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 624, Fri Apr 21 03:06:21 2000 UTC revision 843, Tue Jun 19 21:53:04 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: 2001/06/19 17:55:00 EDT
18    Tag: blume-20010619-instantiate
19    Description:
20    
21    This un-breaks the fix for bug 1432.
22    (The bug was originally fixed in 110.9 but I broke it again some
23    time after that.)
24    
25    ----------------------------------------------------------------------
26    Name: Matthias Blume
27    Date: 2001/06/19 17:25:00 EDT
28    Tag: blume-20010619-signals
29    Description:
30    
31    This should (hopefully) fix the long-standing signal handling bug.
32    (The runtime system was constructing a continuation record with an
33    incorrect descriptor which would cause the GC to drop data on the floor...)
34    
35    ----------------------------------------------------------------------
36    Name: Matthias Blume
37    Date: 2001/06/15 15:05:00 EDT
38    Tag: blume-20010615-moresparc
39    Description:
40    
41    Here is a short late-hour update related to Sparc c-calls:
42    
43     -- made handling of double-word arguments a bit smarter
44    
45     -- instruction selection phase tries to collapse certain clumsily
46        constructed ML-Trees; typical example:
47    
48            ADD(ty,ADD(_,e,LI d1),LI d2)  ->  ADD(ty,e,LI(d1+d2))
49    
50        This currently has no further impact on SML/NJ since mlriscGen does
51        not seem to generate such patterns in the first place, and c-calls
52        (which did generate them in the beginning) has meanwhile been fixed
53        so as to avoid them as well.
54    
55    ----------------------------------------------------------------------
56    Name: Matthias Blume
57    Date: 2001/06/15 15:05:00 EDT
58    Tag: blume-20010615-sparc
59    Description:
60    
61    The purpose of this update is to provide an implementation of NLFFI
62    on Sparc machines.
63    
64    Here are the changes in detail:
65    
66       * src/MLRISC/sparc/c-calls/sparc-c-calls.sml is a new file containing
67       the Sparc implementation of the c-calls API.
68       * The Sparc backend of SML/NJ has been modified to uniformely use %fp
69       for accessing the ML frame.  Thus, we have a real frame pointer and
70       can freely modify %sp without need for an omit-frame-ptr phase.
71       The vfp logic in src/compiler/CodeGen/* has been changed to accomodate
72       this case.
73       * ml-nlffigen has been taught to produce code for different architectures
74       and calling conventions.
75       * In a way similar to what was done in the x86 case, the Sparc
76       backend uses its own specific extension to mltree.  (For example,
77       it needs to be able to generate UNIMP instructions which are part
78       of the calling convention.)
79       * ml-nlffi-lib was reorganized to make it more modular (in particular,
80       to make it easier to plug in new machine- and os-dependent parts).
81    
82    There are some other fairly unrelated bug fixes and cleanups as well:
83    
84       * I further hacked the .cm files for MLRISC tools (like MDLGen) so
85       that they properly share their libraries with existing SML/NJ libraries.
86       * I fixed a minor cosmetic bug in CM, supressing certain spurious
87       follow-up error messages.
88       * Updates to CM/CMB documentation.
89    
90    TODO items:
91    
92       * MLRISC should use a different register as its asmTemp on the Sparc.
93         (The current %o2 is a really bad choice because it is part of the
94         calling conventions, so things might interfere in unexpected ways.)
95    
96    ----------------------------------------------------------------------
97    Name: Matthias Blume
98    Date: 2001/06/07
99    Tag: blume-20010607-calls
100    Description:
101    
102    A number of internal changes related to C calls and calling conventions:
103    
104    1. ML-Tree CALL statements now carry a "pops" field.  It indicates the
105       number of bytes popped implicitly (by the callee).  In most cases
106       this field is 0 but on x86/win32 it is some non-zero value.  This
107       is information provided for the benefit of the "omit-frameptr" pass.
108    2. The CALL instruction on the x86 carries a similar "pops" field.
109       The instruction selection phase copies its value from the ML-Tree
110       CALL statement.
111    3. On all other architectures, the instruction selection phase checks
112       whether "pops=0" and complains if not.
113    4. The c-calls implementation for x86 now accepts two calling conventions:
114       "ccall" and "stdcall".  When "ccall" is selected, the caller cleans
115       up after the call and pops is set to 0.  For "stdcall", the caller
116       does nothing, leaving the cleanup to the callee; pops is set to
117       the number of bytes that were pushed onto the stack.
118    5. The cproto decoder (compiler/Semant/types/cproto.sml) now can
119       distinguish between "ccall" and "stdcall".
120    6. The UNIMP instruction has been added to the supported Sparc instruction
121       set. (This is needed for implementing the official C calling convention
122       on this architecture.)
123    7. I fixed some of the .cm files under src/MLRISC/Tools to make them
124       work with the latest CM.
125    
126    ----------------------------------------------------------------------
127    Name: Matthias Blume
128    Date: 2001/06/05 15:10:00 EDT
129    Tag: blume-20010605-cm-index
130    Description:
131    
132    0. The "lambdasplit" parameter for class "sml" in CM has been documented.
133    
134    1. CM can now generate "index files".  These are human-readable files
135       that list on a per-.cm-file basis each toplevel symbol defined or
136       imported.  The location of the index file for
137       <p>/<d>.cm is <p>/CM/INDEX/<d>.cm.
138       To enable index-file generation, set CM.Control.generate_index to true
139       or export an environment-symbol: export CM_GENERATE_INDEX=true.
140    
141       The CM manual has been updated accordingly.
142    
143    2. I made some slight modifications to the c-calls API in MLRISC.
144    
145         a) There is now a callback to support saving/restoring of
146            dedicated but caller-save registers around the actual call
147            instruction.
148         b) One can optionally specify a comment-annotation for the
149            call instruction.
150    
151    3. SML/NJ (mlriscGen.sml) uses this new API for the rawccall primop.
152       (For example, the comment annotation shows the C prototype of
153        the function being called.)
154    
155    ----------------------------------------------------------------------
156    Name: Matthias Blume
157    Date: 2001/06/01 13:30:00 EDT
158    Tag: blume-20010601-nlffi-cleanup
159    Description:
160    
161    This is mostly a cleanup of MLFFI stuff:
162    
163       - some signature files have been put into a more exposed place
164       - the ugly 'f type parameter is gone (simplifies types tremendously!)
165       - ml-nlffigen changed accordingly
166       - tutorial updated
167    
168    Other changes:
169    
170       - author's affiliation in CM manual(s) updated
171       - some more recognized keywords added to Allen's sml.sty
172    
173    ----------------------------------------------------------------------
174    Name: Matthias Blume
175    Date: 2001/05/25 15:30:00 EDT
176    Tag: blume-20010525-iptr
177    Description:
178    
179      - put the official 110.33-README (as it appears on the ftp server) under
180        CVS
181      - fixed a small bug related to incomplete pointer types in
182        ml-nlffigen
183      - small cosmetic change to the ml-nlffi-lib's "arr" type constructor
184        (it does not need the 'f type parameter)
185    
186    ----------------------------------------------------------------------
187    Name: Matthias Blume
188    Date: 2001/05/23 14:30:00 EDT
189    Tag: Release_110_33
190    Description:
191    
192    New version number (110.33).  New bootfiles.
193    
194    ----------------------------------------------------------------------
195    Name: Matthias Blume
196    Date: 2001/05/22 18:06:00 EDT
197    Tag: blume-20010522-targets
198    Description:
199    
200    Made install.sh use file config/targets.customized if it exists, falling
201    back to config/targets if it doesn't.  This way one can have a customized
202    version of the targets file without touching the "real thing", thus
203    eliminating the constant fear of accidentally checking something bogus
204    back into the CVS repository...  (File config/targets.customized must
205    not be added to the repository!)
206    
207    ----------------------------------------------------------------------
208    Name: Matthias Blume
209    Date: 2001/05/22 16:30:00 EDT
210    Tag: blume-20010522-minitut
211    Description:
212    
213    1. Bug fix in ml-nlffigen; now (hopefully) correctly handling
214       struct returns.
215    2. Added src/ml-nlffi-lib/Doc/mini-tutorial.txt.  This is some very
216       incomplete, preliminary documentation for NLFFI.
217    
218    ----------------------------------------------------------------------
219    Name: Matthias Blume
220    Date: 2001/05/14 11:30:00 EDT
221    Tag: blume-20010514-script
222    Description:
223    
224    Some bugs in install script fixed.
225    
226    In addition to that I also made a slight change to the NLFFI API:
227    Functors generated by ml-nlffigen now take the dynamic library as a
228    straight functor argument, not as a suspended one.  (The original
229    functor code used to force the suspension right away anyway, so there
230    was nothing gained by this complication of the interface.)
231    
232    ----------------------------------------------------------------------
233    Name: Matthias Blume
234    Date: 2001/05/11 14:35:00 EDT
235    Tag: blume-20010511-ml-nlffi
236    Description:
237    
238    I finally took the plunge and added my new FFI code to the main
239    repository.  For x86-linux it is now ready for prime-time.
240    
241    There are two new subdirectories of "src":
242    
243      - ml-nlffi-lib:
244           The utility library for programs using the FFI interface.
245           Here is the implementation of $/c.cm and its associated low-level
246           partners $/c-int.cm and $/memory.cm.
247      - ml-nlffigen:
248           A stand-alone program for generating ML glue code from C source
249           code.
250    
251    Building ml-nlffigen requires $/ckit-lib.cm.
252    
253    The config/install.sh script has been updates to do the Right Thing
254    (hopefully).
255    
256    Notice that the source tree for the C-Kit will not be put under "src"
257    but directly under the installation root directory.  (This is the
258    structure that currently exists on the CVS server when you check out
259    module "sml".)  Fortunately, config/install.sh knows about this oddity.
260    
261    Bugs: No documentation yet.
262    
263    ----------------------------------------------------------------------
264    Name: Matthias Blume
265    Date: 2001/05/09 16:35:00 EDT
266    Tag: blume-20010509-cpscontract
267    Description:
268    
269    Fixed a bug in the accounting code in cpsopt/contract.sml.  (The
270    wrapper/unwrapper elimination did not decrement usage counts and some
271    dead variables got overlooked by the dead-up logic.)
272    
273    ----------------------------------------------------------------------
274    Name: Lal George
275    Date: 2001/05/08  17:26:09 EDT
276    Tag: george-20010508-omit-frameptr
277    Description:
278    
279    Changes to implement the omit-frame-pointer optimization to support
280    raw C calls. For now, there is only support on the Intel x86, but
281    other architectures will follow as more experience is gained with this.
