for details.
1593    
1594    3. Parallel make is trying to be slightly smarter:  When the master
1595       process finds a "bottleneck", i.e., when there is only one
1596       compilation unit that can be compiled and everybody else is
1597       waiting on it, then it will simply compile it directly instead
1598       of clumsily telling one of the slaves to do it.
1599    
1600    4. Support for "unsharing" added.  This is necessary in order to be
1601       able to have two different versions of the same library running
1602       at the same time (e.g., for trying out a new MLRISC while still
1603       having the old MLRISC linked into the current compiler, etc.)
1604       See the CM manual.
1605    
1606    5. Simple "makedepend" functionality added for generating Makefile
1607       dependency information.  (This is rather crude at the moment.
1608       Expect some changes here in the future.)
1609    
1610    6. ".fun" added as a recognized suffix for ML files. Also documented
1611       explicitly in the manual that the fallback behavior (unknown suffix
1612       -> ML file) is not an official feature!
1613    
1614    7. Small changes to the pickler for stable libraries.
1615    
1616    8. Several internal changes to CM (for cleanup/improvement).
1617    
1618    
1619    !!!! NEW BINFILES !!!!
1620    
1621    ----------------------------------------------------------------------
1622    Name: Matthias Blume
1623    Date: 2000/04/28 17:30:00 JST
1624    Tag: blume-20000428-pathconfig
1625    Description:
1626    
1627    1. I changed config/install.sh to remove duplicate entries from the
1628       lib/pathconfig file at the end.  Moreover, the final version of
1629       lib/pathconfig is sorted alphabetically.  The same (sorting) is done
1630       in src/system/installml.
1631    
1632    2. The config/install.sh script now consistently uses relative
1633       pathnames in lib/pathconfig whenever the anchor is in the lib
1634       directory.  (So far this was true for the libraries that come
1635       pre-compiled and bundled as part of the bootfiles but not for
1636       libraries that are compiled by the script itself.)
1637    
1638    ----------------------------------------------------------------------
1639    Name: Matthias Blume
1640    Date: 2000/04/26 13:10:00 JST
1641    Tag: blume-20000426-fun_suffix
1642    Description:
1643    
1644    Added ".fun" as a recognized file name suffix (for ML code).
1645    
1646    ----------------------------------------------------------------------
1647    Name: Allen Leung
1648    Date: 2000/04/25 17:00:00 EST
1649    Tag: leunga-20000425-alpha-ra
1650    Description:
1651    
1652    1. Alpha
1653    
1654        PSEUDOARITH was missing in AlphaRewrite.  This causes an endless loop
1655    in C--.
1656    
1657    2. RA
1658    
1659       Added a flag "ra-dump-size" to print out the size of the flowgraph
1660       and the interference graph.
1661    
1662    ----------------------------------------------------------------------
1663    Name: Dave MacQueen
1664    Date: 2000/04/25/
1665    Tag: dbm-20000425-mlyacc_doc_examples
1666    Description:
1667      Updated mlyacc.tex sections 5 and 7 for SML '97 and CM.
1668      Updated all three examples in src/ml-yacc/examples to run
1669      under 110.* using CM.make.
1670    
1671    ----------------------------------------------------------------------
1672    Name: Allen Leung
1673    Date: 2000/04/20 23:04:00 EST
1674    Tag: leunga-20000420-ssa-c---stuff
1675    Description:
1676    
1677      This update synchronizes my repository with Yale's.  Most of these
1678    changes, however, do not affect SML/NJ at all (the RA is an exception).
1679    
1680    1. Register Allocator
1681    
1682       a. An improvement in the interference graph construction:
1683          Given a copy
1684    
1685                s <- t
1686    
1687          no interference edge between s and t is added for this definition of s.
1688    
1689       b. I've added two new spill heuristic modules that Fermin and I developed
1690          (in the new library RA.cm). These are unused in SML/NJ but maybe
1691          useful for others (Moby?)
1692    
1693    2. X86
1694    
1695       a. Various fixes in the backend provided by Fermin [C--] and Lal.
1696    
1697    3. Alpha
1698    
1699       a. Added the BSR instruction and code generation that goes with it [C--]
1700       b. Other fixes too numerous to recount provided by Fermin [C--]
1701    
1702    4. Regmaps
1703    
1704       a. The regmaps are not initialized with the identity physical bindings
1705          at creation time.  This is unneeded.
