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 644, Fri May 12 09:18:31 2000 UTC revision 810, Fri Apr 20 13:26:46 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: Lal George
17    Date: 2001/04/20 09:15:28 EDT
18    Tag: george-20010420-macosX
19    Description:
20    
21     - Changes to port to Mac OS X; Darwin.
22    
23     - In the process I found that sqrt was broken on the PPC, because the
24       fsqrt instruction is not implemented.
25    
26    ----------------------------------------------------------------------
27    Name: Matthias Blume
28    Date: 2001/04/18 12:45:00 EDT
29    Tag: blume-20010418-ccalls
30    Description:
31    
32     - fixed two off-by-4 errors in the x86-specific c-calls implementation
33       (this bug prevented structure arguments containing pointers from being
34        passed correctly)
35     - changed the raw-C-call code in mlriscGen.sml in such a way that
36       structure arguments are represented as a pointer to the beginning
37       of the structure (instead of having a series of synthesized arguments,
38       one for each structure member)
39    
40     - made makeml script's verbosity level configurable via environment
41       variable (MAKEML_VERBOSITY)
42    
43     - eliminteded placeholder implementations for f32l, w16s, i16s, and f32s
44       in rawmem-x86.sml; we are now using the real thing
45    
46    ----------------------------------------------------------------------
47    Name: Matthias Blume
48    Date: 2001/03/22 16:25:00 EST
49    Tag: blume-20010322-bootfiles
50    Description:
51    
52    Created a new set of bootfiles (for your automatic installation convenience).
53    
54    ----------------------------------------------------------------------
55    Name: Matthias Blume
56    Date: 2001/03/22 15:10:00 EST
57    Tag: blume-20010322-rawmem-parcm
58    Description:
59    
60    1. All "raw memory access" primitives for the new FFI are implemented now
61       (at least on the x86).
62    2. Some further cleanup of CM's parallel make mechanism.
63    
64    ----------------------------------------------------------------------
65    Name: Matthias Blume
66    Date: 2001/03/19 17:53:00 EST
67    Tag: blume-20010319-parallel
68    Description:
69    
70    Parallel make (using compile servers) now works again.
71    
72    To this end, CM.stabilize and CMB.make have been modified to work in
73    two passes when compile servers are attached:
74       1. Compile everything, do not perform stabilization; this pass
75          uses compile servers
76       2. Stabilize everything; this pass does not use compile servers
77    If there are no compile servers, the two passes are combined into one
78    (as before).  Splitting the passes increases the inherent parallelism
79    in the dependency graph because the entire graph including all
80    libraries is available at the same time.  This, in turn, improves
81    server utilization.  The downside is that the master process will
82    have to do some extra work after compilation is done (because for
83    technical reasons it must re-read all the binfiles during stabilization).
84    
85    ----------------------------------------------------------------------
86    Name: Matthias Blume
87    Date: 2001/03/16 12:22:00 EST
88    Tag: blume-20010316-bootfiles
89    Description:
90    
91    Created a new set of bootfiles (for your automatic installation convenience).
92    
93    ----------------------------------------------------------------------
94    Name: Matthias Blume
95    Date: 2001/03/16 11:00:00 EST
96    Tag: blume-20010316-MLTREE-fixup
97    Description:
98    
99    This is a minor fixup for an (untagged) earlier commit by Allen.
100    (A file was missing).
101    
102    ----------------------------------------------------------------------
103    Name: Allen Leung
104    Date: Mon Mar  5 18:54:57 EST 2001
105    Tag: leunga-20010305-cut-support
106    
107    1. New support for alternative control-flow in MLTREE.
108       Currently we support
109    
110          FLOW_TO(CALL ...., [k1,...,kn])
111    
112       This is needed for 'cuts to' in C-- and try/handle-like constructs
113       in Moby
114    
115       New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
116    
117    2. Register Allocator
118    
119       Changes in interface [from Fermin, John]
120    
121    3. Alpha 8-bit SLL support [Fermin]
122    
123    4. All architectures
124    
125       A new module (ClusterExpandCopies) for expanding parallel copies.
126    
127    ----------------------------------------------------------------------
128    Name: Allen Leung
129    Date: 2001/02/27 23:07:00 EST
130    Tag: leunga-20010227-minor-stuff
131    
132    1. Alpha bug fix for CMOVNE
133    2. Handle mltree COND(..,FCMP ...,...)
134    3. Bug fix in simplifier
135    
136    ----------------------------------------------------------------------
137    Name: Matthias Blume
138    Date: 2001/01/30 17:50:00 EST
139    Tag: blume-20010130-sync
140    Description:
141    
142    This is just a minor update to sync my devel branch with the main brach.
143    The only visible change is the addition of some README files.
144    
145    ----------------------------------------------------------------------
146    Name: Matthias Blume
147    Date: 2001/01/12 23:30:00 JST
148    Tag: blume-20010112-bootfiles
149    Description:
150    
151    Made a new set of bootfiles that goes with the current state of the
152    repository.
153    
154    ----------------------------------------------------------------------
155    Name: Matthias Blume
156    Date: 2001/01/12 21:20:00 JST
157    Tag: blume-20010112-sync
158    Description:
159    
160    I am just flushing out some minor changes that had accumulated in
161    my private branch in order to sync with the main tree.  (This is
162    mainly because I had CVS trouble when trying to merge _into_ my
163    private branch.)
164    
165    Most people should be completely unaffected by this.
166    
167    ----------------------------------------------------------------------
168    Name: Allen Leung
169    Date: Thu Jan 11 21:03:00 EST 2001
170    Tag: leunga-20010111-labexp=mltree
171    Description:
172    
173    1.  Removed the type LabelExp and replace it by MLTree.
174    2.  Rewritten mltree-simplify with the pattern matcher tool.
175    3.  There were some bugs in alpha code generator which would break
176        64-bit code generation.
177    4.  Redo the tools to generate code with the
178    5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
179        are now generated by perl scripts.
180    
181    ----------------------------------------------------------------------
182    Name: Matthias Blume
183    Date: 2001/01/10 21:55:00 JST
184    Tag: blume-20010110-rcc
185    Description:
186    
187    The RCC stuff now seems to work (but only on the x86).
188    This required hacking of the c-calls interface (and -implementation) in
189    MLRISC.
190    
191    Normal compiler users should be unaffected.
192    
193    ----------------------------------------------------------------------
194    Name: Matthias Blume
195    Date: 2001/01/09 01:20:00 JST
196    Tag: blume-20010109-rcc
197    Description:
198    
199    This is a fairly big patch, flushing out a large number of pending
200    changes that I made to my development copy over the last couple of days.
201    
202    Of practical relevance at this moment is a workaround for a pickling
203    bug that Allen ran into the other day.  The cause of the bug itself is
204    still unknown and it might be hard to fix it properly, but the
205    workaround has some merits of its own (namely somewhat reducing pickling
206    overhead for certain libraries).  Therefore, I think this solution should
207    be satisfactory at this time.
208    
209    The rest of the changes (i.e., the vast majority) has to do with my
210    ongoing efforts of providing direct support for C function calls from
211    ML.  At the moment there is a new primop "RAW_CCALL", typing magic
212    in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
213    case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
214    translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
215    handle RCC.