282    
283    
284    ----------------------------------------------------------------------
285    Name: Matthias Blume
286    Date: 2001/05/07 14:40:00 EDT
287    Tag: blume-20010507-proxies
288    Description:
289    
290    I made into "proxy libraries" all libraries that qualify for such a
291    change.  (A qualifying library is a library that has another library or
292    groups as its sole member and repeats that member's export list
293    verbatim.  A proxy library avoids this repetition by omitting its export
294    list, effectively inheriting the list that its (only) member exports.
295    See the CM manual for more explanation.)
296    The main effect is that explicit export lists for these libraries
297    do not have to be kepts in sync, making maintenance a bit easier.
298    
299    I also added copyright notices to many .cm-files.
300    
301    Last but not least, I made a new set of bootfiles.
302    
303    ----------------------------------------------------------------------
304    Name: Matthias Blume
305    Date: 2001/05/04 17:00:00 EDT
306    Tag: blume-20010504-cm-lsplit
307    Description:
308    
309    0. John merged pending changes to $/smlnj-lib.cm
310    
311    1. Allen's previous change accidentally backed out of one of Lal's
312       earlier changes.  I undid this mistake (re-introducing Lal's change).
313    
314    2. I used the new topOrder' function from graph-scc.sml (from $/smlnj-lib.cm)
315       within the compiler where applicable.  There is some code simplification
316       because of that.
317    
318    3. The "split" phase (in FLINT) is now part of the default list of phases.
319       Compiler.Control.LambdaSplitting.* can be used to globally control the
320       lambda-splitting (cross-module-inlining) engine.  In addition to that,
321       it can now also be controlled on a per-source basis: CM has been taught
322       a new tool parameter applicable to ML source files.
323    
324       - To turn lambda-splitting off completely:
325            local open Compiler.Control.LambdaSplitting in
326                val _ = set Off
327            end
328       - To make "no lambda-splitting" the global default (but allow per-source
329         overriding); this is the initial setting:
330            local open Compiler.Control.LambdaSplitting in
331                val _ = set (Default NONE)
332            end
333       - To make "lambda-splitting with aggressiveness a" the global default
334         (and allow per-source overriding):
335            local open Compiler.Control.LambdaSplitting in
336                val _ = set (Default (SOME a))
337            end
338    
339       - To turn lambda-splitting off for a given ML souce file (say: a.sml)
340         write (in the respective .cm-file):
341            a.sml (lambdasplitting:off)
342       - To turn lambda-splitting for a.sml on with minimal aggressiveness:
343            a.sml (lambdasplitting:on)
344       - To turn lambda-splitting for a.sml on with aggressiveness <a> (where
345         <a> is a decimal non-negative integer):
346            a.sml (lambdasplitting:<a>)
347       - To turn lambda-splitting for a.sml on with maximal aggressiveness:
348            a.sml (lambdasplitting:infinity)
349       - To use the global default for a.sml:
350            a.sml (lambdasplitting:default)
351         or simply
352            a.sml
353    
354    ----------------------------------------------------------------------
355    Name: Allen Leung
356    Date: 2001/05/04 01:57:00 EDT
357    Tag: leunga-20010504-sync
358    Description:
359    
360      MLRISC features.
361    
362      1. Fix to CMPXCHG instructions.
363      2. Changed RA interface to allow annotations in callbacks.
364      3. Added a new method to the stream interface to allow annotations updates.
365    
366    ----------------------------------------------------------------------
367    Name: Matthias Blume
368    Date: 2001/05/01 11:45:00 EDT
369    Tag: blume-20010501-pcedittmp
370    Description:
371    
372    Changed install.sh to use the current working directory instead of
373    /usr/tmp for a temporary file (pcedittmp).  The previous choice
374    of /usr/tmp caused trouble with MacOS X because of file premission
375    problems.
376    
377    ----------------------------------------------------------------------
378    Name: Matthias Blume
379    Date: 2001/04/20 11:10:00 EDT
380    Tag: blume-20010420-inMLflag
381    Description:
382    
383     - added vp_limitPtrMask to vproc-state.h
384       (for use by the raw-C-calls mechanism to implement proper interrupt
385        handling)
386     - made the ML compiler aware of various data-structure offsets so it
387       can generate code for accessing the vp_inML flag and vp_limitPtrMask
388     - tweaked mlriscGen.sml to have it emit interrupt-handling code for
389       raw C-calls
390    
391    ----------------------------------------------------------------------
392    Name: Lal George
393    Date: 2001/04/20 09:15:28 EDT
394    Tag: george-20010420-macosX
395    Description:
396    
397     - Changes to port to Mac OS X; Darwin.
398    
399     - In the process I found that sqrt was broken on the PPC, because the
400       fsqrt instruction is not implemented.
401    
402    ----------------------------------------------------------------------
403    Name: Matthias Blume
404    Date: 2001/04/18 12:45:00 EDT
405    Tag: blume-20010418-ccalls
406    Description:
407    
408     - fixed two off-by-4 errors in the x86-specific c-calls implementation
409       (this bug prevented structure arguments containing pointers from being
410        passed correctly)
411     - changed the raw-C-call code in mlriscGen.sml in such a way that
412       structure arguments are represented as a pointer to the beginning
413       of the structure (instead of having a series of synthesized arguments,
414       one for each structure member)
415    
416     - made makeml script's verbosity level configurable via environment
417       variable (MAKEML_VERBOSITY)
418    
419     - eliminated placeholder implementations for f32l, w16s, i16s, and f32s
420       in rawmem-x86.sml; we are now using the real thing
421    
422    ----------------------------------------------------------------------
423    Name: Matthias Blume
424    Date: 2001/03/22 16:25:00 EST
425    Tag: blume-20010322-bootfiles
426    Description:
427    
428    Created a new set of bootfiles (for your automatic installation convenience).
429    
430    ----------------------------------------------------------------------
431    Name: Matthias Blume
432    Date: 2001/03/22 15:10:00 EST
433    Tag: blume-20010322-rawmem-parcm
434    Description:
435    
436    1. All "raw memory access" primitives for the new FFI are implemented now
437       (at least on the x86).
438    2. Some further cleanup of CM's parallel make mechanism.
439    
440    ----------------------------------------------------------------------
441    Name: Matthias Blume
442    Date: 2001/03/19 17:53:00 EST
443    Tag: blume-20010319-parallel
444    Description:
445    
446    Parallel make (using compile servers) now works again.
447    
448    To this end, CM.stabilize and CMB.make have been modified to work in
449    two passes when compile servers are attached:
450       1. Compile everything, do not perform stabilization; this pass
451          uses compile servers
452       2. Stabilize everything; this pass does not use compile servers
453    If there are no compile servers, the two passes are combined into one
454    (as before).  Splitting the passes increases the inherent parallelism
455    in the dependency graph because the entire graph including all
456    libraries is available at the same time.  This, in turn, improves
457    server utilization.  The downside is that the master process will
458    have to do some extra work after compilation is done (because for
459    technical reasons it must re-read all the binfiles during stabilization).
460    
461    ----------------------------------------------------------------------
462    Name: Matthias Blume
463    Date: 2001/03/16 12:22:00 EST
464    Tag: blume-20010316-bootfiles
465    Description:
466    
467    Created a new set of bootfiles (for your automatic installation convenience).
468    
469    ----------------------------------------------------------------------
470    Name: Matthias Blume
471    Date: 2001/03/16 11:00:00 EST
472    Tag: blume-20010316-MLTREE-fixup
473    Description:
474    
475    This is a minor fixup for an (untagged) earlier commit by Allen.
476    (A file was missing).
477    
478    ----------------------------------------------------------------------
479    Name: Allen Leung
480    Date: Mon Mar  5 18:54:57 EST 2001
481    Tag: leunga-20010305-cut-support
482    
483    1. New support for alternative control-flow in MLTREE.
484       Currently we support
485    
486          FLOW_TO(CALL ...., [k1,...,kn])
487    
488       This is needed for 'cuts to' in C-- and try/handle-like constructs
489       in Moby
490    
491       New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
492    
493    2. Register Allocator
494    
495       Changes in interface [from Fermin, John]
496    
497    3. Alpha 8-bit SLL support [Fermin]
498    
499    4. All architectures
500    
501       A new module (ClusterExpandCopies) for expanding parallel copies.
502    
503    ----------------------------------------------------------------------
504    Name: Allen Leung
505    Date: 2001/02/27 23:07:00 EST
506    Tag: leunga-20010227-minor-stuff
507    
508    1. Alpha bug fix for CMOVNE
509    2. Handle mltree COND(..,FCMP ...,...)
510    3. Bug fix in simplifier
511    
512    ----------------------------------------------------------------------
513    Name: Matthias Blume
514    Date: 2001/01/30 17:50:00 EST
515    Tag: blume-20010130-sync
516    Description:
517    
518    This is just a minor update to sync my devel branch with the main brach.
519    The only visible change is the addition of some README files.
520    
521    ----------------------------------------------------------------------
522    Name: Matthias Blume
523    Date: 2001/01/12 23:30:00 JST
524    Tag: blume-20010112-bootfiles
525    Description:
526    
527    Made a new set of bootfiles that goes with the current state of the
528    repository.
529    
530    ----------------------------------------------------------------------
531    Name: Matthias Blume
532    Date: 2001/01/12 21:20:00 JST
533    Tag: blume-20010112-sync
534    Description:
535    
536    I am just flushing out some minor changes that had accumulated in
537    my private branch in order to sync with the main tree.  (This is
538    mainly because I had CVS trouble when trying to merge _into_ my
539    private branch.)
540    
541    Most people should be completely unaffected by this.
542    
543    ----------------------------------------------------------------------
544    Name: Allen Leung
545    Date: Thu Jan 11 21:03:00 EST 2001
546    Tag: leunga-20010111-labexp=mltree
547    Description:
548    
549    1.  Removed the type LabelExp and replace it by MLTree.
550    2.  Rewritten mltree-simplify with the pattern matcher tool.
551    3.  There were some bugs in alpha code generator which would break
552        64-bit code generation.
553    4.  Redo the tools to generate code with the
554    5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
555        are now generated by perl scripts.
556    
557    ----------------------------------------------------------------------
558    Name: Matthias Blume
559    Date: 2001/01/10 21:55:00 JST
560    Tag: blume-20010110-rcc
561    Description:
562    
563    The RCC stuff now seems to work (but only on the x86).
564    This required hacking of the c-calls interface (and -implementation) in
565    MLRISC.
566    
567    Normal compiler users should be unaffected.
568    
569    ----------------------------------------------------------------------
570    Name: Matthias Blume
571    Date: 2001/01/09 01:20:00 JST
572    Tag: blume-20010109-rcc
573    Description:
574    
575    This is a fairly big patch, flushing out a large number of pending
576    changes that I made to my development copy over the last couple of days.
