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 977, Wed Nov 14 16:53:16 2001 UTC revision 1717, Thu Nov 25 04:11:57 2004 UTC
# Line 8  Line 8 
8  The form of an entry should be:  The form of an entry should be:
9    
10  Name:  Name:
11  Date:  Date: yyyy/mm/dd
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    
15  ----------------------------------------------------------------------  ----------------------------------------------------------------------
16    Name: Matthias Blume (blume (at) tti - c (dot) org)
17    Date: 2004/11/24 22:15:00 CST
18    Tag: blume-20041124-cml
19    Description:
20    
21    Made sure CML compiles when Position = Int64.
22    
23    ----------------------------------------------------------------------
24    Name: Matthias Blume (blume (at) tti - c (dot) org)
25    Date: 2004/11/24 14:30:00 CST
26    Tag: blume-20041124-position
27    Description:
28    
29    The compiler can now be compiled in a mode that makes structure Position
30    equal to Int64.  The default, however, is unchanged (Position = Int31)
31    for the time being.
32    
33    To enable 64-bit positions, use the following procedure:
34    
35    1. Start sml
36    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
37    3. Type
38          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
39    4. Run CMB.make() as usual.
40    
41    This is barely tested.  The only test so far was a little SML program
42    counting the number of characters in an 8-gigabyte file by
43    reading it character-by-character.  That test was successful.
44    
45    In support of 64-bit positions, a number of new functions have been
46    added to the runtime system.
47    
48    ----------------------------------------------------------------------
49    Name: Matthias Blume (blume (at) tti - c (dot) org)
50    Date: 2004/11/23 14:45:00 CST
51    Tag: blume-20041123-useFile
52    Description:
53    
54    Fixed a problem with unhelpful error messages related to problems with
55    .cm- or .sml files that appear as part of the sml command line.
56    
57    ----------------------------------------------------------------------
58    Name: Matthias Blume (blume (at) tti - c (dot) org)
59    Date: 2004/11/18 15:40:00 CST
60    Tag: Release_110_51
61    Description:
62    
63    New working version (110.51).  NEW BOOTFILES!
64    
65    ----------------------------------------------------------------------
66    Name: Matthias Blume (blume (at) tti - c (dot) org)
67    Date: 2004/11/18 15:35:00 CST
68    Tag: <none>
69    Description:
70    
71    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
72    Joemann).
73    
74    ----------------------------------------------------------------------
75    Name: Allen Leung (leunga (at) reservoir (dot) com)
76    Date: 2004/11/17 16:05:21 EST 2004
77    Tag: leunga-20041117-mlrisc-live-kill
78    Description:
79    
80       Added support for MLTree constructs LIVE and KILL
81    to all the architectures.
82    
83    ----------------------------------------------------------------------
84    Name: Matthias Blume (blume (at) tti - c (dot) org)
85    Date: 2004/11/13 00:20:00 CST
86    Tag: blume-20041113-versiontool
87    Description:
88    
89    - Stripped down the versiontool: It now only handles the version number.
90      The date string is generated at bootstrap time (during makeml).
91    
92    - In a previous commit, fixed a minor issue with how polyequal is being
93      translated.  In particular, the code now "looks through" abstractions.
94      This results in slightly fewer polyEqual warnings and hopefully slightly
95      more efficient code.  Important examples for where this matters are
96      the new int64 and word64 types.
97    
98    ----------------------------------------------------------------------
99    Name: Matthias Blume (blume (at) tti - c (dot) org)
100    Date: 2004/11/12 00:30:00 CST
101    Tag: blume-20041112-int64
102    Description:
103    
104    Structure Int64 fully hooked in.  (The implementation is not very
105    efficient, though.)
106    
107    ----------------------------------------------------------------------
108    Name: Matthias Blume (blume (at) tti - c (dot) org)
109    Date: 2004/11/11 17:30:00 CST
110    Tag: blume-20041111-more64
111    Description:
112    
113    All the pieces of Word64 are now there, with the exception of the
114    conversions from and to LargeWord. (Eventually these need to be identities,
115    but for the time being they don't even make sense because LargeWord is
116    32-bit wide.)
117    
118    Also started to add similar support for Int64, but major pieces of that
119    are still missing.