216    
217    The last part (the changes to mlriscGen.sml) are still known to be
218    wrong on the x86 and not implemented on all other architectures.  But
219    the infrastructure is in place. I had to change a few functor
220    signatures in the backend to be able to route the CCalls interface
221    from MLRISC there, and I had to specialize the mltree type (on the
222    x86) to include the necessary extensions. (The extensions themselves
223    were already there and redy to go in MLRISC/x86).
224    
225    Everything should be very happy as soon as someone helps me with
226    mlriscGen.sml...
227    
228    In any case, nothing of this should matter to anyone as long as the
229    new primop is not being used (which is going to be the case unless you
230    find it where I hid it :). The rest of the compiler is completely
231    unaffected.
232    
233    ----------------------------------------------------------------------
234    Name: Matthias Blume
235    Date: 2001/01/05 00:30:00 JST
236    Tag: blume-20010105-primops
237    Description:
238    
239    Added some experimental support for work that I am doing right now.
240    These changes mostly concern added primops, but there is also a new
241    experimental C library in the runtime system (but currently not enabled
242    anywhere except on Linux/X86).
243    
244    In the course of adding primops (and playing with them), I discovered that
245    Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
246    broken.  (Zhong was very right he labeled this stuff as "major gross hack".)
247    To recover, I made type information in INL_PRIM mandatory and changed
248    prim.sml as well as built-in.sml accordingly.  The InLine structure now
249    has complete, correct type information (i.e., no bottom types).
250    
251    Since all these changes mean that we need new binfiles, I also bumped the
252    version number to 110.32.1.
253    
254    ----------------------------------------------------------------------
255    Name: Matthias Blume
256    Date: 2000/12/30 22:10:00 JST
257    Tag: blume-20001230-various
258    Description:
259    
260    Added proxy libraries for MLRISC and let MLRISC libraries refer
261    to each other using path anchors.  (See CM manual for explanation.)
262    
263    Updated CM documentation.
264    
265    Fixed some bugs in CM.
266    
267    Implemented "proxy" libraries (= syntactic sugar for CM).
268    
269    Added "-quiet" option to makeml and changed runtime system accordingly.
270    
271    Added cleanup handler for exportML to reset timers and compiler stats.
272    
273    ----------------------------------------------------------------------
274    Name: Lal George
275    Date: 2000/12/22 22:22:58 EST 2000
276    Tag: Release_110_32
277    Description:
278    
279            Infinite precision used throughout MLRISC.
280            see MLRISC/mltree/machine-int.sig
281    
282    ----------------------------------------------------------------------
283    Name: Matthias Blume
284    Date: 2000/12/22 23:16:00 JST
285    Tag: blume-20001222-warn
286    Description:
287    
288    Corrected wording and formatting of some CM warning message which I
289    broke in my previous patch.
290    
291    ----------------------------------------------------------------------
292    Name: Matthias Blume
293    Date: 2000/12/22 21:20:00 JST
294    Tag: blume-20001222-anchorenv
295    Description:
296    
297    Fixed CM's handling of anchor environments in connection with CMB.make.
298    
299    ----------------------------------------------------------------------
300    Name: Matthias Blume
301    Date: 2000/12/22 13:15:00 JST
302    Tag: blume-20001222-cleanup
303    Description:
304    
305    Removed src/cm/ffi which does not (and did not) belong here.
306    
307    ----------------------------------------------------------------------
308    Name: Matthias Blume
309    Date: 2000/12/21 23:55:00 JST
310    Tag: blume-20001221-exn
311    Description:
312    
313    Probably most important: CM no longer silently swallows all exceptions
314    in the compiler.
315    Plus: some other minor CM changes.  For example, CM now reports some
316    sizes for generated binfiles (code, data, envpickle, lambdapickle).
317    
318    ----------------------------------------------------------------------
319    Name: Matthias Blume
320    Date: 2000/12/15 00:01:05 JST
321    Tag: blume-20001215-dirtool
322    Description:
323    
324    - "dir" tool added.
325    - improvements and cleanup to Tools structure
326    - documentation updates
327    
328    ----------------------------------------------------------------------
329    Name: Allen Leung
330    Date: Thu Dec 14 03:45:24 EST 2000
331    Description:
332    Tag:  leunga-20001214-int-inf
333    Description:
334    
335       In IntInf, added these standard functions, which are missing from our
336    implementation:
337    
338        andb : int * int -> int
339        xorb : int * int -> int
340        orb  : int * int -> int
341        notb : int -> int
342         <<   : int * word -> int
343        ~>>  : int * word -> int
344    
345       Not tested, I hope they are correct.
346    
347    ----------------------------------------------------------------------
348    Name: Allen Leung
349    Date: Fri Dec  8 19:23:26 EST 2000
350    Description:
351    Tag:  leunga-20001208-nowhere
352    Description:
353    
354      Slight improvements to the 'nowhere' tool to handle OR-patterns,
355    to generate better error messages etc.  Plus a brief manual.
356    
357    ----------------------------------------------------------------------
358    Name: Lal George
359    Date: 2000/12/08 09:54:02 EST 2000
360    Tag: Release_110_31
361    Description:
362    
363    - Version 110.31
364    ----------------------------------------------------------------------
365    Name: Allen Leung
366    Date: Thu Dec  7 22:01:04 EST 2000
367    Tag:  leunga-20001207-cell-monster-hack
368    Description:
369    
370    Major MLRISC internal changes.  Affect all clients.
371    Summary:
372    
373    1.  Type CELLS.cell = int is now replaced by a datatype.
374        As a result, the old regmap is now gone.  Almost all interfaces
375        in MLRISC change as a consequence.
376    
377    2.  A new brand version of machine description tool (v3.0) that generates
378        modules expecting the new interface.  The old version is removed.
379    
380    3.  The RA interface has been further abstracted into two new functors.
381        RISC_RA and X86RA.  These functors have much simpler interfaces.
382        [See also directory MLRISC/demo.]
383    
384    4.  Some other new source->source code generation tools are available:
385    
386        a. MLRISC/Tools/RewriteGen -- generate rewriters from rules.
387        b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules.
388           I use this tool to generate the peephole optimizers---with the new
389           cell type changes, peephole rules are becoming difficult to write
390           without conditional pattern matching.
391    
392    5.  More Intmap -> IntHashTable change.  Previous changes by Matthias didn't
393        cover the entire MLRISC source tree so many things broke.
394    
395    6.  CM files have been moved to the subdirectory MLRISC/cm.
396        They are moved because there are a lot of them and they clutter up the
397        root dir.
398    
399    7.  More detailed documentation to come...
400    
401        NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do
402        a makeml -rebuild first.  This is because of other other
403        changes that Matthias has made (see below).
404    
405    
406    ----------------------------------------------------------------------
407    Name: Matthias Blume
408    Date: 2000/11/30 23:12:00 JST
409    Tag: blume-20001130-filereorg
410    Description:
411    
412    Some manual updates and some file reorganizations in CM.
413    
414    ----------------------------------------------------------------------
415    Name: Matthias Blume
416    Date: 2000/11/24 17:45:00 JST
417    Tag: blume-20001124-link
418    Description:
419    
420    Drastically improved link traversal code for the case that the dynamic
421    value was already loaded at bootstrap time.  As a result, CM and CMB
422    now both load blazingly fast -- even on a very slow machine.  Also,
423    memory consumption has been further reduced by this.