577    
578    Of practical relevance at this moment is a workaround for a pickling
579    bug that Allen ran into the other day.  The cause of the bug itself is
580    still unknown and it might be hard to fix it properly, but the
581    workaround has some merits of its own (namely somewhat reducing pickling
582    overhead for certain libraries).  Therefore, I think this solution should
583    be satisfactory at this time.
584    
585    The rest of the changes (i.e., the vast majority) has to do with my
586    ongoing efforts of providing direct support for C function calls from
587    ML.  At the moment there is a new primop "RAW_CCALL", typing magic
588    in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
589    case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
590    translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
591    handle RCC.
592    
593    The last part (the changes to mlriscGen.sml) are still known to be
594    wrong on the x86 and not implemented on all other architectures.  But
595    the infrastructure is in place. I had to change a few functor
596    signatures in the backend to be able to route the CCalls interface
597    from MLRISC there, and I had to specialize the mltree type (on the
598    x86) to include the necessary extensions. (The extensions themselves
599    were already there and redy to go in MLRISC/x86).
600    
601    Everything should be very happy as soon as someone helps me with
602    mlriscGen.sml...
603    
604    In any case, nothing of this should matter to anyone as long as the
605    new primop is not being used (which is going to be the case unless you
606    find it where I hid it :). The rest of the compiler is completely
607    unaffected.
608    
609    ----------------------------------------------------------------------
610    Name: Matthias Blume
611    Date: 2001/01/05 00:30:00 JST
612    Tag: blume-20010105-primops
613    Description:
614    
615    Added some experimental support for work that I am doing right now.
616    These changes mostly concern added primops, but there is also a new
617    experimental C library in the runtime system (but currently not enabled
618    anywhere except on Linux/X86).
619    
620    In the course of adding primops (and playing with them), I discovered that
621    Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
622    broken.  (Zhong was very right he labeled this stuff as "major gross hack".)
623    To recover, I made type information in INL_PRIM mandatory and changed
624    prim.sml as well as built-in.sml accordingly.  The InLine structure now
625    has complete, correct type information (i.e., no bottom types).
626    
627    Since all these changes mean that we need new binfiles, I also bumped the
628    version number to 110.32.1.
629    
630    ----------------------------------------------------------------------
631    Name: Matthias Blume
632    Date: 2000/12/30 22:10:00 JST
633    Tag: blume-20001230-various
634    Description:
635    
636    Added proxy libraries for MLRISC and let MLRISC libraries refer
637    to each other using path anchors.  (See CM manual for explanation.)
638    
639    Updated CM documentation.
640    
641    Fixed some bugs in CM.
642    
643    Implemented "proxy" libraries (= syntactic sugar for CM).
644    
645    Added "-quiet" option to makeml and changed runtime system accordingly.
646    
647    Added cleanup handler for exportML to reset timers and compiler stats.
648    
649    ----------------------------------------------------------------------
650    Name: Lal George
651    Date: 2000/12/22 22:22:58 EST 2000
652    Tag: Release_110_32
653    Description:
654    
655            Infinite precision used throughout MLRISC.
656            see MLRISC/mltree/machine-int.sig
657    
658    ----------------------------------------------------------------------
659    Name: Matthias Blume
660    Date: 2000/12/22 23:16:00 JST
661    Tag: blume-20001222-warn
662    Description:
663    
664    Corrected wording and formatting of some CM warning message which I
665    broke in my previous patch.
666    
667    ----------------------------------------------------------------------
668    Name: Matthias Blume
669    Date: 2000/12/22 21:20:00 JST
670    Tag: blume-20001222-anchorenv
671    Description:
672    
673    Fixed CM's handling of anchor environments in connection with CMB.make.
674    
675    ----------------------------------------------------------------------
676    Name: Matthias Blume
677    Date: 2000/12/22 13:15:00 JST
678    Tag: blume-20001222-cleanup
679    Description:
680    
681    Removed src/cm/ffi which does not (and did not) belong here.
682    
683    ----------------------------------------------------------------------
684    Name: Matthias Blume
685    Date: 2000/12/21 23:55:00 JST
686    Tag: blume-20001221-exn
687    Description:
688    
689    Probably most important: CM no longer silently swallows all exceptions
690    in the compiler.
691    Plus: some other minor CM changes.  For example, CM now reports some
692    sizes for generated binfiles (code, data, envpickle, lambdapickle).
693    
694    ----------------------------------------------------------------------
695    Name: Matthias Blume
696    Date: 2000/12/15 00:01:05 JST
697    Tag: blume-20001215-dirtool
698    Description:
699    
700    - "dir" tool added.
701    - improvements and cleanup to Tools structure
702    - documentation updates
703    
704    ----------------------------------------------------------------------
705    Name: Allen Leung
706    Date: Thu Dec 14 03:45:24 EST 2000
707    Description:
708    Tag:  leunga-20001214-int-inf
709    Description:
710    
711       In IntInf, added these standard functions, which are missing from our
712    implementation:
713    
714        andb : int * int -> int
715        xorb : int * int -> int
716        orb  : int * int -> int
717        notb : int -> int
718         <<   : int * word -> int
719        ~>>  : int * word -> int
720    
721       Not tested, I hope they are correct.
722    
723    ----------------------------------------------------------------------
724    Name: Allen Leung
725    Date: Fri Dec  8 19:23:26 EST 2000
726    Description:
727    Tag:  leunga-20001208-nowhere
728    Description:
729    
730      Slight improvements to the 'nowhere' tool to handle OR-patterns,
731    to generate better error messages etc.  Plus a brief manual.
732    
733    ----------------------------------------------------------------------
734    Name: Lal George
735    Date: 2000/12/08 09:54:02 EST 2000
736    Tag: Release_110_31
737    Description:
738    
739    - Version 110.31
740    ----------------------------------------------------------------------
741    Name: Allen Leung
742    Date: Thu Dec  7 22:01:04 EST 2000
743    Tag:  leunga-20001207-cell-monster-hack
744    Description:
745    
746    Major MLRISC internal changes.  Affect all clients.
747    Summary:
748    
749    1.  Type CELLS.cell = int is now replaced by a datatype.
750        As a result, the old regmap is now gone.  Almost all interfaces
751        in MLRISC change as a consequence.
752    
753    2.  A new brand version of machine description tool (v3.0) that generates
754        modules expecting the new interface.  The old version is removed.
755    
756    3.  The RA interface has been further abstracted into two new functors.
757        RISC_RA and X86RA.  These functors have much simpler interfaces.
758        [See also directory MLRISC/demo.]
759    
760    4.  Some other new source->source code generation tools are available:
761    
762        a. MLRISC/Tools/RewriteGen -- generate rewriters from rules.
763        b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules.
764           I use this tool to generate the peephole optimizers---with the new
765           cell type changes, peephole rules are becoming difficult to write
766           without conditional pattern matching.
767    
768    5.  More Intmap -> IntHashTable change.  Previous changes by Matthias didn't
769        cover the entire MLRISC source tree so many things broke.
770    
771    6.  CM files have been moved to the subdirectory MLRISC/cm.
772        They are moved because there are a lot of them and they clutter up the
773        root dir.
774    
775    7.  More detailed documentation to come...
776    
777        NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do
778        a makeml -rebuild first.  This is because of other other
779        changes that Matthias has made (see below).
780    
781    
782    ----------------------------------------------------------------------
783    Name: Matthias Blume
784    Date: 2000/11/30 23:12:00 JST
785    Tag: blume-20001130-filereorg
786    Description:
787    
788    Some manual updates and some file reorganizations in CM.
789    
790    ----------------------------------------------------------------------
791    Name: Matthias Blume
792    Date: 2000/11/24 17:45:00 JST
793    Tag: blume-20001124-link
794    Description:
795    
796    Drastically improved link traversal code for the case that the dynamic
797    value was already loaded at bootstrap time.  As a result, CM and CMB
798    now both load blazingly fast -- even on a very slow machine.  Also,
799    memory consumption has been further reduced by this.
800    
801    Warning: The format of the PIDMAP file has changed.  THerefore, to
802    bootstrap you have to do this:
803    
804    1. Run CMB.make
805    2. Make a symbolic link for the boot directory:
806         ln -s sml.boot.ARCH-OS xxx
807    3. "Rebuild" the boot directory:
808         ./makeml -boot xxx -rebuild sml ; rm xxx
809    4. Boot normally:
810          ./makeml
811    
812    ----------------------------------------------------------------------
813    Name: Matthias Blume
814    Date: 2000/11/21 21:20:00 JST
815    Tag: blume-20001121-tools
816    Description:
817    
818    Continued hacking on autoloading problem -- with success this time.
819    Also changed tool-plugin mechanism.  See new CM manual.
820    
821    ----------------------------------------------------------------------
822    Name: Matthias Blume
823    Date: 2000/11/19 14:30:00 JST
824    Tag:  blume-20001119-autoload
825    Description:
826    
827    Some hacking to make autoloading faster.  Success for CMB, no success
828    so far for CM.  There is a reduced structure CM' that autoloads faster.
829    (This is a temporary, non-documented hack to be eliminated again when
830    the general problem is solved.)
831    
832    ----------------------------------------------------------------------
833    Name: Matthias Blume
834    Date: 2000/11/17 14:10:00 JST
835    Tag: blume-20001117-pickle-lib
836    Description:
837    
838    1. Eliminated comp-lib.cm
839    2. Made pickle-lib.cm
840    3. Eliminated all uses of intset.sml (from comp-lib.cm)
841    4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
842       equivalent constructs from smlnj-lib.cm (INtHashTable).
843    5. Point 4. also goes for those uses of intmap.* in MLRISC.
844       Duplicated intmap modules thrown out.
845    6. Hunted down all duplicated SCC code and replaced it with
846       equivalent stuff (GraphSCCFn from smlnj-lib.cm).
847    7. Rewrote Feedback module.
848    8. Moved sortedlist.sml into viscomp-lib.cm.  Eventually it
849       should be thrown out and equivalent modules from smlnj-lib.cm
850       should be used (IntRedBlackSet, IntListSet, ...).
851    
852    Confirmed that compiler compiles to fixpoint.
853    
854    ----------------------------------------------------------------------
855    Name: Allen Leung
856    Date: 2000/11/10 18:00:00
857    Tag: leunga-20001110-new-x86-fp
858    
859    A new x86 floating point code generator has been added.
860    By default this is turned off.  To turn this on, do:
861    
862        CM.autoload "$smlnj/compiler.cm";
863        Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
864    
865    Changes:
866    
867    1.  Changed FTAN to FPTAN so that the assembly output is correct.