120    
121    ----------------------------------------------------------------------
122    Name: Matthias Blume (blume (at) tti - c (dot) org)
123    Date: 2004/11/11 00:15:00 CST
124    Tag: blume-20041111-word64
125    Description:
126    
127    Structure Word64 is now (almost) complete, word literals and patterns
128    seem to work.  There are a few odd pieces missing.  In particular,
129    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
130    at the moment.
131    
132    Making Word64 official would mean that LargeWord becomes Word64.  But
133    this requires extreme care because most word-word conversions have to
134    go through LargeWord, so making a mistake means loss of efficiency or
135    worse.  Eventually there will be a solution similar to (but actually
136    simpler than) what I did with IntInf.
137    
138    ----------------------------------------------------------------------
139    Name: Matthias Blume (blume (at) tti - c (dot) org)
140    Date: 2004/11/10 18:12:00 CST
141    Tag: blume-20041110-64bit
142    Description:
143    
144    More 64-bit hacking (but still not even half-way there yet).
145    Also, some assorted improvements to the handling of 8-bit words.
146    
147    ----------------------------------------------------------------------
148    Name: Matthias Blume (blume (at) tti - c (dot) org)
149    Date: 2004/11/09 17:50:00 CST
150    Tag: <none>
151    Description:
152    
153    Started adding some infrastructure for supporting 64-bit int- and
154    word-types.  (Still in its very early stages.)
155    
156    ----------------------------------------------------------------------
157    Name: Matthias Blume (blume (at) tti - c (dot) org)
158    Date: 2004/10/28 10:45:00 CDT
159    Tag: Release_110_50
160    Description:
161    
162    New working version (110.50).  NEW BOOTFILES!
163    
164                 =====================
165    
166    Also:
167    
168      - Changend config/srcarchiveurl from a file just
169        containing the URL string into a file containing
170        shell script code.  The code has access to the $VERSION variable.
171      - Made corresponding changes to config/install.sh and config/unpack.
172      - Default contents of config/srcarchiveurl uses $VERSION and
173        normally does not have to be edited to reflect a version change.
174    
175      (As a result, a version change can be done by just editing
176       config/version, the rest is now automatic.)
177    
178    ----------------------------------------------------------------------
179    Name: Matthias Blume (blume (at) tti - c (dot) org)
180    Date: 2004/10/27 17:50:00 CDT
181    Tag: blume-20041027-btrace-msg
182    Description:
183    
184    BackTrace.monitor now also reports the source of the exception that
185    triggered the trace.
186    
187    ----------------------------------------------------------------------
188    Name: Matthias Blume (blume (at) tti - c (dot) org)
189    Date: 2004/10/27 17:20:00 CDT
190    Tag: blume-20041027-x86-c-calls
191    Description:
192    
193    This is the HISTORY entry for two earlier commits, both concerning
194    the x86 c-calls code in MLRISC:
195    
196      - added a missing LOAD in the code that deals with struct arguments
197      - made sure the caller does not add the wrong number of bytes to the
198        stack pointer after a call of a function returning a struct
199        (the callee already pops the implicit argument which points to
200         the space reserved for the result)
201    
202    ----------------------------------------------------------------------
203    Name: Allen Leung (leunga (at) reservoir (dot) com)
204    Date: 2004/10/24 14:00:00 EST
205    Tag: leunga-20041024-x86-gas-fucomip
206    Description:
207    
208       John discovered a bug in the syntax of fucomip.
209       The opcodes FU?COMIP? have been changed to
210    
211           fu?comip? %st(i), %st
212    
213    ----------------------------------------------------------------------
214    Name: Matthias Blume (blume (at) tti - c (dot) org)
215    Date: 2004/10/20 15:06:00 CDT
216    Tag: blume-20041020-standalone-backtrace
217    Description:
218    
219    Added a mechanism for getting back-trace information from standalone
220    programs.  Here is how it works:
221    
222       1. The part of the program from which you want to get backtrace
223          information (usually the whole program) should be wrapped with
224          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
225          main program could be modified from something like
226    
227             fun main (pgm, args) = ...
228    
229         to
230    
231            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
232    
233       2. To be able to access BackTrace.monitor, you have to add
234          library $smlnj-tdp/plugins.cm to the .cm file that contains your
235          main function.