424    
425    Warning: The format of the PIDMAP file has changed.  THerefore, to
426    bootstrap you have to do this:
427    
428    1. Run CMB.make
429    2. Make a symbolic link for the boot directory:
430         ln -s sml.boot.ARCH-OS xxx
431    3. "Rebuild" the boot directory:
432         ./makeml -boot xxx -rebuild sml ; rm xxx
433    4. Boot normally:
434          ./makeml
435    
436    ----------------------------------------------------------------------
437    Name: Matthias Blume
438    Date: 2000/11/21 21:20:00 JST
439    Tag: blume-20001121-tools
440    Description:
441    
442    Continued hacking on autoloading problem -- with success this time.
443    Also changed tool-plugin mechanism.  See new CM manual.
444    
445    ----------------------------------------------------------------------
446    Name: Matthias Blume
447    Date: 2000/11/19 14:30:00 JST
448    Tag:  blume-20001119-autoload
449    Description:
450    
451    Some hacking to make autoloading faster.  Success for CMB, no success
452    so far for CM.  There is a reduced structure CM' that autoloads faster.
453    (This is a temporary, non-documented hack to be eliminated again when
454    the general problem is solved.)
455    
456    ----------------------------------------------------------------------
457    Name: Matthias Blume
458    Date: 2000/11/17 14:10:00 JST
459    Tag: blume-20001117-pickle-lib
460    Description:
461    
462    1. Eliminated comp-lib.cm
463    2. Made pickle-lib.cm
464    3. Eliminated all uses of intset.sml (from comp-lib.cm)
465    4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
466       equivalent constructs from smlnj-lib.cm (INtHashTable).
467    5. Point 4. also goes for those uses of intmap.* in MLRISC.
468       Duplicated intmap modules thrown out.
469    6. Hunted down all duplicated SCC code and replaced it with
470       equivalent stuff (GraphSCCFn from smlnj-lib.cm).
471    7. Rewrote Feedback module.
472    8. Moved sortedlist.sml into viscomp-lib.cm.  Eventually it
473       should be thrown out and equivalent modules from smlnj-lib.cm
474       should be used (IntRedBlackSet, IntListSet, ...).
475    
476    Confirmed that compiler compiles to fixpoint.
477    
478    ----------------------------------------------------------------------
479    Name: Allen Leung
480    Date: 2000/11/10 18:00:00
481    Tag: leunga-20001110-new-x86-fp
482    
483    A new x86 floating point code generator has been added.
484    By default this is turned off.  To turn this on, do:
485    
486        CM.autoload "$smlnj/compiler.cm";
487        Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
488    
489    Changes:
490    
491    1.  Changed FTAN to FPTAN so that the assembly output is correct.
492    2.  Changed the extension callback for FTANGENT to generate:
493    
494              fptan
495              fstp  %st(0)
496        instead of
497              fptan
498              fstpl ftempmem
499    
500    3.  Numerous assembly fixes for x86.
501    
502    5.  Cleaned up the machine code output module x86/x86MC.sml and added
503        support for a whole bunch of instructions and addressing modes:
504    
505          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st, %st(n)
506          faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp  %st, %st(n)
507          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st(n), %st
508          fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
509          fxch %st(n)
510          fld %st(n)
511          fst %st(n)
512          fst mem
513          fstp %st(n)
514          fucom %st(n)
515          fucomp %st(n)
516    
517        All these are now generated when the fast fp mode is turned on.
518    
519    6.  Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
520    
521    ----------------------------------------------------------------------
522    Name: Matthias Blume
523    Date: 2000/11/09 11:20:00 JST
524    Tag: blume-20001109-scc
525    Description:
526    
527    Eliminated some code duplication:
528    
529    1. Added "where" clause to GraphSCCFn in SML/NJ Library.
530       (Otherwise the functor is useless.)
531    2. Used GraphSCCFn where SCCUtilFun was used previously.
532    3. Got rid of SCCUtilFun (in comp-lib.cm).
533    
534    ----------------------------------------------------------------------
535    Name: Lal George
536    Date: 2000/11/06 09:02:21 EST 2000
537    Tag: Release_110_30
538    Description:
539    
540    - Version 110.30
541    ----------------------------------------------------------------------
542    Name: Matthias Blume
543    Date: 2000/11/04 14:45:00
544    Tag: blume-20001104-mlbuild
545    Description:
546    
547    - Made ml-build faster on startup.
548    - Documentation fixes.
549    
550    ----------------------------------------------------------------------
551    Name: Matthias Blume
552    Date: 2000/11/02 17:00:00 JST
553    Tag: blume-20001102-condcomp
554    Description:
555    
556    - Small tweaks to pickler -- new BOOTFILES!
557    - Version bumped to 110.29.2.
558    - Added conditional compilation facility to init.cmi (see comment there).
559    ----------------------------------------------------------------------
560    Name: Allen Leung
561    Date: 2000/10/23 19:31:00
562    Tag: leunga-20001023-demo-ra
563    
564    1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
565    2. Test programs for the graph library updated
566    3. Some new MLRISC demo programs added
567    
568    ----------------------------------------------------------------------
569    Name: Matthias Blume
570    Date: 2000/08/31 22:15:00 JST
571    Tag: blume-20001017-errmsg
572    Description:
573    
574    More error message grief: Where there used to be no messages, there
575    now were some that had bogus error regions.  Fixed.
576    
577    ----------------------------------------------------------------------
578    Name: Matthias Blume
579    Date: 2000/08/31 17:30:00 JST
580    Tag: blume-20001017-v110p29p1
581    Description:
582    
583    I made a version 110.29.1 with new bootfiles.
584    
585    Changes:  Modified pickler/unpickler for faster and leaner unpickling.
586              CM documentation changes and a small bugfix in CM's error reporting.
587    
588    ----------------------------------------------------------------------
589    Name: Lal George
590    Date: 2000/09/27 14:42:35 EDT
591    Tag: george-20000927-nodestatus
592    Description:
593    
594    Changed the type of the nodestatus, so that:
595    
596            SPILLED(~1)             is now SPILLED
597            SPILLED(m) where m>=0   is now MEMREG(m)
598            SPILLED(s) where s<~1   is now SPILL_LOC(~s)
599    
600    ----------------------------------------------------------------------
601    Name: Matthias Blume
602    Date: 2000/09/07 14:45:00 JST
603    Tag: blume-20000907-cmerrmsg
604    Description:
605    
606    Small tweak to CM to avoid getting ML syntax error messages twice.
607    
608    ----------------------------------------------------------------------
609    Name: Matthias Blume
610    Date: 2000/08/31 18:00:00 JST
611    Tag: blume-20000831-cvsbootfiles
612    Description:
613    
614    New URL for boot files (because the 110.29 files on the BL server do
615    now work correctly with my updated install scripts for yacc and lex).
616    
617    ----------------------------------------------------------------------
618    Name: Matthias Blume
619    Date: 2000/08/08 12:33:00 JST
620    Tag: blume-20000808-manual
621    Description:
622    
623    Tiny update to CM manual.
624    
625    ----------------------------------------------------------------------
626    Name: Allen Leung
627    Date: 2000/08/7 19:31:00
628    Tag: leunga-20000807-a-whole-bunch-of-stuff
629    
630      Moby, C--, SSA, x86, machine descriptions etc.  Should only affect C--
631    and Mobdy.
632    
633    1.  x86
634    
635       a.  Fixes to peephole module by John and Dan.