868    2.  Changed the extension callback for FTANGENT to generate:
869    
870              fptan
871              fstp  %st(0)
872        instead of
873              fptan
874              fstpl ftempmem
875    
876    3.  Numerous assembly fixes for x86.
877    
878    5.  Cleaned up the machine code output module x86/x86MC.sml and added
879        support for a whole bunch of instructions and addressing modes:
880    
881          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st, %st(n)
882          faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp  %st, %st(n)
883          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st(n), %st
884          fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
885          fxch %st(n)
886          fld %st(n)
887          fst %st(n)
888          fst mem
889          fstp %st(n)
890          fucom %st(n)
891          fucomp %st(n)
892    
893        All these are now generated when the fast fp mode is turned on.
894    
895    6.  Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
896    
897    ----------------------------------------------------------------------
898    Name: Matthias Blume
899    Date: 2000/11/09 11:20:00 JST
900    Tag: blume-20001109-scc
901    Description:
902    
903    Eliminated some code duplication:
904    
905    1. Added "where" clause to GraphSCCFn in SML/NJ Library.
906       (Otherwise the functor is useless.)
907    2. Used GraphSCCFn where SCCUtilFun was used previously.
908    3. Got rid of SCCUtilFun (in comp-lib.cm).
909    
910    ----------------------------------------------------------------------
911    Name: Lal George
912    Date: 2000/11/06 09:02:21 EST 2000
913    Tag: Release_110_30
914    Description:
915    
916    - Version 110.30
917    ----------------------------------------------------------------------
918    Name: Matthias Blume
919    Date: 2000/11/04 14:45:00
920    Tag: blume-20001104-mlbuild
921    Description:
922    
923    - Made ml-build faster on startup.
924    - Documentation fixes.
925    
926    ----------------------------------------------------------------------
927    Name: Matthias Blume
928    Date: 2000/11/02 17:00:00 JST
929    Tag: blume-20001102-condcomp
930    Description:
931    
932    - Small tweaks to pickler -- new BOOTFILES!
933    - Version bumped to 110.29.2.
934    - Added conditional compilation facility to init.cmi (see comment there).
935    ----------------------------------------------------------------------
936    Name: Allen Leung
937    Date: 2000/10/23 19:31:00
938    Tag: leunga-20001023-demo-ra
939    
940    1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
941    2. Test programs for the graph library updated
942    3. Some new MLRISC demo programs added
943    
944    ----------------------------------------------------------------------
945    Name: Matthias Blume
946    Date: 2000/08/31 22:15:00 JST
947    Tag: blume-20001017-errmsg
948    Description:
949    
950    More error message grief: Where there used to be no messages, there
951    now were some that had bogus error regions.  Fixed.
952    
953    ----------------------------------------------------------------------
954    Name: Matthias Blume
955    Date: 2000/08/31 17:30:00 JST
956    Tag: blume-20001017-v110p29p1
957    Description:
958    
959    I made a version 110.29.1 with new bootfiles.
960    
961    Changes:  Modified pickler/unpickler for faster and leaner unpickling.
962              CM documentation changes and a small bugfix in CM's error reporting.
963    
964    ----------------------------------------------------------------------
965    Name: Lal George
966    Date: 2000/09/27 14:42:35 EDT
967    Tag: george-20000927-nodestatus
968    Description:
969    
970    Changed the type of the nodestatus, so that:
971    
972            SPILLED(~1)             is now SPILLED
973            SPILLED(m) where m>=0   is now MEMREG(m)
974            SPILLED(s) where s<~1   is now SPILL_LOC(~s)
975    
976    ----------------------------------------------------------------------
977    Name: Matthias Blume
978    Date: 2000/09/07 14:45:00 JST
979    Tag: blume-20000907-cmerrmsg
980    Description:
981    
982    Small tweak to CM to avoid getting ML syntax error messages twice.
983    
984    ----------------------------------------------------------------------
985    Name: Matthias Blume
986    Date: 2000/08/31 18:00:00 JST
987    Tag: blume-20000831-cvsbootfiles
988    Description:
989    
990    New URL for boot files (because the 110.29 files on the BL server do
991    now work correctly with my updated install scripts for yacc and lex).
992    
993    ----------------------------------------------------------------------
994    Name: Matthias Blume
995    Date: 2000/08/08 12:33:00 JST
996    Tag: blume-20000808-manual
997    Description:
998    
999    Tiny update to CM manual.
1000    
1001    ----------------------------------------------------------------------
1002    Name: Allen Leung
1003    Date: 2000/08/7 19:31:00
1004    Tag: leunga-20000807-a-whole-bunch-of-stuff
1005    
1006      Moby, C--, SSA, x86, machine descriptions etc.  Should only affect C--
1007    and Mobdy.
1008    
1009    1.  x86
1010    
1011       a.  Fixes to peephole module by John and Dan.
1012       b.  Assembly fix to SETcc by Allen.
1013       c.  Fix to c-call by John.
1014       d.  Fix to spilling by John.  (This one deals with the missing FSTPT case)
1015       e.  Instruction selection optimization to SETcc as suggested by John.
1016    
1017           For example,
1018    
1019            MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
1020    
1021           should generate:
1022    
1023            MOVL a, x
1024            SUBL b, x
1025            SHRL 31, x
1026    
1027    2.  IR stuff
1028    
1029         A bunch of new DJ-graph related algorithms added.  These
1030         speed up SSA construction.
1031    
1032    3.  SSA + Scheduling
1033    
1034         Added code for SSA and scheduling to the repository
1035    
1036    ----------------------------------------------------------------------
1037    Name: Lal George
1038    Date: 2000/07/27 11:53:14 EDT
1039    
1040    Tag: lal-20000727-linux-ppc
1041    Description:
1042    
1043     Made changes to support Linux PPC.
1044     p.s. I have confirmation that the 110.29 boot files work fine.
1045    
1046    ----------------------------------------------------------------------
1047    Name: Matthias Blume
1048    Date: 2000/07/27 17:40:00 JST
1049    Tag: blume-20000727-scripts
1050    Description:
1051    
1052    !!!! WARNING !!!!
1053    You must recompile the runtime system!
1054    !!!! WARNING !!!!
1055    
1056    This is basically another round of script-enhancements:
1057    
1058    1. sml, ml-build, and ml-makedepend accept options -D and -U to define
1059       and undefine CM preprocessor symbols.
1060    
1061    2. ml-build avoids generating a new heap image if it finds that the
1062       existing one is still ok.  (The condition is that no ML file had to
1063       be recompiled and all ML files are found to be older that the heap
1064       file.)
1065    
1066       To make this work smoothly, I also hacked the runtime system as
1067       well as SMLofNJ.SysInfo to get access to the heap image suffix
1068       (.sparc-solaris, ...) that is currently being used.
1069    
1070       Moreover, the signature of CM.mk_standalone has changed.  See the
1071       CM manual.
1072    
1073    3. ml-makedepend accepts additional options -n, -a, and -o.  (See the
1074       CM manual for details.)
1075    
1076    4. More CM manual updates:
1077        - all of the above has been documented.
1078        - there is now a section describing the (CM-related) command line
1079          arguments that are accepted by the "sml" command
1080    
1081    ----------------------------------------------------------------------
1082    Name: Matthias Blume
1083    Date: 2000/07/25 16:20:00 JST
1084    Tag: blume-20000725-makedepend
1085    Description:
1086    
1087    Added a script called ml-makedepend.  This can be used in makefiles
1088    for Unix' make in a way very similar to the "makedepend" command for
1089    C.
1090    
1091    The script internally uses function CM.sources.
1092    
1093    Synopsis:
1094    
1095        ml-makedepend [-f makefile] cmfile targetname
1096    
1097    The default for the makefile is "makefile" (or "Makefile" should
1098    "makefile" not exist).
1099    
1100    ml-makedepend adds a cmfile/targetname-specific section to this
1101    makefile (after removing the previous version of this section).  The
1102    section contains a single dependency specification with targetname on
1103    the LHS (targetname is an arbitrary name), and a list of files derived
1104    from the cmfile on the RHS.  Some of the files on the RHS are
1105    ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
1106    "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
1107    path names.  The makefile writer is responsible for making sure that
1108    these variables have correct at the time "make" is invoked.
1109    
1110    ----------------------------------------------------------------------
1111    Name: Matthias Blume
1112    Date: 2000/07/22 23:30:00 JST
1113    Tag: blume-20000722-urlupdate
1114    Description:
1115    
1116    Changed BOOT and config/srcarchiveurl to point to BL server:
1117    
1118        ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
1119    
1120    ----------------------------------------------------------------------
1121    Name: Matthias Blume
1122    Date: 2000/07/18 18:00:00 JST
1123    Tag: blume-20000718-Version_110_29
1124    Description:
1125    
1126    1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
1127    
1128    2. Updated config/version to 110.29
1129    
1130    3. Updated config/srcarchiveurl
1131    
1132    3. New boot files!
1133       ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
1134    
1135    ----------------------------------------------------------------------
1136    Name: Matthias Blume
1137    Date: 2000/07/11 13:58:00 JST
1138    Tag: blume-20000711-doctypo
1139    Description:
1140    
1141    Fixed a few typos in CM manual.
1142    
1143    ----------------------------------------------------------------------
1144    Name: Allen Leung
1145    Date: 2000/06/15 00:38:00
1146    Tag: leunga-20000704-sparc-x86
1147    
1148    1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
1149    2. fix to x86 RET bug [found by Dan Grossman]
1150    3. sparc assembly bug fix for ticc instructions [found by Fermin]
1151    
1152       Affects c-- and moby only
1153    
1154    ----------------------------------------------------------------------
1155    Name: Matthias Blume
1156    Date: 2000/07/04 15:26:00
1157    Tag: blume-20000704-trigger
1158    Description:
1159    
1160    1. Improvements to CM manual.
1161    2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
1162       of getting a back-trace.  The function, when called, raises an
1163       internal exception which explicitly carries the full back-trace history,
1164       so it is unaffected by any intervening handle-raise pairs ("trivial"
1165       or not).  The interactive loop will print that history once it arrives
1166       at top level.