236    
237       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
238          CM might have created in the past for your project).
239    
240       4. Build the system using this command line:
241    
242             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
243                      myprog.cm MyProg.main myprog
244    
245          instead of the usual
246    
247             ml-build myprog.cm MyProg.main myprog
248    
249    I changed a library name:
250    
251       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
252    
253    New libraries:
254    
255       $smlnj-tdp/back-trace.cm
256          -- when loaded causes the back-trace plugin to be installed
257       $smlnj-tdp/coverage.cm
258          -- when loaded causes the coverage plugin to be installed
259    
260    ----------------------------------------------------------------------
261    Name: Matthias Blume (blume (at) tti - c (dot) org)
262    Date: 2004/10/18 16:45:00 CDT
263    Tag: blume-20041018-groupowner
264    Description:
265    
266    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
267    
268    Eliminated group owner specs from .cm files throughout the source tree.
269    
270    ----------------------------------------------------------------------
271    Name: Matthias Blume (blume (at) tti - c (dot) org)
272    Date: 2004/10/15 15:45:00 CDT
273    Tag: blume-20041015-coverage
274    Description:
275    
276    * Test coverage tool added!
277    * Further reorganization of tracing-, debugging-, and profiling support:
278    
279        - moved original BTImp -- now called BackTrace -- into a separate
280          library called $/trace-debug-profile.cm
281        - eliminated all mentions of BTrace from SMLofNJ.Internals
282        - only the instrumentation mechanism is now left in the compiler proper
283        - BackTrace module is a plugin which is NOT plugged in by default
284        - Coverage module is another such plugin
285    
286    To get the benefits of any of these plugin modules, the code in
287    question must be compiled with tdp instrumentation turned on.  This
288    can be done by setting SMLofNJ.Internals.TDP.mode to true.
289    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
290    
291    Plugins are selected at link time.  (Pre-compiled instrumented code
292    can be re-loaded with different plugins in effect.)  When an
293    instrumented module is linked, whatever plugins are at that time
294    enabled will come into effect for that module.
295    
296    To enable the back-trace plugin, load library $/trace-debug-profile.cm
297    and invoke BackTrace.install() (e.g., from the interactive prompt).
298    To enable the coverage plugin, load the same library and invoke
299    Coverage.install().
300    
301    Back-traces are generated automatically on uncaught exceptions and
302    when the code in question explicitly invokes BackTrace.trigger().
303    
304    Coverage (and execution frequency-) information must be queried
305    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
306    
307    ----------------------------------------------------------------------
308    Name: Matthias Blume (blume (at) tti - c (dot) org)
309    Date: 2004/10/14 17:40:00 CDT
310    Tag: blume-20041014-tdp-core
311    Description:
312    
313    Snapshot of a significant overhaul of how the trace/debug/profile support
314    is hooked into the system (specifically: Core and SMLofNJ.Internals).
315    
316    ----------------------------------------------------------------------
317    Name: Matthias Blume (blume (at) tti - c (dot) org)
318    Date: 2004/10/13 16:34:00 CDT
319    Tag: blume-20041013-tdp
320    Description:
321    
322    Some rationalization of names:
323    
324       structure BTrace -> structure TDPInstrument
325       etc.
326    
327    This is is preparation of using the original back-trace
328    instrumentation for other purposes.  "TDP" stands for
329    Trace/Debug/Profile.
330    
331    The control flag controlling whether instrumentation is on or off is now
332    registered under a different name, so instead of running sml as
333    
334     sml -Cinstrument.btrace-mode=true
335    
336    one has to say
337    
338     sml -Ctdp.instrument=true
339    
340    ----------------------------------------------------------------------
341    Name: Matthias Blume (blume (at) tti - c (dot) org)
342    Date: 2004/10/11 16:37:00 CDT
343    Tag: blume-20041011-regions
344    Description:
345    
346    Made some minor modifications to elabcore.sml to have source regions
347    be propagated more tightly -- resulting in better (i.e., smaller)
348    regions being reported in error- and debug messages.
349    
350    ----------------------------------------------------------------------
351    Name: Matthias Blume (blume (at) tti - c (dot) org)
352    Date: 2004/10/08 22:50:00 CDT
353    Tag: blume-20041008-cmkw
354    Description:
355    
356    Fixed handling of keywords in .cm files: After seeing "is" the lexer
357    treats subsequent occurrences of "group", "library", "source", "is",
358    "*", and "-" as ordinary identifiers rather than keywords.
359    
360    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
361    is supposed to accept a tool argument called "source", but this did
362    not work because of the clash with the keyword.