636       b.  Assembly fix to SETcc by Allen.
637       c.  Fix to c-call by John.
638       d.  Fix to spilling by John.  (This one deals with the missing FSTPT case)
639       e.  Instruction selection optimization to SETcc as suggested by John.
640    
641           For example,
642    
643            MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
644    
645           should generate:
646    
647            MOVL a, x
648            SUBL b, x
649            SHRL 31, x
650    
651    2.  IR stuff
652    
653         A bunch of new DJ-graph related algorithms added.  These
654         speed up SSA construction.
655    
656    3.  SSA + Scheduling
657    
658         Added code for SSA and scheduling to the repository
659    
660    ----------------------------------------------------------------------
661    Name: Lal George
662    Date: 2000/07/27 11:53:14 EDT
663    
664    Tag: lal-20000727-linux-ppc
665    Description:
666    
667     Made changes to support Linux PPC.
668     p.s. I have confirmation that the 110.29 boot files work fine.
669    
670    ----------------------------------------------------------------------
671    Name: Matthias Blume
672    Date: 2000/07/27 17:40:00 JST
673    Tag: blume-20000727-scripts
674    Description:
675    
676    !!!! WARNING !!!!
677    You must recompile the runtime system!
678    !!!! WARNING !!!!
679    
680    This is basically another round of script-enhancements:
681    
682    1. sml, ml-build, and ml-makedepend accept options -D and -U to define
683       and undefine CM preprocessor symbols.
684    
685    2. ml-build avoids generating a new heap image if it finds that the
686       existing one is still ok.  (The condition is that no ML file had to
687       be recompiled and all ML files are found to be older that the heap
688       file.)
689    
690       To make this work smoothly, I also hacked the runtime system as
691       well as SMLofNJ.SysInfo to get access to the heap image suffix
692       (.sparc-solaris, ...) that is currently being used.
693    
694       Moreover, the signature of CM.mk_standalone has changed.  See the
695       CM manual.
696    
697    3. ml-makedepend accepts additional options -n, -a, and -o.  (See the
698       CM manual for details.)
699    
700    4. More CM manual updates:
701        - all of the above has been documented.
702        - there is now a section describing the (CM-related) command line
703          arguments that are accepted by the "sml" command
704    
705    ----------------------------------------------------------------------
706    Name: Matthias Blume
707    Date: 2000/07/25 16:20:00 JST
708    Tag: blume-20000725-makedepend
709    Description:
710    
711    Added a script called ml-makedepend.  This can be used in makefiles
712    for Unix' make in a way very similar to the "makedepend" command for
713    C.
714    
715    The script internally uses function CM.sources.
716    
717    Synopsis:
718    
719        ml-makedepend [-f makefile] cmfile targetname
720    
721    The default for the makefile is "makefile" (or "Makefile" should
722    "makefile" not exist).
723    
724    ml-makedepend adds a cmfile/targetname-specific section to this
725    makefile (after removing the previous version of this section).  The
726    section contains a single dependency specification with targetname on
727    the LHS (targetname is an arbitrary name), and a list of files derived
728    from the cmfile on the RHS.  Some of the files on the RHS are
729    ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
730    "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
731    path names.  The makefile writer is responsible for making sure that
732    these variables have correct at the time "make" is invoked.
733    
734    ----------------------------------------------------------------------
735    Name: Matthias Blume
736    Date: 2000/07/22 23:30:00 JST
737    Tag: blume-20000722-urlupdate
738    Description:
739    
740    Changed BOOT and config/srcarchiveurl to point to BL server:
741    
742        ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
743    
744    ----------------------------------------------------------------------
745    Name: Matthias Blume
746    Date: 2000/07/18 18:00:00 JST
747    Tag: blume-20000718-Version_110_29
748    Description:
749    
750    1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
751    
752    2. Updated config/version to 110.29
753    
754    3. Updated config/srcarchiveurl
755    
756    3. New boot files!
757       ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
758    
759    ----------------------------------------------------------------------
760    Name: Matthias Blume
761    Date: 2000/07/11 13:58:00 JST
762    Tag: blume-20000711-doctypo
763    Description:
764    
765    Fixed a few typos in CM manual.
766    
767    ----------------------------------------------------------------------
768    Name: Allen Leung
769    Date: 2000/06/15 00:38:00
770    Tag: leunga-20000704-sparc-x86
771    
772    1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
773    2. fix to x86 RET bug [found by Dan Grossman]
774    3. sparc assembly bug fix for ticc instructions [found by Fermin]
775    
776       Affects c-- and moby only
777    
778    ----------------------------------------------------------------------
779    Name: Matthias Blume
780    Date: 2000/07/04 15:26:00
781    Tag: blume-20000704-trigger
782    Description:
783    
784    1. Improvements to CM manual.
785    2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
786       of getting a back-trace.  The function, when called, raises an
787       internal exception which explicitly carries the full back-trace history,
788       so it is unaffected by any intervening handle-raise pairs ("trivial"
789       or not).  The interactive loop will print that history once it arrives
790       at top level.
791       Short of having all exceptions implicitly carry the full history, the
792       recommended way of using this facility is:
793         - compile your program with instrumentation "on"
794         - run it, when it raises an exception, look at the history
795         - if the history is "cut off" because of some handler, go and modify
796           your program so that it explicitly calls BTrace.trigger
797         - recompile (still instrumented), and rerun; look at the full history
798    
799    ----------------------------------------------------------------------
800    Name: Matthias Blume
801    Date: 2000/07/03 15:36:00 JST
802    Tag: blume-20000702-manual
803    Description:
804    
805    Small corrections and updates to CM manual.
806    
807    ----------------------------------------------------------------------
808    Name: Matthias Blume
809    Date: 2000/06/29 16:04:00 JST
810    Tag: blume-20000629-yacctool
811    Description:
812    
813    Changes:
814    
815    1. Class "mlyacc" now takes separate arguments to pass options to
816       generated .sml- and .sig-files independently.
817    2. Corresponding CM manual updates.
818    3. BTrace module now also reports call sites.  (However, for loop clusters
819       it only shows from where the cluster was entered.)  There are associated
820       modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
821    
822    ----------------------------------------------------------------------
823    Name: Matthias Blume
824    Date: 2000/06/27 16:51:00 JST
825    Tag: blume-20000627-noweb
826    Description:
827    
828    Changes:
829    
830     1. Implemented "subdir" and "witness" options for noweb tool.
831        This caused some slight internal changes in CM's tool implementation.
832     2. Fixed bug in "tool plugin" mechanism.  This is essentially cleaning
833        some remaining issues from earlier path anchor changes.
834     3. Updated CM manual accordingly.
835    
836     4. Changed implementation of back-tracing so that I now consider it
837        ready for prime-time.
838    
839        In particular, you don't have to explicitly trigger the back-trace
840        anymore.  Instead, if you are running BTrace-instrumented code and
841        there is an uncaught exception (regardless of whether or not it was
842        raised in instrumented code), the top-level evalloop will print
843        the back-trace.
844    
845        Features:
846    
847          - Instrumented and uninstrumented code work together seemlessly.
848            (Of course, uninstrumented code is never mentioned in actual
849             back-traces.)
850    
851          - Asymptotic time- and space-complexity of instrumented code is
852            equal to that of uninstrumented code.  (This means that
853            tail-recursion is preserved by the instrumentation phase.)