1167       Short of having all exceptions implicitly carry the full history, the
1168       recommended way of using this facility is:
1169         - compile your program with instrumentation "on"
1170         - run it, when it raises an exception, look at the history
1171         - if the history is "cut off" because of some handler, go and modify
1172           your program so that it explicitly calls BTrace.trigger
1173         - recompile (still instrumented), and rerun; look at the full history
1174    
1175    ----------------------------------------------------------------------
1176    Name: Matthias Blume
1177    Date: 2000/07/03 15:36:00 JST
1178    Tag: blume-20000702-manual
1179    Description:
1180    
1181    Small corrections and updates to CM manual.
1182    
1183    ----------------------------------------------------------------------
1184    Name: Matthias Blume
1185    Date: 2000/06/29 16:04:00 JST
1186    Tag: blume-20000629-yacctool
1187    Description:
1188    
1189    Changes:
1190    
1191    1. Class "mlyacc" now takes separate arguments to pass options to
1192       generated .sml- and .sig-files independently.
1193    2. Corresponding CM manual updates.
1194    3. BTrace module now also reports call sites.  (However, for loop clusters
1195       it only shows from where the cluster was entered.)  There are associated
1196       modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
1197    
1198    ----------------------------------------------------------------------
1199    Name: Matthias Blume
1200    Date: 2000/06/27 16:51:00 JST
1201    Tag: blume-20000627-noweb
1202    Description:
1203    
1204    Changes:
1205    
1206     1. Implemented "subdir" and "witness" options for noweb tool.
1207        This caused some slight internal changes in CM's tool implementation.
1208     2. Fixed bug in "tool plugin" mechanism.  This is essentially cleaning
1209        some remaining issues from earlier path anchor changes.
1210     3. Updated CM manual accordingly.
1211    
1212     4. Changed implementation of back-tracing so that I now consider it
1213        ready for prime-time.
1214    
1215        In particular, you don't have to explicitly trigger the back-trace
1216        anymore.  Instead, if you are running BTrace-instrumented code and
1217        there is an uncaught exception (regardless of whether or not it was
1218        raised in instrumented code), the top-level evalloop will print
1219        the back-trace.
1220    
1221        Features:
1222    
1223          - Instrumented and uninstrumented code work together seemlessly.
1224            (Of course, uninstrumented code is never mentioned in actual
1225             back-traces.)
1226    
1227          - Asymptotic time- and space-complexity of instrumented code is
1228            equal to that of uninstrumented code.  (This means that
1229            tail-recursion is preserved by the instrumentation phase.)
1230    
1231          - Modules whose code has been instrumented in different sessions
1232            work together without problem.
1233    
1234          - There is no penalty whatsoever on uninstrumented code.
1235    
1236          - There is no penalty on "raise" expressions, even in
1237            instrumented code.
1238    
1239        A potential bug (or perhaps it is a feature, too):
1240    
1241          A back-trace reaches no further than the outermost instrumented
1242          non-trivial "raise".  Here, a "trivial" raise is one that is the
1243          sole RHS of a "handle" rule.  Thus, back-traces reach trough
1244    
1245               <exp> handle e => raise e
1246    
1247          and even
1248    
1249               <exp> handle Foo => raise Bar
1250    
1251          and, of course, through
1252    
1253               <exp> handle Foo => ...
1254    
1255         if the exception was not Foo.
1256    
1257         Back-traces always reach right through any un-instrumented code
1258         including any of its "handle" expressions, trivial or not.
1259    
1260       To try this out, do the following:
1261    
1262         - Erase all existing binfiles for your program.
1263           (You may keep binfiles for those modules where you think you
1264            definitely don't need back-tracing.)
1265         - Turn on back-trace instrumentation:
1266              SMLofNJ.Internals.BTrace.mode (SOME true);
1267         - Recompile your program.  (I.e., run "CM.make" or "use".)
1268         - You may now turn instrumentation off again (if you want):
1269              SMLofNJ.Internals.BTrace.mode (SOME false);
1270         - Run your program as usual.  If it raises an exception that
1271           reaches the interactive toplevel, then a back-trace will
1272           automatically be printed.  After that, the toplevel loop
1273           will print the exception history as usual.
1274    
1275    ----------------------------------------------------------------------
1276    Name: Matthias Blume
1277    Date: 2000/06/26 09:56:46 JST
1278    Tag: blume-20000626-setup
1279    Description:
1280    
1281    CM: - setup-parameter to "sml" added; this can be used to run arbitrary
1282          ML code before and after compiling a file (e.g., to set compiler
1283          flags)
1284    
1285    Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
1286              - associated changes to btrace.sml (BTrace instrumentation pass)
1287              - cleaner implementation of btimp.sml (BTrace tracing and report
1288                module)
1289    
1290    CM manual: * new path encoding documented
1291               * description of setup-parameter to "sml" added
1292    
1293    The biggest user-visible change to back-tracing is that it is no
1294    longer necessary to compile all traced modules within the same
1295    session.  (This was a real limitation.)
1296    
1297    ----------------------------------------------------------------------
1298    Name: Matthias Blume
1299    Date: 2000/06/24 12:40:00 JST
1300    Tag: blume-20000624-startup
1301    Description:
1302    
1303    Fixes startup slowdown problem.  (I was calling SrcPath.sync a _tad_
1304    bit too often -- to put it mildly. :)
1305    
1306    ----------------------------------------------------------------------
1307    Name: Matthias Blume
1308    Date: 2000/06/23 18:20:00 JST
1309    Tag: blume-20000623-btrace
1310    Description:
1311    
1312    This updates adds a backtrace facility to aid programmers in debugging
1313    their programs.  This involves the following changes:
1314    
1315    1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
1316       keeping track of the current call stack.  When programs are compiled
1317       in a special mode, the compiler will insert calls to these hooks
1318       into the user program.
1319       "Hook" means that it is possible for different implementations of
1320       back-tracing to register themselves (at different times).
1321    
1322    2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
1323       as an Absyn.dec->Absyn.dec rewrite.  Normally this phase is turned off.
1324       It can be turned on using this call:
1325         SMLofNJ.Internals.BTrace.mode (SOME true);
1326       Turning it off again:
1327         SMLofNJ.Internals.BTrace.mode (SOME false);
1328       Querying the current status:
1329         SMLofNJ.Internals.BTrace.mode NONE;
1330       Annotated programs are about twice as big as normal ones, and they
1331       run a factor of 2 to 4 slower with a dummy back-trace plugin (one
1332       where all hooks do nothing).  The slowdown with a plugin that is
1333       actually useful (such as the one supplied by default) is even greater,
1334       but in the case of the default plugin it is still only an constant
1335       factor (amortized).
1336    
1337    3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
1338       with a sub-structure BTrace for controlling back-tracing.  In particular,
1339       the above-mentioned function "mode" controls whether the annotation
1340       phase is invoked by the compiler.  Another important function is
1341       "trigger": when called it aborts the current execution and causes
1342       the top-level loop to print a full back-trace.
1343    
1344    4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
1345       for back-tracing.  It keeps track of the dynamic call stack and in
1346       addition to that it keeps a partial history at each "level" of that
1347       stack.  For example, if a tail-calls b, b tail-calls c, and c tail-calls
1348       d and b (at separate times, dynamically), then the report will show:
1349    
1350       GOTO   d
1351             /c
1352       GOTO  \b
1353       CALL   a
1354    
1355       This shows that there was an initial non-tail call of a, then a
1356       tail-call to b or c, looping behavior in a cluster of functions that
1357       consist of b and c, and then a goto from that cluster (i.e., either from
1358       b or from c) to d.
1359    
1360       Note that (depending on the user program) the amount of information
1361       that the back-trace module has to keep track of at each level is bounded
1362       by a constant.  Thus, the whole implementation has the same asymptotical
1363       complexity as the original program (both in space and in time).
1364    
1365    5. compiler/TopLevel/interact/evalloop.sml has been modified to
1366       handle the special exception SMLofNJ.Internals.BTrace.BTrace
1367       which is raised by the "trigger" function mentioned above.
1368    
1369    Notes on usage:
1370    
1371    - Annotated code works well together with unannotated code:
1372    Unannotated calls simply do not show up at all in the backtrace.
1373    
1374    - It is not a good idea to let modules that were annotated during
1375    different sessions run at the same time.  This is because the compiler
1376    chooses small integers to identify individual functions, and there
1377    will be clashes if different modules were compiled in separate sessions.
1378    (Nothing will crash, and you will even be told about the clashes, but
1379    back-trace information will in general not be useful.)
1380    
1381    - Back-tracing can be confused by callcc and capture.
1382    
1383    - The only way of getting a back-trace right now is to explicitly
1384    invoke the "trigger" function from your user program.  Eventually, we
1385    should make every exception carry back-trace information (if
1386    available).  But since this creates more overhead at "raise"-time
1387    (similar to the current exnHistory overhead), I have not yet
1388    implemented this.  (The implementation will be rather easy.)  With
1389    exceptions carrying back-trace information, this facility will be even
1390    more useful because users don't need to modify their programs...
1391    
1392    - While it is possible to compile the compiler with back-trace
1393    annotations turned on (I did it to get some confidence in
1394    correctness), you must make absolutely sure that core.sml and
1395    btimp.sml are compiled WITHOUT annotation!  (core.sml cannot actually
1396    be compiled with annotation because there is no core access yet, but
1397    if you compile btimp.sml with annotation, then the system will go into
1398    an infinite recursion and crash.)
1399    Since CM currently does not know about BTrace, the only way to turn
1400    annotations on and off for different modules of the compiler is to
1401    interrupt CMB.make, change the settings, and re-invoke it.  Of course,
1402    this is awkward and clumsy.