363    
364    ----------------------------------------------------------------------
365    Name: Matthias Blume (blume (at) tti - c (dot) org)
366    Date: 2004/10/07 16:00:00 CDT
367    Tag: blume-20041007-cleanup
368    Description:
369    
370    Assorted cleanup work:
371    
372      - got rid of intstrmap in favor of using the library's
373        hash table implementation
374      - threw out most of the pathnames stuff, as it was not used anyway
375      - simplified tokentable implementation
376      - fixed some minor spelling errors
377    
378    ----------------------------------------------------------------------
379    Name: Matthias Blume (blume (at) tti - c (dot) org)
380    Date: 2004/10/06 15:15:15 CDT
381    Tag: blume-20041006-handler
382    Description:
383    
384    Cleaned up the absyn to reflect the invariant that HANDLE always
385    carries a FNexp as part of the type definition.  This eliminates some
386    superfluous sanity checks at runtime down the road.
387    
388    Some minor cleanup of the btrace code.
389    
390    ----------------------------------------------------------------------
391    Name: Matthias Blume (blume (at) tti - c (dot) org)
392    Date: 2004/10/01 10:20:30 CDT
393    Tag: blume-20041001-slave
394    Description:
395    
396    Added hack to make slave mode work in the presence of the version
397    tool.  (Still, since the master does two passes over the code for
398    CMB.make, the release number gets bumped twice when slaves are
399    attached.  I don't know if this is worth fixing...)
400    
401    ----------------------------------------------------------------------
402    Name: Matthias Blume (blume (at) tti - c (dot) org)
403    Date: 2004/09/30 10:55:00 CDT
404    Tag: blume-20040930-version
405    Description:
406    
407    * Moved the "version" magic into its own little library under
408      src/system/smlnj/internal.  This avoids expensive reconstruction of
409      a stable src/compiler/core.cm.
410    
411    * At the same time, structure CompilerVersion is now known as
412      structure SMLNJVersion.
413    
414    * Arranged for the version tool to NOT kick in when rebuilding the system
415      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
416      Also, loading the versiontool does not work when rebuilding the system
417      because CM is not properly initialized at that time.
418    
419    ----------------------------------------------------------------------
420    Name: Matthias Blume (blume (at) tti - c (dot) org)
421    Date: 2004/09/29 14:00:00 CDT
422    Tag: blume-20040929-autoversion
423    Description:
424    
425    Implemented some CM magic to have
426        file src/compiler/TopLevel/main/version.sml
427    generated automagically.
428    The version is taken from two files: config/version and config/release.
429    The first is expected to contain a two-part version number such as 110.49.
430    The second should contain a single number, but it may be missing.
431    
432    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
433    time the version tool is loaded (which is the first time you say CMB.make),
434    then the tool will increment the value stored in config/release every
435    time CMB.make is invoked.
436    
437    The binfile format is now insensitive to anything beyond the first
438    two components of a version number, so bumping the release does not render
439    binfiles incompatible.  Auto-bumping can be used to keep track of versions
440    during development without invalidating existing binfiles.
441    
442    In any case, every CMB.make updates the date information in version.sml.
443    (This is the date that is printed in the banner.)
444    
445    ----------------------------------------------------------------------
446    Name: Matthias Blume (blume (at) tti - c (dot) org)
447    Date: 2004/09/28 10:53:00 CDT
448    Tag: blume-20040928-controls
449    Description:
450    
451    Some cleanup of the controls code.
452    
453    ----------------------------------------------------------------------
454    Name: Matthias Blume (blume (at) tti - c (dot) org)
455    Date: 2004/09/27 22:08:00 CDT
456    Tag: blume-20040927-controls
457    Description:
458    
459    Added two pieces of functionality to the Controls interface:
460    
461    1.   val save'restore: 'a control -> unit -> unit
462    
463       grabs the current value of the control in stage 1 and restores it
464       in stage 2.
465    
466    2.   val set' : 'a control * 'a -> unit -> unit
467    
468       stores the given value into the control in stage 2 (i.e., delayed)
469       but does all error checking in stage 1.
470       (This is for string controls that need to do parse their argument --
471        something that might fail.  In some cases, notably in CM, one
472        already knows the intended argument but wants to delay the actual
473        assignment until a time when error recovery would be more difficult.)
474    
475    Changed the handling of controls in tool arguments to classes "sml" and
476    "lazysml":
477       - use Controls.save'restore as a more robust way of restoring the
478         old value (in particular: without having to re-parse the string)
479       - use controls to handle the "overload" keyword in the init group
480         (I believe this change actually fixes a long-standing obscure bug.)