854    
855          - Modules whose code has been instrumented in different sessions
856            work together without problem.
857    
858          - There is no penalty whatsoever on uninstrumented code.
859    
860          - There is no penalty on "raise" expressions, even in
861            instrumented code.
862    
863        A potential bug (or perhaps it is a feature, too):
864    
865          A back-trace reaches no further than the outermost instrumented
866          non-trivial "raise".  Here, a "trivial" raise is one that is the
867          sole RHS of a "handle" rule.  Thus, back-traces reach trough
868    
869               <exp> handle e => raise e
870    
871          and even
872    
873               <exp> handle Foo => raise Bar
874    
875          and, of course, through
876    
877               <exp> handle Foo => ...
878    
879         if the exception was not Foo.
880    
881         Back-traces always reach right through any un-instrumented code
882         including any of its "handle" expressions, trivial or not.
883    
884       To try this out, do the following:
885    
886         - Erase all existing binfiles for your program.
887           (You may keep binfiles for those modules where you think you
888            definitely don't need back-tracing.)
889         - Turn on back-trace instrumentation:
890              SMLofNJ.Internals.BTrace.mode (SOME true);
891         - Recompile your program.  (I.e., run "CM.make" or "use".)
892         - You may now turn instrumentation off again (if you want):
893              SMLofNJ.Internals.BTrace.mode (SOME false);
894         - Run your program as usual.  If it raises an exception that
895           reaches the interactive toplevel, then a back-trace will
896           automatically be printed.  After that, the toplevel loop
897           will print the exception history as usual.
898    
899    ----------------------------------------------------------------------
900    Name: Matthias Blume
901    Date: 2000/06/26 09:56:46 JST
902    Tag: blume-20000626-setup
903    Description:
904    
905    CM: - setup-parameter to "sml" added; this can be used to run arbitrary
906          ML code before and after compiling a file (e.g., to set compiler
907          flags)
908    
909    Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
910              - associated changes to btrace.sml (BTrace instrumentation pass)
911              - cleaner implementation of btimp.sml (BTrace tracing and report
912                module)
913    
914    CM manual: * new path encoding documented
915               * description of setup-parameter to "sml" added
916    
917    The biggest user-visible change to back-tracing is that it is no
918    longer necessary to compile all traced modules within the same
919    session.  (This was a real limitation.)
920    
921    ----------------------------------------------------------------------
922    Name: Matthias Blume
923    Date: 2000/06/24 12:40:00 JST
924    Tag: blume-20000624-startup
925    Description:
926    
927    Fixes startup slowdown problem.  (I was calling SrcPath.sync a _tad_
928    bit too often -- to put it mildly. :)
929    
930    ----------------------------------------------------------------------
931    Name: Matthias Blume
932    Date: 2000/06/23 18:20:00 JST
933    Tag: blume-20000623-btrace
934    Description:
935    
936    This updates adds a backtrace facility to aid programmers in debugging
937    their programs.  This involves the following changes:
938    
939    1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
940       keeping track of the current call stack.  When programs are compiled
941       in a special mode, the compiler will insert calls to these hooks
942       into the user program.
943       "Hook" means that it is possible for different implementations of
944       back-tracing to register themselves (at different times).
945    
946    2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
947       as an Absyn.dec->Absyn.dec rewrite.  Normally this phase is turned off.
948       It can be turned on using this call:
949         SMLofNJ.Internals.BTrace.mode (SOME true);
950       Turning it off again:
951         SMLofNJ.Internals.BTrace.mode (SOME false);
952       Querying the current status:
953         SMLofNJ.Internals.BTrace.mode NONE;
954       Annotated programs are about twice as big as normal ones, and they
955       run a factor of 2 to 4 slower with a dummy back-trace plugin (one
956       where all hooks do nothing).  The slowdown with a plugin that is
957       actually useful (such as the one supplied by default) is even greater,
958       but in the case of the default plugin it is still only an constant
959       factor (amortized).
960    
961    3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
962       with a sub-structure BTrace for controlling back-tracing.  In particular,
963       the above-mentioned function "mode" controls whether the annotation
964       phase is invoked by the compiler.  Another important function is
965       "trigger": when called it aborts the current execution and causes
966       the top-level loop to print a full back-trace.
967    
968    4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
969       for back-tracing.  It keeps track of the dynamic call stack and in
970       addition to that it keeps a partial history at each "level" of that
971       stack.  For example, if a tail-calls b, b tail-calls c, and c tail-calls
972       d and b (at separate times, dynamically), then the report will show:
973    
974       GOTO   d
975             /c
976       GOTO  \b
977       CALL   a
978    
979       This shows that there was an initial non-tail call of a, then a
980       tail-call to b or c, looping behavior in a cluster of functions that
981       consist of b and c, and then a goto from that cluster (i.e., either from
982       b or from c) to d.
983    
984       Note that (depending on the user program) the amount of information
985       that the back-trace module has to keep track of at each level is bounded
986       by a constant.  Thus, the whole implementation has the same asymptotical
987       complexity as the original program (both in space and in time).
988    
989    5. compiler/TopLevel/interact/evalloop.sml has been modified to
990       handle the special exception SMLofNJ.Internals.BTrace.BTrace
991       which is raised by the "trigger" function mentioned above.
992    
993    Notes on usage:
994    
995    - Annotated code works well together with unannotated code:
996    Unannotated calls simply do not show up at all in the backtrace.
997    
998    - It is not a good idea to let modules that were annotated during
999    different sessions run at the same time.  This is because the compiler
1000    chooses small integers to identify individual functions, and there
1001    will be clashes if different modules were compiled in separate sessions.
1002    (Nothing will crash, and you will even be told about the clashes, but
1003    back-trace information will in general not be useful.)
1004    
1005    - Back-tracing can be confused by callcc and capture.
1006    
1007    - The only way of getting a back-trace right now is to explicitly
1008    invoke the "trigger" function from your user program.  Eventually, we
1009    should make every exception carry back-trace information (if
1010    available).  But since this creates more overhead at "raise"-time
1011    (similar to the current exnHistory overhead), I have not yet
1012    implemented this.  (The implementation will be rather easy.)  With
1013    exceptions carrying back-trace information, this facility will be even
1014    more useful because users don't need to modify their programs...
1015    
1016    - While it is possible to compile the compiler with back-trace
1017    annotations turned on (I did it to get some confidence in
1018    correctness), you must make absolutely sure that core.sml and
1019    btimp.sml are compiled WITHOUT annotation!  (core.sml cannot actually
1020    be compiled with annotation because there is no core access yet, but
1021    if you compile btimp.sml with annotation, then the system will go into
1022    an infinite recursion and crash.)
1023    Since CM currently does not know about BTrace, the only way to turn
1024    annotations on and off for different modules of the compiler is to
1025    interrupt CMB.make, change the settings, and re-invoke it.  Of course,
1026    this is awkward and clumsy.