1403    
1404    Sample sessions:
1405    
1406    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
1407    - SMLofNJ.Internals.BTrace.mode (SOME true);
1408    [autoloading]
1409    [autoloading done]
1410    val it = false : bool
1411    - structure X = struct
1412    -     fun main n = let
1413    -         fun a (x, 0) = d x
1414    -           | a (x, n) = b (x, n - 1)
1415    -         and b (x, n) = c (x, n)
1416    -         and c (x, n) = a (x, n)
1417    -         and d x = e (x, 3)
1418    -         and e (x, 0) = f x
1419    -           | e (x, n) = e (x, n - 1)
1420    -         and f 0 = SMLofNJ.Internals.BTrace.trigger ()
1421    -           | f n = n * g (n - 1)
1422    -         and g n = a (n, 3)
1423    -     in
1424    -         f n
1425    -     end
1426    - end;
1427    structure X : sig val main : int -> int end
1428    - X.main 3;
1429    *** BACK-TRACE ***
1430    GOTO   stdIn:4.2-13.20: X.main[2].f
1431    GOTO-( stdIn:4.2-13.20: X.main[2].e
1432    GOTO   stdIn:4.2-13.20: X.main[2].d
1433         / stdIn:4.2-13.20: X.main[2].a
1434         | stdIn:4.2-13.20: X.main[2].b
1435    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1436    CALL   stdIn:4.2-13.20: X.main[2].g
1437    GOTO   stdIn:4.2-13.20: X.main[2].f
1438    GOTO-( stdIn:4.2-13.20: X.main[2].e
1439    GOTO   stdIn:4.2-13.20: X.main[2].d
1440         / stdIn:4.2-13.20: X.main[2].a
1441         | stdIn:4.2-13.20: X.main[2].b
1442    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1443    CALL   stdIn:4.2-13.20: X.main[2].g
1444    GOTO   stdIn:4.2-13.20: X.main[2].f
1445    GOTO-( stdIn:4.2-13.20: X.main[2].e
1446    GOTO   stdIn:4.2-13.20: X.main[2].d
1447         / stdIn:4.2-13.20: X.main[2].a
1448         | stdIn:4.2-13.20: X.main[2].b
1449    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1450    CALL   stdIn:4.2-13.20: X.main[2].g
1451    GOTO   stdIn:4.2-13.20: X.main[2].f
1452    CALL   stdIn:2.15-17.4: X.main[2]
1453    -
1454    
1455    (Note that because of a FLINt bug the above code currently does not
1456    compile without BTrace turned on.)
1457    
1458    Here is another example, using my modified Tiger compiler:
1459    
1460    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
1461    - SMLofNJ.Internals.BTrace.mode (SOME true);
1462    [autoloading]
1463    [autoloading done]
1464    val it = false : bool
1465    - CM.make "sources.cm";
1466    [autoloading]
1467    ...
1468    [autoloading done]
1469    [scanning sources.cm]
1470    [parsing (sources.cm):parse.sml]
1471    [creating directory CM/SKEL ...]
1472    [parsing (sources.cm):tiger.lex.sml]
1473    ...
1474    [wrote CM/sparc-unix/semant.sml]
1475    [compiling (sources.cm):main.sml]
1476    [wrote CM/sparc-unix/main.sml]
1477    [New bindings added.]
1478    val it = true : bool
1479    - Main.compile ("../testcases/merge.tig", "foo.out");
1480    *** BACK-TRACE ***
1481    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
1482    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1483    CALL   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1484    GOTO   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1485    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1486    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1487    CALL   lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
1488         / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
1489    CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
1490    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1491    CALL   lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
1492    CALL   main.sml:1.18-118.4: Main.compile[2]
1493    -
1494    
1495    ----------------------------------------------------------------------
1496    Name: Matthias Blumen
1497    Date: 2000/06/21 18:00:00 JST
1498    Tag: blume-20000621-manual
1499    Description:
1500    
1501    CM manual update: Path environments documented.
1502    
1503    ----------------------------------------------------------------------
1504    Name: Matthias Blume
1505    Date: 2000/06/19 13:40:00
1506    Tag: blume-20000619-manual
1507    Description:
1508    
1509    CM manual and system/README update.  This only covers the fact that
1510    there are no more implicit anchors.  (Path environments and the "bind"
1511    option to "cm" have yet to be documented.)
1512    
1513    ----------------------------------------------------------------------
1514    Name: Matthias Blume
1515    Date: 2000/06/19 11:05:00 JST
1516    Tag: blume-20000619-chdir-bugfix
1517    Description:
1518    
1519    Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
1520    
1521    ----------------------------------------------------------------------
1522    Name: Matthias Blume
1523    Date: 2000/06/18 22:00:10 JST
1524    Tag: blume-20000618-implicit-anchors-really-gone
1525    Description:
1526    
1527    I updates the previous HISTORY entry where I forgot to mention that
1528    implicit anchors are no longer with us.
1529    
1530    The current update also gets rid of the (now useless) controller
1531    CM.Control.implicit_anchors.
1532    
1533    ----------------------------------------------------------------------
1534    Name: Matthias Blume
1535    Date: 2000/06/16 17:30:00 JST
1536    Tag: blume-20000616-anchorenv
1537    Description:
1538    
1539    This patch implements the long anticipated (just kidding :) "anchor
1540    environment" mechanism.  In the course of doing this, I also
1541    re-implemented CM's internal "SrcPath" module from scratch.  The new
1542    one should be more robust in certain boundary cases.  In any case, it
1543    is a lot cleaner than its predecessor (IMHO).
1544    
1545    This time, although there is yet another boot file format change, I
1546    kept the unpickler backward-compatible.  As a result, no new bootfiles
1547    are necessary and bootstrapping is straightforward.  (You cannot read
1548    new bootfiles into an old system, but the other way around is no
1549    problem.)
1550    
1551    Visible changes:
1552    
1553    ** 0. Implicit path anchors (without the leading $-symbol) are no
1554    longer recognized at all. This means that such path names are not
1555    illegal either.  For example, the name basis.cm simply refers to a
1556    local file called "basis.cm" (i.e, the name is an ordinary path
1557    relative to .cm-files directory).  Or, to put it differently, only
1558    names that start with $ are anchored paths.
1559    
1560    ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
1561    vanished.
1562    
1563    John (Reppy) had critizised this as soon as I originally proposed and
1564    implemented it, but at that time I did not really deeply believe
1565    him. :) Now I came full-circle because I need the $<singlearc> syntax
1566    in another place where it cannot be seen as an abbreviation for
1567    $/<singlearc>.  To avoid the confusion, $<singlearc> now means what it
1568    seems to mean (i.e., it "expands" into the corresponding anchor
1569    value).
1570    
1571    However, when paths are used as members in CM description files, it
1572    continues to be true that there must be at least another arc after the
1573    anchor.  This is now enforced separately during semantic analysis
1574    (i.e., from a lexical/syntactical point of view, the notation is ok.)
1575    
1576    ** 2. The "cm" class now accepts an option "bind".  The option's value
1577    is a sub-option list of precisely two items -- one labeled "anchor"
1578    and the other one labeled "value".  As you might expect, "anchor" is
1579    used to specify an anchor name to be bound, and "value" specifies what
1580    the anchor is being bound to.
1581    
1582    The value must be a directory name and can be given in either standard
1583    syntax (including the possibility that it is itself an anchored path)
1584    or native syntax.
1585    
1586    Examples:
1587    
1588       foo.cm (bind:(anchor:bar value:$mystuff/bar))
1589       lib.cm (bind:(anchor:a value:"H:\\x\\y\\z"))  (* only works under windows *)
1590    
1591    and so on.
1592    
1593    The meaning of this is that the .cm-file will be processed with an
1594    augmented anchor environment where the given anchor(s) is/are bound to
1595    the given values(s).
1596    
1597    The rationale for having this feature is this: Suppose you are trying
1598    to use two different (already stable) libraries a.cm and b.cm (that
1599    you perhaps didn't write yourself).  Further, suppose each of these
1600    two libraries internally uses its own auxiliary library $aux/lib.cm.
1601    Normally you would now have a problem because the anchor "lib" can not
1602    be bound to more than one value globally.  Therefore, the project that
1603    uses both a.cm and b.cm must locally redirect the anchor to some other
1604    place:
1605    
1606       a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
1607       b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
1608    
1609    This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
1610    /usr/lib/smlnj/b-stuff/aux.cm, respectively.
1611    
1612    Hard-wiring path names is a bit inflexible (and CM will verbosely warn
1613    you when you do so at the time of CM.stabilize).  Therefore, you can
1614    also use an anchored path as the value:
1615    
1616      a.cm (bind:(anchor:lib value:$a-lib))
1617      b.cm (bind:(anchor:lib value:$b-lib))
1618    
1619    Now you can globally configure (using the usual CM.Anchor.anchor or
1620    pathconfig machinery) bindings for "a-lib" and "b-lib".  Since "lib"
1621    itself is always locally bound, setting it globally is no longer
1622    meaningful or necessary (but it does not hurt either).  In fact, "lib"
1623    can still be used as a global anchor for separate purposes.  As a
1624    matter of fact, one can locally define "lib" in terms of a global
1625    "lib":
1626    
1627      a.cm (bind:(anchor:lib value:$lib/a))
1628      b.cm (bind:(anchor:lib value:$lib/b))
1629    
1630    ** 3: The encoding of path names has changed.  This affects the way
1631    path names are shown in CM's progress report and also the internal
1632    protocol encoding used for parallel make.
1633    
1634    The encoding now uses one or more ':'-separated segments.  Each
1635    segments corresponds to a file that has been specified relative to the
1636    file given by its preceding segment.  The first segment is either
1637    relative to the CWD, absolute, or anchored.  Each segment itself is
1638    basically a Unix pathname; all segments but the first are relative.
1639    
1640    Example:
1641    
1642       $foo/bar/baz.cm:a/b/c.sml
1643    
1644    This path denotes the file bar/a/b/c.sml relative to the directory
1645    denoted by anchor "foo".  Notice that the encoding also includes
1646    baz.cm which is the .cm-file that listed a/b/c.sml.  As usual, such
1647    paths are resolved relative to the .cm-files directory, so baz.cm must
1648    be ignored to get the "real" pathname.
1649    
1650    To make this fact more obvious, CM puts the names of such "virtual
1651    arcs" into parentheses when they appear in progress reports. (No
1652    parentheses will appear in the internal protocol encoding.)  Thus,
1653    what you really see is:
1654    
1655      $foo/bar/(baz.cm):a/b/c.sml
1656    
1657    I find this notation to be much more informative than before.
1658    
1659    Another new feature of the encoding is that special characters
1660    including parentheses, colons, (back)slashes, and white space are
1661    written as \ddd (where ddd is the decimal encoding of the character).
1662    
1663    *** The CM manual still needs to be updated.
1664    
1665    ----------------------------------------------------------------------
1666    Name: Allen Leung
1667    Date: 2000/06/15 00:38:00
1668    Tag: leunga-20000615-x86-peephole
1669    
1670    x86 Peephole fix by Fermin.  Affects c-- and moby only.
1671    
1672    ----------------------------------------------------------------------
1673    Name: Matthias Blume
1674    Date: 2000/06/12 11:40:00
1675    Tag: blume-20000612-parmakefix
1676    Description:
1677    
1678    More cleanup after changing the file naming scheme: This time I
1679    repaired the parallel make mechanism for CMB.make which I broke earlier.
1680    
1681    ----------------------------------------------------------------------
1682    Name: Allen Leung
1683    Date: 2000/06/09 01:25:00
1684    Tag: leunga-20000609-various
1685    
1686    None of these things should affect normal SML/NJ operations
1687    
1688    1. Peephole improvements provided by Fermin (c--)
1689    2. New annotation DEFUSE for adding extra dependence (moby)
1690    3. New X86 LOCK instructions (moby)
1691    4. New machine description language for reservation tables (scheduling)
1692    5. Fixes to various optimization/analysis modules (branch chaining, dominator
1693       trees etc.)