1706    
1707    5. MLRISC Optimizations
1708    
1709       a. The DJ-Graph module can now compute the iterated dominance frontiers
1710          intersects with liveness incrementally in linear time! Woohoo!
1711          This is now used in my new SSA construction algorithm.
1712    
1713       b. THe branch reorganization module is now smarter about linear chains of
1714          basic blocks.
1715    
1716    
1717    ----------------------------------------------------------------------
1718    Name: Matthias Blume
1719    Date: 2000/04/12 13:52:00 JST
1720    Tag: blume_main_v110p27_1
1721    Description:
1722    
1723    Changed install.sh script to handle archive files without version number
1724    and to use "boot.<arch>-<os>" instead of "sml.boot.<arch>-<os>" for the
1725    name of the boot file archive.
1726    
1727    ----------------------------------------------------------------------
1728    Name: Dave MacQueen
1729    Date: 2000/04/09 14:00 EDT
1730    Tag: dbm-20000410-Version_110_27
1731    Description:
1732    
1733    1. Updated src/compiler/TopLevel/main/version.sml to version 110.27
1734    
1735    2. Updated src/config/version to 110.27
1736    
1737    3. New boot files!
1738    
1739    ----------------------------------------------------------------------
1740    Name: Allen Leung
1741    Date: 2000/04/09 19:09:00 EST
1742    Tag: leunga-20000409-misc
1743    Description:
1744    
1745    1.  Yet another fix for x86 assembly for idivl, imull, mull and friends.
1746    
1747    2.  Miscellaneous improvements to MLRISC (unused in sml/nj)
1748    
1749    ----------------------------------------------------------------------
1750    Name: Stefan
1751    Date: 2000/04/07 10:00:00 EDT
1752    Tag: monnier-20000406-branch-handling
1753    Description:
1754    
1755    Improved handling of branches (mostly those generated from
1756    polymorphic equality), removed switchoff and changed the
1757    default optimization settings (more cpsopt and less flintopt).
1758    
1759    ----------------------------------------------------------------------
1760    Name: Allen Leung
1761    Date: 2000/04/06 01:30:00 EST
1762    Tag: leunga-20000406-peephole-x86-SSA-2
1763    Description:
1764    
1765       Forgot a few files.
1766    
1767    ----------------------------------------------------------------------
1768    Name: Allen Leung
1769    Date: 2000/04/06 00:36:00 EST
1770    Tag: leunga-20000406-peephole-x86-SSA
1771    Description:
1772    
1773    1.  New Peephole code
1774    
1775    2.  Minor improvement to X86 instruction selection
1776    
1777    3.  Various fixes to SSA and machine description -> code translator
1778    
1779    ----------------------------------------------------------------------
1780    Name: Matthias Blume
1781    Date: 2000/04/05 12:30:00 JST
1782    Tag: blume_main_v110p26p2_3
1783    Description:
1784    
1785    This update just merges three minor cosmetic updates to CM's sources
1786    to get ready for the 110.27 code freeze on Friday.  No functionality
1787    has changed.
1788    
1789    ----------------------------------------------------------------------
1790    Name: Allen Leung
1791    Date: 2000/04/04 19:39:00 EST
1792    Tag: leunga-20000404-x86-asm
1793    Description:
1794    
1795    1.  Fixed a problem in X86 assembly.
1796    
1797        Things like
1798    
1799           jmp %eax
1800           jmp (%eax)
1801    
1802        should be output as
1803    
1804           jmp *%eax
1805           jmp *(%eax)
1806    
1807    2.  Assembly output
1808    
1809          Added a new flag
1810    
1811              "asm-indent-copies" (default to false)
1812    
1813          When this flag is on, parallel copies will be indented an extra level.
1814    
1815  ----------------------------------------------------------------------  ----------------------------------------------------------------------
1816  Name: Allen Leung  Name: Allen Leung
1817  Date: 2000/04/04 03:18:00 EST  Date: 2000/04/04 03:18:00 EST
# Line 31  Line 1832 
1832    
1833  3.  Assembly  3.  Assembly
1834    
1835      When generating assemby, resolve the value of client defined constants,      When generating assembly, resolve the value of client defined constants,
1836      instead of generating symbolic values.  This is controlled by the      instead of generating symbolic values.  This is controlled by the
1837      new flag "asm-resolve-constants", which is default to true.      new flag "asm-resolve-constants", which is default to true.