1027    
1028    Sample sessions:
1029    
1030    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
1031    - SMLofNJ.Internals.BTrace.mode (SOME true);
1032    [autoloading]
1033    [autoloading done]
1034    val it = false : bool
1035    - structure X = struct
1036    -     fun main n = let
1037    -         fun a (x, 0) = d x
1038    -           | a (x, n) = b (x, n - 1)
1039    -         and b (x, n) = c (x, n)
1040    -         and c (x, n) = a (x, n)
1041    -         and d x = e (x, 3)
1042    -         and e (x, 0) = f x
1043    -           | e (x, n) = e (x, n - 1)
1044    -         and f 0 = SMLofNJ.Internals.BTrace.trigger ()
1045    -           | f n = n * g (n - 1)
1046    -         and g n = a (n, 3)
1047    -     in
1048    -         f n
1049    -     end
1050    - end;
1051    structure X : sig val main : int -> int end
1052    - X.main 3;
1053    *** BACK-TRACE ***
1054    GOTO   stdIn:4.2-13.20: X.main[2].f
1055    GOTO-( stdIn:4.2-13.20: X.main[2].e
1056    GOTO   stdIn:4.2-13.20: X.main[2].d
1057         / stdIn:4.2-13.20: X.main[2].a
1058         | stdIn:4.2-13.20: X.main[2].b
1059    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1060    CALL   stdIn:4.2-13.20: X.main[2].g
1061    GOTO   stdIn:4.2-13.20: X.main[2].f
1062    GOTO-( stdIn:4.2-13.20: X.main[2].e
1063    GOTO   stdIn:4.2-13.20: X.main[2].d
1064         / stdIn:4.2-13.20: X.main[2].a
1065         | stdIn:4.2-13.20: X.main[2].b
1066    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1067    CALL   stdIn:4.2-13.20: X.main[2].g
1068    GOTO   stdIn:4.2-13.20: X.main[2].f
1069    GOTO-( stdIn:4.2-13.20: X.main[2].e
1070    GOTO   stdIn:4.2-13.20: X.main[2].d
1071         / stdIn:4.2-13.20: X.main[2].a
1072         | stdIn:4.2-13.20: X.main[2].b
1073    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1074    CALL   stdIn:4.2-13.20: X.main[2].g
1075    GOTO   stdIn:4.2-13.20: X.main[2].f
1076    CALL   stdIn:2.15-17.4: X.main[2]
1077    -
1078    
1079    (Note that because of a FLINt bug the above code currently does not
1080    compile without BTrace turned on.)
1081    
1082    Here is another example, using my modified Tiger compiler:
1083    
1084    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
1085    - SMLofNJ.Internals.BTrace.mode (SOME true);
1086    [autoloading]
1087    [autoloading done]
1088    val it = false : bool
1089    - CM.make "sources.cm";
1090    [autoloading]
1091    ...
1092    [autoloading done]
1093    [scanning sources.cm]
1094    [parsing (sources.cm):parse.sml]
1095    [creating directory CM/SKEL ...]
1096    [parsing (sources.cm):tiger.lex.sml]
1097    ...
1098    [wrote CM/sparc-unix/semant.sml]
1099    [compiling (sources.cm):main.sml]
1100    [wrote CM/sparc-unix/main.sml]
1101    [New bindings added.]
1102    val it = true : bool
1103    - Main.compile ("../testcases/merge.tig", "foo.out");
1104    *** BACK-TRACE ***
1105    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
1106    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1107    CALL   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1108    GOTO   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1109    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1110    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1111    CALL   lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
1112         / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
1113    CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
1114    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1115    CALL   lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
1116    CALL   main.sml:1.18-118.4: Main.compile[2]
1117    -
1118    
1119    ----------------------------------------------------------------------
1120    Name: Matthias Blumen
1121    Date: 2000/06/21 18:00:00 JST
1122    Tag: blume-20000621-manual
1123    Description:
1124    
1125    CM manual update: Path environments documented.
1126    
1127    ----------------------------------------------------------------------
1128    Name: Matthias Blume
1129    Date: 2000/06/19 13:40:00
1130    Tag: blume-20000619-manual
1131    Description:
1132    
1133    CM manual and system/README update.  This only covers the fact that
1134    there are no more implicit anchors.  (Path environments and the "bind"
1135    option to "cm" have yet to be documented.)
1136    
1137    ----------------------------------------------------------------------
1138    Name: Matthias Blume
1139    Date: 2000/06/19 11:05:00 JST
1140    Tag: blume-20000619-chdir-bugfix
1141    Description:
1142    
1143    Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
1144    
1145    ----------------------------------------------------------------------
1146    Name: Matthias Blume
1147    Date: 2000/06/18 22:00:10 JST
1148    Tag: blume-20000618-implicit-anchors-really-gone
1149    Description:
1150    
1151    I updates the previous HISTORY entry where I forgot to mention that
1152    implicit anchors are no longer with us.
1153    
1154    The current update also gets rid of the (now useless) controller
1155    CM.Control.implicit_anchors.
1156    
1157    ----------------------------------------------------------------------
1158    Name: Matthias Blume
1159    Date: 2000/06/16 17:30:00 JST
1160    Tag: blume-20000616-anchorenv
1161    Description:
1162    
1163    This patch implements the long anticipated (just kidding :) "anchor
1164    environment" mechanism.  In the course of doing this, I also
1165    re-implemented CM's internal "SrcPath" module from scratch.  The new
1166    one should be more robust in certain boundary cases.  In any case, it
1167    is a lot cleaner than its predecessor (IMHO).
1168    
1169    This time, although there is yet another boot file format change, I
1170    kept the unpickler backward-compatible.  As a result, no new bootfiles
1171    are necessary and bootstrapping is straightforward.  (You cannot read
1172    new bootfiles into an old system, but the other way around is no
1173    problem.)
1174    
1175    Visible changes:
1176    
1177    ** 0. Implicit path anchors (without the leading $-symbol) are no
1178    longer recognized at all. This means that such path names are not
1179    illegal either.  For example, the name basis.cm simply refers to a
1180    local file called "basis.cm" (i.e, the name is an ordinary path
1181    relative to .cm-files directory).  Or, to put it differently, only
1182    names that start with $ are anchored paths.
1183    
1184    ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
1185    vanished.
1186    
1187    John (Reppy) had critizised this as soon as I originally proposed and
1188    implemented it, but at that time I did not really deeply believe
1189    him. :) Now I came full-circle because I need the $<singlearc> syntax
1190    in another place where it cannot be seen as an abbreviation for
1191    $/<singlearc>.  To avoid the confusion, $<singlearc> now means what it
1192    seems to mean (i.e., it "expands" into the corresponding anchor
1193    value).
1194    
1195    However, when paths are used as members in CM description files, it
1196    continues to be true that there must be at least another arc after the
1197    anchor.  This is now enforced separately during semantic analysis
1198    (i.e., from a lexical/syntactical point of view, the notation is ok.)
1199    
1200    ** 2. The "cm" class now accepts an option "bind".  The option's value
1201    is a sub-option list of precisely two items -- one labeled "anchor"
1202    and the other one labeled "value".  As you might expect, "anchor" is
1203    used to specify an anchor name to be bound, and "value" specifies what
1204    the anchor is being bound to.
1205    
1206    The value must be a directory name and can be given in either standard
1207    syntax (including the possibility that it is itself an anchored path)
1208    or native syntax.
1209    
1210    Examples:
1211    
1212       foo.cm (bind:(anchor:bar value:$mystuff/bar))
1213       lib.cm (bind:(anchor:a value:"H:\\x\\y\\z"))  (* only works under windows *)
1214    
1215    and so on.
1216    
1217    The meaning of this is that the .cm-file will be processed with an
1218    augmented anchor environment where the given anchor(s) is/are bound to
1219    the given values(s).