1694    6. I've changed the CM files so that they can work with versions
1695       110.0.6, 110.25 and 110.28
1696    
1697    ----------------------------------------------------------------------
1698    Name: Matthias Blume
1699    Date: 2000/06/09 12:40:00
1700    Tag: blume-20000609-log
1701    Description:
1702    
1703    - Removed all(?) remaining RCS Log entries from sources.
1704    
1705    - Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
1706      anchored paths).
1707    
1708    ----------------------------------------------------------------------
1709    Name: Matthias Blume
1710    Date: 2000/06/07 17:00:00 JST
1711    Tag: blume-20000607-no-implicit-anchors
1712    Description:
1713    
1714    1. This update changes the default setting for
1715    CM.Control.implicit_anchors from true to false.  This means that
1716    implicit anchors are no longer permitted by default.  I also tried to
1717    make sure that nothing else still relies on implicit anchors.
1718    (This is the next step on the schedule towards a CM that does not even
1719    have the notion of implicit anchors anymore.)
1720    
1721    2. More CM manual updates.
1722    
1723    3. I managed to track down and fix the pickling bug I mentioned last
1724    time.  Because of the previously existing workaround, this entails no
1725    immediate practical changes.
1726    
1727    ----------------------------------------------------------------------
1728    Name: Matthias Blume
1729    Date: 2000/06/06 11:15:00 JST
1730    Tag: blume-20000606-lazierpickle
1731    Description:
1732    
1733    !!!! NEW BOOT FILES !!!!
1734    
1735    * The main purpose of this update is to make library pickles lazier in
1736    order to reduce the initial space penalty for autoloading a library.
1737    As a result, it is now possible to have $smlnj/compiler.cm
1738    pre-registered.  This should take care of the many complaints or
1739    inquiries about missing structure Compiler.  This required changes to
1740    CM's internal data structures and small tweaks to some algorithms.
1741    
1742    As a neat additional effect, it is no longer necessary (for the sake
1743    of lean heap image files) to distinguish between a "minimal" CM and a
1744    "full" CM.  Now, there is only one CM (i.e., the "full" version:
1745    $smlnj/cm.cm aka $smlnj/cm/full.cm), and it is always available at the
1746    interactive top level. ($smlnj/cm/minimal.cm is gone.)
1747    
1748    To make the life of compiler-hackers easier, "makeml" now also
1749    pre-registers $smlnj/cmb.cm (aka $smlnj/cmb/current.cm).  In other
1750    words, after you bootstrap a new sml for the first time, you will not
1751    have to autoload $smlnj/cmb.cm again afterwards.  (The first time
1752    around you will still have to do it, though.)
1753    
1754    * A second change consists of major updates to the CM manual.  There
1755    are now several appendices with summary information and also a full
1756    specification of the CM description file syntax.
1757    
1758    * In directory src/system I added the script "allcross".  This script
1759    invokes sml and cross-compiles the compiler for all supported
1760    architectures.  (Useful when providing a new set of boot files.)
1761    
1762    * There seems to be a latent bug in my "lazy pickles" mechanism.  I
1763    added a small tweak to pickle-util.sml to work around this problem,
1764    but it is not a proper fix yet.  I will investigate further.  (The
1765    effect of the bug was an inflation of library pickle size.)
1766    
1767    * Version number increased to 110.28.1 (to avoid compatibility problems).
1768    
1769    ----------------------------------------------------------------------
1770    Name: Allen Leung
1771    Date: 2000/05/25 17:28 EDT
1772    Tag: leunga-20000525-ra
1773    Description:
1774    
1775      Fixed a bug in freezing phase of the register allocator.
1776    
1777    ----------------------------------------------------------------------
1778    Name: Allen Leung
1779    Date: 2000/05/15 22:53 EDT
1780    Tag: leunga-20000515-alpha-x86-ra
1781    Description:
1782    
1783      1. Alpha
1784    
1785          Slight cleanup.  Removed the instruction SGNXL
1786    
1787      2. X86
1788    
1789          Added the following instructions to the instruction set:
1790    
1791            ROLx, RORx,
1792            BTx, BTSx, BTLx, BTRx,
1793            XCHGx, and variants with the LOCK prefix
1794    
1795      3. Register Allocation
1796    
1797          The module ra-rewrite-with-renaming has been improved.
1798    
1799      These have no effect on SML/NJ.
1800    
1801    ----------------------------------------------------------------------
1802    Name: Matthias Blume
1803    Date: 2000/05/15 16:20:00 JST
1804    Tag: blume-20000515-lightrebuild
1805    Description:
1806    
1807    1. I added an alternative to "-rebuild" to "makeml".  The difference is
1808       that prior to calling CMB.make' the CM-variable "LIGHT" will be
1809       defined.  In effect, the command will not build any cross-compiler
1810       backends and therefore finish more quickly.
1811    
1812       The "fixpt" script also takes a "-light" switch to be able to use
1813       this new facility while compiling for a fixpoint.
1814    
1815    2. I replaced all mentions of anchored paths in group owner specifications
1816       with simple relative paths (usually starting with "..").
1817       The rationale is that a library's internal workings should not be
1818       compromised by the lack of some anchor.  (An anchor is necessary
1819       for someone who wants to refer to the library by an anchored path,
1820       but it should not be necessary to build the same library in the first
1821       place.)
1822    
1823    3. I changed the way CM's tool mechanism determines the shell command
1824       string used for things like ml-yacc etc. so that it does not break
1825       when CM.Control.implicit_anchors is turned off.
1826    
1827    ----------------------------------------------------------------------
1828    Name: Matthias Blume
1829    Date: 2000/05/12 18:20:00 JST
1830    Tag: blume-20000512-ml-build
1831    Description:
1832    
1833    Fixed a bug in config/_ml-build that prevented ml-yacc and ml-lex from
1834    getting installed properly (by config/install.sh).
1835    
1836    ----------------------------------------------------------------------
1837    Name: Matthias Blume
1838    Date: 2000/05/12 17:30:00 JST
1839    Tag: blume-20000512-anchors
1840    Description:
1841    
1842    !!! NEW BOOT FILES !!!
1843    
1844    This change is in preparation of fading out support for "implicitly
1845    anchored path names".  I went through all sources and used the
1846    explicit (and relatively new) $-notation.  See system/README and the
1847    CM manual for more info on this.
1848    
1849    I also modified the anchoring scheme for some things such as "smlnj",
1850    "MLRISC", "cm", etc. to take advantage of the fact that explicit
1851    anchors are more expressive: anchor name and first arc do not have to
1852    coincide.  This entails the following user-visible change:
1853    
1854    You have to write $smlnj/foo/bar instead of smlnj/foo/bar.  In
1855    particular, when you fire up sml with a command-line argument, say,
1856    e.g.:
1857    
1858       sml '$smlnj/cmb.cm'
1859    
1860    At the ML toplevel prompt:
1861    
1862       CM.autoload "$smlnj/cmb.cm";
1863    
1864    There is also a new controller in CM.Control that can be used to turn
1865    off all remaining support for implicit anchors by saying:
1866    
1867        CM.autoload "$smlnj/
1868        #set CM.Control.implicit_anchors false;
1869    
1870    This causes CM to reject implicitly anchored paths.  This is (for the
1871    time being) less permissive than the "final" version where there will
1872    be no more such implicit anchors and relative paths will be just that:
1873    relative.
1874    
1875    The next step (version after next version?) will be to make the
1876    default for CM.Control.implicit_anchors false.  After the dust has
1877    settled, I can then produce the "final" version of this...
1878    
1879    Note: Since bootstrapping is a bit tricky, I provided new boot files.
1880    
1881    ----------------------------------------------------------------------
1882    Name: Matthias Blume
1883    Date: 2000/05/11 16:30:00 JST
1884    Tag: blume-20000511-sources
1885    Description:
1886    
1887    The main change is that I added function CM.sources as a generalized
1888    version of the earlier CM.makedepend.  This entails the following
1889    additional changes:
1890    
1891      - CM.makedepend has been dropped.
1892    
1893      - CM manual has been updated.
1894    
1895      - TOOLS signature and API have been changed.
1896    
1897    ----------------------------------------------------------------------
1898    Name: Allen Leung
1899    Date: 2000/05/10 21:17 EDT
1900    Tag: leunga-20000510-moby-c--ssa
1901    Description:
1902    
1903      Various bug fixes and new features for C--, Moby and MLRISC optimizations.
1904    None of these affect SML/NJ.
1905    
1906    1. Register Allocation
1907    
1908        a. A new ra spilling module (ra/ra-spill-with-renaming) is implemented.
1909           This module tries to remove local (i.e. basic block level) redundancies
1910           during spilling.
1911    
1912        b. A new framework for performing region based register allocation.
1913           Not yet entirely functional.
1914    
1915    2. X86
1916    
1917       a. DefUse for POP was missing the stack pointer [found by Lal]
1918       b. Reload for CALL was incorrect in X86Spill [found by John]
1919       c. Various fixes in X86Spill so that it can be used correctly for
1920          the new spilling module.
1921    
1922    3. SSA/IR
1923    
1924       a. New module ir/dj-dataflow.sml implements elimination based
1925          data flow analysis.
1926    
1927    4. MLRiscGen
1928    
1929       a. Fix for gc type annotation
1930    
1931    5. MDGen
1932    
1933       Various fixes for machine description -> ml code translation.  For ssa
1934       only.
1935    
1936    ----------------------------------------------------------------------
1937    Name: Allen Leung
1938    Date: 2000/05/08 22:17 EDT
1939    Tag: leunga-20000508-labexp
1940    Description:
1941    
1942      Fermin has found a few assembly problems with constant expressions
1943      generated in LabelExp.  Mostly, the problems involve extra parentheses,
1944      which choke on dumb assemblers.  This is his fix.
1945    
1946    ----------------------------------------------------------------------
1947    Name: Dave MacQueen
1948    Date: 2000/04/09 14:00 EDT
1949    Tag: dbm-20000502-Version_110_28
1950    Description:
1951    
1952    1. Updated src/compiler/TopLevel/main/version.sml to version 110.28
1953    
1954    2. Updated config/version to 110.28
1955    
1956    3. Updated config/srcarchiveurl
1957    
1958    3. New boot files!