1838    
# Line 54  Line 1855 
1855    
1856  To this end, I arranged that instead of "structure Core" as "structure  To this end, I arranged that instead of "structure Core" as "structure
1857  _Core" is bound in the pervasive environment.  Core access is done via  _Core" is bound in the pervasive environment.  Core access is done via
1858  _Core (which can never be accidentially rebound because _Core is not a  _Core (which can never be accidentally rebound because _Core is not a
1859  legal surface-syntax symbol).  legal surface-syntax symbol).
1860    
1861  The current solution is much cleaner because the core environment is  The current solution is much cleaner because the core environment is
# Line 64  Line 1865 
1865  with dynamic and symbolic parts of the core environment.  with dynamic and symbolic parts of the core environment.
1866    
1867  Remaining hackery (to bind the "magic" symbol _Core) is localized in the  Remaining hackery (to bind the "magic" symbol _Core) is localized in the
1868  compilation mananger's bootstrap compiler (actually: in the "init group"  compilation manager's bootstrap compiler (actually: in the "init group"
1869  handling).  See the comments in src/system/smlnj/init/init.cmi for  handling).  See the comments in src/system/smlnj/init/init.cmi for
1870  more details.  more details.
1871    
# Line 179  Line 1980 
1980     (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
1981     used).     used).
1982    
1983     This was done to accomodate the new "make" and "shell" tools which     This was done to accommodate the new "make" and "shell" tools which
1984     facilitate fairly seemless hookup to portions of code managed using     facilitate fairly seamless hookup to portions of code managed using
1985     Makefiles or Shell scripts.     Makefiles or Shell scripts.
1986    
1987     There are no classes "shared" or "private" anymore.  Instead, the     There are no classes "shared" or "private" anymore.  Instead, the
# Line 192  Line 1993 
1993    
1994     All existing tools are described in the CM manual.     All existing tools are described in the CM manual.
1995    
1996     - Slightly better error handling.  (CM now surpresses many followup     - Slightly better error handling.  (CM now suppresses many followup
1997     error messages that tended to be more annoying than helpful.)     error messages that tended to be more annoying than helpful.)
1998    
1999  2. Major changes to the compiler's static environment data structures.  2. Major changes to the compiler's static environment data structures.
# Line 326  Line 2127 
2127    
2128       I've changed andl to testl in the floating point test sequence       I've changed andl to testl in the floating point test sequence
2129       whenever appropriate.  The Intel optimization guide states that       whenever appropriate.  The Intel optimization guide states that
2130       testl is perferable to andl.       testl is preferable to andl.
2131    
2132  7. RA (x86 only)  7. RA (x86 only)
2133    
# Line 508  Line 2309 
2309    
2310  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
2311  additional argument called "template" which is an optional string that  additional argument called "template" which is an optional string that
2312  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
2313  explanation.  explanation.
2314    
2315  2. A special-purpose tool can be "regisitered" by simply dropping the  2. A special-purpose tool can be "registered" by simply dropping the
2316  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
2317  directory where the .cm file lives that uses this tool.  (The  directory where the .cm file lives that uses this tool.  (The
2318  behavior/misfeature until now was to look for the tool description  behavior/misfeature until now was to look for the tool description
# Line 555  Line 2356 
2356  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
2357  <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,
2358  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
2359  perhaps got accidentially killed).  In this case, fixpt will quickly  perhaps got accidentally killed).  In this case, fixpt will quickly
2360  move through what exists before continuing where it left off earlier,  move through what exists before continuing where it left off earlier,
2361  and, thus, saves a lot of time.  and, thus, saves a lot of time.
2362    
# Line 605  Line 2406 
2406    it from that remote directory.    it from that remote directory.
2407    This should simplify installation further:  For machines that have    This should simplify installation further:  For machines that have
2408    access to the internet, just fetch <version>-config.tgz, unpack it,    access to the internet, just fetch <version>-config.tgz, unpack it,
2409    edit config/targets, and go (run config/install.sh).  The scipt will    edit config/targets, and go (run config/install.sh).  The script will
2410    fetch everything else that it might need all by itself.    fetch everything else that it might need all by itself.
2411    
2412    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.593  
changed lines
  Added in v.801

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