1220    
1221    The rationale for having this feature is this: Suppose you are trying
1222    to use two different (already stable) libraries a.cm and b.cm (that
1223    you perhaps didn't write yourself).  Further, suppose each of these
1224    two libraries internally uses its own auxiliary library $aux/lib.cm.
1225    Normally you would now have a problem because the anchor "lib" can not
1226    be bound to more than one value globally.  Therefore, the project that
1227    uses both a.cm and b.cm must locally redirect the anchor to some other
1228    place:
1229    
1230       a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
1231       b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
1232    
1233    This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
1234    /usr/lib/smlnj/b-stuff/aux.cm, respectively.
1235    
1236    Hard-wiring path names is a bit inflexible (and CM will verbosely warn
1237    you when you do so at the time of CM.stabilize).  Therefore, you can
1238    also use an anchored path as the value:
1239    
1240      a.cm (bind:(anchor:lib value:$a-lib))
1241      b.cm (bind:(anchor:lib value:$b-lib))
1242    
1243    Now you can globally configure (using the usual CM.Anchor.anchor or
1244    pathconfig machinery) bindings for "a-lib" and "b-lib".  Since "lib"
1245    itself is always locally bound, setting it globally is no longer
1246    meaningful or necessary (but it does not hurt either).  In fact, "lib"
1247    can still be used as a global anchor for separate purposes.  As a
1248    matter of fact, one can locally define "lib" in terms of a global
1249    "lib":
1250    
1251      a.cm (bind:(anchor:lib value:$lib/a))
1252      b.cm (bind:(anchor:lib value:$lib/b))
1253    
1254    ** 3: The encoding of path names has changed.  This affects the way
1255    path names are shown in CM's progress report and also the internal
1256    protocol encoding used for parallel make.
1257    
1258    The encoding now uses one or more ':'-separated segments.  Each
1259    segments corresponds to a file that has been specified relative to the
1260    file given by its preceding segment.  The first segment is either
1261    relative to the CWD, absolute, or anchored.  Each segment itself is
1262    basically a Unix pathname; all segments but the first are relative.
1263    
1264    Example:
1265    
1266       $foo/bar/baz.cm:a/b/c.sml
1267    
1268    This path denotes the file bar/a/b/c.sml relative to the directory
1269    denoted by anchor "foo".  Notice that the encoding also includes
1270    baz.cm which is the .cm-file that listed a/b/c.sml.  As usual, such
1271    paths are resolved relative to the .cm-files directory, so baz.cm must
1272    be ignored to get the "real" pathname.
1273    
1274    To make this fact more obvious, CM puts the names of such "virtual
1275    arcs" into parentheses when they appear in progress reports. (No
1276    parentheses will appear in the internal protocol encoding.)  Thus,
1277    what you really see is:
1278    
1279      $foo/bar/(baz.cm):a/b/c.sml
1280    
1281    I find this notation to be much more informative than before.
1282    
1283    Another new feature of the encoding is that special characters
1284    including parentheses, colons, (back)slashes, and white space are
1285    written as \ddd (where ddd is the decimal encoding of the character).
1286    
1287    *** The CM manual still needs to be updated.
1288    
1289    ----------------------------------------------------------------------
1290    Name: Allen Leung
1291    Date: 2000/06/15 00:38:00
1292    Tag: leunga-20000615-x86-peephole
1293    
1294    x86 Peephole fix by Fermin.  Affects c-- and moby only.
1295    
1296    ----------------------------------------------------------------------
1297    Name: Matthias Blume
1298    Date: 2000/06/12 11:40:00
1299    Tag: blume-20000612-parmakefix
1300    Description:
1301    
1302    More cleanup after changing the file naming scheme: This time I
1303    repaired the parallel make mechanism for CMB.make which I broke earlier.
1304    
1305    ----------------------------------------------------------------------
1306    Name: Allen Leung
1307    Date: 2000/06/09 01:25:00
1308    Tag: leunga-20000609-various
1309    
1310    None of these things should affect normal SML/NJ operations
1311    
1312    1. Peephole improvements provided by Fermin (c--)
1313    2. New annotation DEFUSE for adding extra dependence (moby)
1314    3. New X86 LOCK instructions (moby)
1315    4. New machine description language for reservation tables (scheduling)
1316    5. Fixes to various optimization/analysis modules (branch chaining, dominator
1317       trees etc.)
1318    6. I've changed the CM files so that they can work with versions
1319       110.0.6, 110.25 and 110.28
1320    
1321    ----------------------------------------------------------------------
1322    Name: Matthias Blume
1323    Date: 2000/06/09 12:40:00
1324    Tag: blume-20000609-log
1325    Description:
1326    
1327    - Removed all(?) remaining RCS Log entries from sources.
1328    
1329    - Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
1330      anchored paths).
1331    
1332    ----------------------------------------------------------------------
1333    Name: Matthias Blume
1334    Date: 2000/06/07 17:00:00 JST
1335    Tag: blume-20000607-no-implicit-anchors
1336    Description:
1337    
1338    1. This update changes the default setting for
1339    CM.Control.implicit_anchors from true to false.  This means that
1340    implicit anchors are no longer permitted by default.  I also tried to
1341    make sure that nothing else still relies on implicit anchors.
1342    (This is the next step on the schedule towards a CM that does not even
1343    have the notion of implicit anchors anymore.)
1344    
1345    2. More CM manual updates.
1346    
1347    3. I managed to track down and fix the pickling bug I mentioned last
1348    time.  Because of the previously existing workaround, this entails no
1349    immediate practical changes.
1350    
1351    ----------------------------------------------------------------------
1352    Name: Matthias Blume
1353    Date: 2000/06/06 11:15:00 JST
1354    Tag: blume-20000606-lazierpickle
1355    Description:
1356    
1357    !!!! NEW BOOT FILES !!!!
1358    
1359    * The main purpose of this update is to make library pickles lazier in
1360    order to reduce the initial space penalty for autoloading a library.
1361    As a result, it is now possible to have $smlnj/compiler.cm
1362    pre-registered.  This should take care of the many complaints or
1363    inquiries about missing structure Compiler.  This required changes to
1364    CM's internal data structures and small tweaks to some algorithms.
1365    
1366    As a neat additional effect, it is no longer necessary (for the sake
1367    of lean heap image files) to distinguish between a "minimal" CM and a
1368    "full" CM.  Now, there is only one CM (i.e., the "full" version:
1369    $smlnj/cm.cm aka $smlnj/cm/full.cm), and it is always available at the
1370    interactive top level. ($smlnj/cm/minimal.cm is gone.)
1371    
1372    To make the life of compiler-hackers easier, "makeml" now also
1373    pre-registers $smlnj/cmb.cm (aka $smlnj/cmb/current.cm).  In other
1374    words, after you bootstrap a new sml for the first time, you will not
1375    have to autoload $smlnj/cmb.cm again afterwards.  (The first time
1376    around you will still have to do it, though.)
1377    
1378    * A second change consists of major updates to the CM manual.  There
1379    are now several appendices with summary information and also a full
1380    specification of the CM description file syntax.
1381    
1382    * In directory src/system I added the script "allcross".  This script
1383    invokes sml and cross-compiles the compiler for all supported
1384    architectures.  (Useful when providing a new set of boot files.)