1959       ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.28/
1960    
1961    ----------------------------------------------------------------------
1962    Name: Matthias Blume
1963    Date: 2000/05/01 19:05:00 JST
1964    Tag: blume-20000501-noweb
1965    Description:
1966    
1967    A new noweb tool has been added.  The existing system is entirely
1968    unaffected by this, but some CM users have asked for renewed noweb
1969    support.  Everything is documented in the CM manual.
1970    
1971    New (plugin) libraries:
1972    
1973       noweb-tool.cm
1974       nw-ext.cm
1975    
1976    ----------------------------------------------------------------------
1977    Name: Dave MacQueen
1978    Date: 2000/04/30 12:40PM EDT
1979    Tag: dbm-20000430-bug_fixes
1980    Description:
1981    
1982    1. Fix for bug 1498
1983       smlnj/src/system/Basis/Implementation/Unsafe/object.sig
1984       smlnj/src/system/Basis/Implementation/Unsafe/object.sml
1985         added toRealArray function
1986       smlnj/src/compiler/MiscUtil/print/ppobj.sml
1987         added check for tag Obj.RealArray to array printing case in ppObj
1988    
1989    2. Fix for bug 1510
1990       smlnj/src/compiler/Semant/types/typesutil.sml
1991         fixed definition of dummyargs (used by equalTycon) so that
1992         dummy args are distinct types
1993    
1994    ----------------------------------------------------------------------
1995    Name: Matthias Blume
1996    Date: 2000/04/30 01:00:00 JST
1997    Tag: blume-20000430-versions
1998    Description:
1999    
2000    1. CM version numbering added.  This is an implementation of Lal's
2001       proposal for adding version numbers and version checking to .cm
2002       files.  Lal said that his proposal was just that -- a proposal.
2003       For the time being I went ahead and implemented it so that people
2004       can comment on it.  Everything is completely backward-compatible
2005       (except for the stable library format, i.e., new bootfiles!).
2006    
2007       As usual, see the CM manual for details.
2008    
2009    2. An alternative syntax for anchored paths has been implemented.
2010       Dave has recently voiced the same concerns that I had when I did
2011       this, so there should be some support.  My take is that eventually
2012       I will let support for the current syntax (where anchors are
2013       "implicit") fade out in favor of the new, explicit syntax.
2014       In order to be backward-compatible, both old and new syntax are
2015       currently supported.
2016    
2017       Again, see the CM manual for details.
2018    
2019    3. Parallel make is trying to be slightly smarter:  When the master
2020       process finds a "bottleneck", i.e., when there is only one
2021       compilation unit that can be compiled and everybody else is
2022       waiting on it, then it will simply compile it directly instead
2023       of clumsily telling one of the slaves to do it.
2024    
2025    4. Support for "unsharing" added.  This is necessary in order to be
2026       able to have two different versions of the same library running
2027       at the same time (e.g., for trying out a new MLRISC while still
2028       having the old MLRISC linked into the current compiler, etc.)
2029       See the CM manual.
2030    
2031    5. Simple "makedepend" functionality added for generating Makefile
2032       dependency information.  (This is rather crude at the moment.
2033       Expect some changes here in the future.)
2034    
2035    6. ".fun" added as a recognized suffix for ML files. Also documented
2036       explicitly in the manual that the fallback behavior (unknown suffix
2037       -> ML file) is not an official feature!
2038    
2039    7. Small changes to the pickler for stable libraries.
2040    
2041    8. Several internal changes to CM (for cleanup/improvement).
2042    
2043    
2044    !!!! NEW BINFILES !!!!
2045    
2046    ----------------------------------------------------------------------
2047    Name: Matthias Blume
2048    Date: 2000/04/28 17:30:00 JST
2049    Tag: blume-20000428-pathconfig
2050    Description:
2051    
2052    1. I changed config/install.sh to remove duplicate entries from the
2053       lib/pathconfig file at the end.  Moreover, the final version of
2054       lib/pathconfig is sorted alphabetically.  The same (sorting) is done
2055       in src/system/installml.
2056    
2057    2. The config/install.sh script now consistently uses relative
2058       pathnames in lib/pathconfig whenever the anchor is in the lib
2059       directory.  (So far this was true for the libraries that come
2060       pre-compiled and bundled as part of the bootfiles but not for
2061       libraries that are compiled by the script itself.)
2062    
2063    ----------------------------------------------------------------------
2064    Name: Matthias Blume
2065    Date: 2000/04/26 13:10:00 JST
2066    Tag: blume-20000426-fun_suffix
2067    Description:
2068    
2069    Added ".fun" as a recognized file name suffix (for ML code).
2070    
2071    ----------------------------------------------------------------------
2072    Name: Allen Leung
2073    Date: 2000/04/25 17:00:00 EST
2074    Tag: leunga-20000425-alpha-ra
2075    Description:
2076    
2077    1. Alpha
2078    
2079        PSEUDOARITH was missing in AlphaRewrite.  This causes an endless loop
2080    in C--.
2081    
2082    2. RA
2083    
2084       Added a flag "ra-dump-size" to print out the size of the flowgraph
2085       and the interference graph.
2086    
2087    ----------------------------------------------------------------------
2088    Name: Dave MacQueen
2089    Date: 2000/04/25/
2090    Tag: dbm-20000425-mlyacc_doc_examples
2091    Description:
2092      Updated mlyacc.tex sections 5 and 7 for SML '97 and CM.
2093      Updated all three examples in src/ml-yacc/examples to run
2094      under 110.* using CM.make.
2095    
2096  ----------------------------------------------------------------------  ----------------------------------------------------------------------
2097  Name: Allen Leung  Name: Allen Leung
2098  Date: 2000/04/20 23:04:00 EST  Date: 2000/04/20 23:04:00 EST
# Line 27  Line 2109 
2109    
2110              s <- t              s <- t
2111    
2112        no interference edge between s and t are added for this definition of s.        no interference edge between s and t is added for this definition of s.
2113    
2114     b. I've added two new spill heuristic modules that Fermin and I developed     b. I've added two new spill heuristic modules that Fermin and I developed
2115        (in the new library RA.cm). These are unused in SML/NJ but maybe        (in the new library RA.cm). These are unused in SML/NJ but maybe
# Line 175  Line 2257 
2257    
2258  3.  Assembly  3.  Assembly
2259    
2260      When generating assemby, resolve the value of client defined constants,      When generating assembly, resolve the value of client defined constants,
2261      instead of generating symbolic values.  This is controlled by the      instead of generating symbolic values.  This is controlled by the
2262      new flag "asm-resolve-constants", which is default to true.      new flag "asm-resolve-constants", which is default to true.
2263    
# Line 198  Line 2280 
2280    
2281  To this end, I arranged that instead of "structure Core" as "structure  To this end, I arranged that instead of "structure Core" as "structure
2282  _Core" is bound in the pervasive environment.  Core access is done via  _Core" is bound in the pervasive environment.  Core access is done via
2283  _Core (which can never be accidentially rebound because _Core is not a  _Core (which can never be accidentally rebound because _Core is not a
2284  legal surface-syntax symbol).  legal surface-syntax symbol).
2285    
2286  The current solution is much cleaner because the core environment is  The current solution is much cleaner because the core environment is
# Line 208  Line 2290 
2290  with dynamic and symbolic parts of the core environment.  with dynamic and symbolic parts of the core environment.
2291    
2292  Remaining hackery (to bind the "magic" symbol _Core) is localized in the  Remaining hackery (to bind the "magic" symbol _Core) is localized in the
2293  compilation mananger's bootstrap compiler (actually: in the "init group"  compilation manager's bootstrap compiler (actually: in the "init group"
2294  handling).  See the comments in src/system/smlnj/init/init.cmi for  handling).  See the comments in src/system/smlnj/init/init.cmi for
2295  more details.  more details.
2296    
# Line 323  Line 2405 
2405     (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
2406     used).     used).
2407    
2408     This was done to accomodate the new "make" and "shell" tools which     This was done to accommodate the new "make" and "shell" tools which
2409     facilitate fairly seemless hookup to portions of code managed using     facilitate fairly seamless hookup to portions of code managed using
2410     Makefiles or Shell scripts.     Makefiles or Shell scripts.
2411    
2412     There are no classes "shared" or "private" anymore.  Instead, the     There are no classes "shared" or "private" anymore.  Instead, the
# Line 336  Line 2418 
2418    
2419     All existing tools are described in the CM manual.     All existing tools are described in the CM manual.
2420    
2421     - Slightly better error handling.  (CM now surpresses many followup     - Slightly better error handling.  (CM now suppresses many followup
2422     error messages that tended to be more annoying than helpful.)     error messages that tended to be more annoying than helpful.)
2423    
2424  2. Major changes to the compiler's static environment data structures.  2. Major changes to the compiler's static environment data structures.
# Line 470  Line 2552 
2552    
2553       I've changed andl to testl in the floating point test sequence       I've changed andl to testl in the floating point test sequence
2554       whenever appropriate.  The Intel optimization guide states that       whenever appropriate.  The Intel optimization guide states that
2555       testl is perferable to andl.       testl is preferable to andl.
2556    
2557  7. RA (x86 only)  7. RA (x86 only)
2558    
# Line 652  Line 2734 
2734    
2735  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
2736  additional argument called "template" which is an optional string that  additional argument called "template" which is an optional string that
2737  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
2738  explanation.  explanation.
2739    
2740  2. A special-purpose tool can be "regisitered" by simply dropping the  2. A special-purpose tool can be "registered" by simply dropping the
2741  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
2742  directory where the .cm file lives that uses this tool.  (The  directory where the .cm file lives that uses this tool.  (The
2743  behavior/misfeature until now was to look for the tool description  behavior/misfeature until now was to look for the tool description
# Line 699  Line 2781 
2781  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
2782  <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,
2783  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
2784  perhaps got accidentially killed).  In this case, fixpt will quickly  perhaps got accidentally killed).  In this case, fixpt will quickly
2785  move through what exists before continuing where it left off earlier,  move through what exists before continuing where it left off earlier,
2786  and, thus, saves a lot of time.  and, thus, saves a lot of time.
2787    
# Line 749  Line 2831 
2831    it from that remote directory.    it from that remote directory.
2832    This should simplify installation further:  For machines that have    This should simplify installation further:  For machines that have
2833    access to the internet, just fetch <version>-config.tgz, unpack it,    access to the internet, just fetch <version>-config.tgz, unpack it,
2834    edit config/targets, and go (run config/install.sh).  The scipt will    edit config/targets, and go (run config/install.sh).  The script will
2835    fetch everything else that it might need all by itself.    fetch everything else that it might need all by itself.
2836    
2837    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.624  
changed lines
  Added in v.843

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