481    
482    ----------------------------------------------------------------------
483    Name: Matthias Blume (blume (at) tti - c (dot) org)
484    Date: 2004/09/27 17:00:00 CDT
485    Tag: blume-20040927-lazysml
486    Description:
487    
488    Added a new tool class called "lazysml" to CM's tool chest.  The only
489    difference to "sml" is that compilation is done with Control.lazysml
490    set to true.  A source of class "lazysml" is automatically recognized
491    by a file name suffix of ".lml".
492    
493    In addition to the above feature, the original class "sml" now also
494    supports a tool argument "lazy" which has the same effect.  As a
495    result, the following three lines are equivalent:
496    
497        foo.sml : lazysml
498        foo.sml : sml (lazy)
499        foo.sml (lazy)
500    
501    The setting goes into effect both during parsing and during
502    compilation.  The original setting is restored right after parsing and
503    after compilation, respectively.
504    
505    In addition to all the above, there is also a general mechanism to set
506    ANY of the "controls" that are available at the command line via
507    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
508    apply as well. (In fact, "lazy" is implemented as a special case of
509    the general mechanism.)
510    
511    The .cm file syntax uses a new keyword tool argument called "with".
512    There are several ways of indicating the desired settings:
513    
514        foo.sml (with:parser.quotations=true)
515        foo.sml (with:(name:parser.quotations value:true))
516        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
517        foo.sml (with:(name1=value1 name2=value2 ...))
518        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
519    
520        etc.
521    
522    Another possible abbreviation is to leave out the =v or value:v part
523    if the name refers to a boolean control (in which case the value is
524    taken to be true).  Thus, one could get lazy sml also by saying:
525    
526        foo.sml (with:parser.lazy-keyword=true)
527        foo.sml (with:parser.lazy-keyword)
528        foo.sml (with:(name:parser.lazy-keyword value:true))
529        foo.sml (with:(name:parser.lazy-keyword))
530    
531    ----------------------------------------------------------------------
532    Name: Matthias Blume (blume (at) tti - c (dot) org)
533    Date: 2004/09/24 16:22:00 CDT
534    Tag: blume-20040924-ppc-long-branch
535    Description:
536    
537    Turned message about "emiting long form of branch" off by default.
538    Added a control flag to turn it back on when desired.
539    
540    ----------------------------------------------------------------------
541    Name: Matthias Blume (blume (at) tti - c (dot) org)
542    Date: 2004/09/24 16:05:00 CDT
543    Tag: blume-20040924-rounding
544    Description:
545    
546    Applied patch for setting rounding modes under Mac OS X.  Thanks to
547    Melissa O'Neill for providing the code!
548    
549    ----------------------------------------------------------------------
550    Name: Matthias Blume (blume (at) tti - c (dot) org)
551    Date: 2004/09/23 17:30:00 CDT
552    Tag: blume-20040923-envvars
553    Description:
554    
555    1. Changed definition of type ControlRegistry.registry_tree to
556       include control_info (i.e., the name of the controlling environment
557       variable).
558    
559    2. Added command-line flags -e and -E to print the names of environment
560       variables that can be used to control internal settings.
561       (This uses the new API mentioned in 1.)
562    
563    ----------------------------------------------------------------------
564    Name: Matthias Blume (blume (at) tti - c (dot) org)
565    Date: 2004/09/13 16:50:00 CDT
566    Tag: Release_110_49
567    Description:
568    
569    New working version (110.49).  NEW BOOTFILES!
570    
571    ----------------------------------------------------------------------
572    Name: Matthias Blume
573    Date: 2004/09/13 16:20:00 CDT
574    Tag: blume-20040913-config-mlrisc
575    Description:
576    
577    Put target "mlrisc" back into the default list.
578    (There is no harm in having it, and some users have expressed their
579    wish to have "mlrisc" included by default.)
580    
581    ----------------------------------------------------------------------
582    Name: John Reppy
583    Date: 2004/09/13
584    Tag: jhr-20040913-signals
585    Description:
586    
587    Fixed the signal masking code to properly nest mask/unmask operations
588    on a per-signal basis.
589    
590    ----------------------------------------------------------------------
591    Name: Matthias Blume (blume (at) tti - c (dot) org)
592    Date: 2004/09/08 13:20:00 CDT
593    Tag: blume-20040908-heap-magic
594    Description:
595    
596    Bumped the heap macig number to 0x09082004 to account for the changed
597    layout of the ML frame under MacOS X.