1385    
1386    * There seems to be a latent bug in my "lazy pickles" mechanism.  I
1387    added a small tweak to pickle-util.sml to work around this problem,
1388    but it is not a proper fix yet.  I will investigate further.  (The
1389    effect of the bug was an inflation of library pickle size.)
1390    
1391    * Version number increased to 110.28.1 (to avoid compatibility problems).
1392    
1393    ----------------------------------------------------------------------
1394    Name: Allen Leung
1395    Date: 2000/05/25 17:28 EDT
1396    Tag: leunga-20000525-ra
1397    Description:
1398    
1399      Fixed a bug in freezing phase of the register allocator.
1400    
1401    ----------------------------------------------------------------------
1402    Name: Allen Leung
1403    Date: 2000/05/15 22:53 EDT
1404    Tag: leunga-20000515-alpha-x86-ra
1405    Description:
1406    
1407      1. Alpha
1408    
1409          Slight cleanup.  Removed the instruction SGNXL
1410    
1411      2. X86
1412    
1413          Added the following instructions to the instruction set:
1414    
1415            ROLx, RORx,
1416            BTx, BTSx, BTLx, BTRx,
1417            XCHGx, and variants with the LOCK prefix
1418    
1419      3. Register Allocation
1420    
1421          The module ra-rewrite-with-renaming has been improved.
1422    
1423      These have no effect on SML/NJ.
1424    
1425    ----------------------------------------------------------------------
1426    Name: Matthias Blume
1427    Date: 2000/05/15 16:20:00 JST
1428    Tag: blume-20000515-lightrebuild
1429    Description:
1430    
1431    1. I added an alternative to "-rebuild" to "makeml".  The difference is
1432       that prior to calling CMB.make' the CM-variable "LIGHT" will be
1433       defined.  In effect, the command will not build any cross-compiler
1434       backends and therefore finish more quickly.
1435    
1436       The "fixpt" script also takes a "-light" switch to be able to use
1437       this new facility while compiling for a fixpoint.
1438    
1439    2. I replaced all mentions of anchored paths in group owner specifications
1440       with simple relative paths (usually starting with "..").
1441       The rationale is that a library's internal workings should not be
1442       compromised by the lack of some anchor.  (An anchor is necessary
1443       for someone who wants to refer to the library by an anchored path,
1444       but it should not be necessary to build the same library in the first
1445       place.)
1446    
1447    3. I changed the way CM's tool mechanism determines the shell command
1448       string used for things like ml-yacc etc. so that it does not break
1449       when CM.Control.implicit_anchors is turned off.
1450    
1451  ----------------------------------------------------------------------  ----------------------------------------------------------------------
1452  Name: Matthias Blume  Name: Matthias Blume
1453  Date: 2000/05/12 18:20:00 JST  Date: 2000/05/12 18:20:00 JST
# Line 444  Line 1881 
1881    
1882  3.  Assembly  3.  Assembly
1883    
1884      When generating assemby, resolve the value of client defined constants,      When generating assembly, resolve the value of client defined constants,
1885      instead of generating symbolic values.  This is controlled by the      instead of generating symbolic values.  This is controlled by the
1886      new flag "asm-resolve-constants", which is default to true.      new flag "asm-resolve-constants", which is default to true.
1887    
# Line 467  Line 1904 
1904    
1905  To this end, I arranged that instead of "structure Core" as "structure  To this end, I arranged that instead of "structure Core" as "structure
1906  _Core" is bound in the pervasive environment.  Core access is done via  _Core" is bound in the pervasive environment.  Core access is done via
1907  _Core (which can never be accidentially rebound because _Core is not a  _Core (which can never be accidentally rebound because _Core is not a
1908  legal surface-syntax symbol).  legal surface-syntax symbol).
1909    
1910  The current solution is much cleaner because the core environment is  The current solution is much cleaner because the core environment is
# Line 477  Line 1914 
1914  with dynamic and symbolic parts of the core environment.  with dynamic and symbolic parts of the core environment.
1915    
1916  Remaining hackery (to bind the "magic" symbol _Core) is localized in the  Remaining hackery (to bind the "magic" symbol _Core) is localized in the
1917  compilation mananger's bootstrap compiler (actually: in the "init group"  compilation manager's bootstrap compiler (actually: in the "init group"
1918  handling).  See the comments in src/system/smlnj/init/init.cmi for  handling).  See the comments in src/system/smlnj/init/init.cmi for
1919  more details.  more details.
1920    
# Line 592  Line 2029 
2029     (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
2030     used).     used).
2031    
2032     This was done to accomodate the new "make" and "shell" tools which     This was done to accommodate the new "make" and "shell" tools which
2033     facilitate fairly seemless hookup to portions of code managed using     facilitate fairly seamless hookup to portions of code managed using
2034     Makefiles or Shell scripts.     Makefiles or Shell scripts.
2035    
2036     There are no classes "shared" or "private" anymore.  Instead, the     There are no classes "shared" or "private" anymore.  Instead, the
# Line 605  Line 2042 
2042    
2043     All existing tools are described in the CM manual.     All existing tools are described in the CM manual.
2044    
2045     - Slightly better error handling.  (CM now surpresses many followup     - Slightly better error handling.  (CM now suppresses many followup
2046     error messages that tended to be more annoying than helpful.)     error messages that tended to be more annoying than helpful.)
2047    
2048  2. Major changes to the compiler's static environment data structures.  2. Major changes to the compiler's static environment data structures.
# Line 739  Line 2176 
2176    
2177       I've changed andl to testl in the floating point test sequence       I've changed andl to testl in the floating point test sequence
2178       whenever appropriate.  The Intel optimization guide states that       whenever appropriate.  The Intel optimization guide states that
2179       testl is perferable to andl.       testl is preferable to andl.
2180    
2181  7. RA (x86 only)  7. RA (x86 only)
2182    
# Line 921  Line 2358 
2358    
2359  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
2360  additional argument called "template" which is an optional string that  additional argument called "template" which is an optional string that
2361  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
2362  explanation.  explanation.
2363    
2364  2. A special-purpose tool can be "regisitered" by simply dropping the  2. A special-purpose tool can be "registered" by simply dropping the
2365  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
2366  directory where the .cm file lives that uses this tool.  (The  directory where the .cm file lives that uses this tool.  (The
2367  behavior/misfeature until now was to look for the tool description  behavior/misfeature until now was to look for the tool description
# Line 968  Line 2405 
2405  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
2406  <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,
2407  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
2408  perhaps got accidentially killed).  In this case, fixpt will quickly  perhaps got accidentally killed).  In this case, fixpt will quickly
2409  move through what exists before continuing where it left off earlier,  move through what exists before continuing where it left off earlier,
2410  and, thus, saves a lot of time.  and, thus, saves a lot of time.
2411    
# Line 1018  Line 2455 
2455    it from that remote directory.    it from that remote directory.
2456    This should simplify installation further:  For machines that have    This should simplify installation further:  For machines that have
2457    access to the internet, just fetch <version>-config.tgz, unpack it,    access to the internet, just fetch <version>-config.tgz, unpack it,
2458    edit config/targets, and go (run config/install.sh).  The scipt will    edit config/targets, and go (run config/install.sh).  The script will
2459    fetch everything else that it might need all by itself.    fetch everything else that it might need all by itself.
2460    
2461    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.644  
changed lines
  Added in v.810

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