598    
599    ----------------------------------------------------------------------
600    Name: Allen Leung (leunga (at) reservoir (dot) com)
601    Date: 2004/09/03 11:26:00 EST
602    Tag: leunga-20040903-cygwin-install
603    Description:
604    
605       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
606    The Cygwin runtime is turned on by setting the environment
607    variable SMLNJ_CYGWIN_RUNTIME to 1.
608    
609    ----------------------------------------------------------------------
610    Name: Matthias Blume (blume (at) tti - c (dot) org)
611    Date: 2004/08/31 17:15:00 CDT
612    Tag: blume-20040831-core
613    Description:
614    
615    Added some exports to src/compiler/core.cm upon request by J. Joemann.
616    
617    ----------------------------------------------------------------------
618    Name: Matthias Blume (blume (at) tti - c (dot) org)
619    Date: 2004/08/30 17:55:00 CDT
620    Tag: blume-20040830-installer
621    Description:
622    
623    Upon request by Johannes Joemann:
624    
625    - improved ML code of installer to fall back to coping when renaming
626      fails (i.e., when source and target are on different file systems);
627      the code compiles but has yet to be tested in anger
628    - removed mlrisc from list of default targets (config/targets)
629    
630    ----------------------------------------------------------------------
631    Name: Matthias Blume (blume (at) tti - c (dot) org)
632    Date: 2004/08/27 17:20:00 CDT
633    Tag: blume-20040827-ptreql
634    Description:
635    
636    Added ptreql primop to structure InlineT (upon request from Larry
637    Paulson).
638    
639    ----------------------------------------------------------------------
640    Name: Allen Leung (leunga (at) reservoir (dot) com)
641    Date: 2004/08/15 21:21:00 EST
642    Tag: leunga-110_48-udgraph
643    Description:
644    
645    Another bug fix from Carl Hauser:
646    
647    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
648    > 48c48
649    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
650    > e::L)
651    > ---
652    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
653    > Without this, any deletion of an edge in an undirected graph does severe
654    > violence to the graph.
655    
656    ----------------------------------------------------------------------
657    Name: Allen Leung (leunga (at) reservoir (dot) com)
658    Date: 2004/08/10 23:35:00 EST
659    Tag: leunga-110_48-ppc
660    Description:
661    
662       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
663    
664    ----------------------------------------------------------------------
665    Name: Matthias Blume (blume (at) tti - c (dot) org)
666    Date: 2004/08/10 12:00:00 CDT
667    Tag: Release_110_48
668    Description:
669    
670    New working version (110.48).  NEW BOOTFILES!
671    
672    ----------------------------------------------------------------------
673    Name: Allen Leung (leunga (at) reservoir (dot) com)
674    Date: 2004/08/09 12:21:00 EST
675    Tag: leunga-110_47-dijsktra
676    Description:
677    
678    Bug fix from Carl Hauser:
679    
680    single_source_shortest_paths in dijkstra.sml was observed to get wrong
681    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
682    
683    The problem is that following the expression A.update(dist,s,Num.zero)
684    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
685    
686    ----------------------------------------------------------------------
687    Name: Matthias Blume (blume (at) tti - c (dot) org)
688    Date: 2004/08/06 18:10:00 CDT
689    Tag: blume-20040806-cmdline
690    Description:
691    
692    Fiddled with handling of command-line options:
693    
694      * sml now quits after processing the command line
695        if -H, -S, -h<n>, or -s<n> appears as the last
696        command-line argument
697      * a new option -q terminates the session when encountered on
698        the command line; subsequent arguments will be ignored
699      * bug fixes: short (erroneous) arguments are no longer ignored
700        completely
701    
702    ----------------------------------------------------------------------
703    Name: Allen Leung (leunga (at) reservoir (dot) com)
704    Date: 2004/08/04 18:17:00 EST
705    Tag: leunga-110_47-ppc-ibm-asm
706    Description:
707    
708      - Added minimal IBM assembly syntax support for PowerPC.
709    
710      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
711        been altered in the runtime.  We need an automatic way to keep the file
712        in sync.
713    
714    ----------------------------------------------------------------------
715    Name: Matthias Blume (blume (at) tti - c (dot) org)
716    Date: 2004/08/04 14:00:00 CDT
717    Tag: Release_110_47
718    Description:
719    
720    New working version (110.47).  NEW BOOTFILES!
721    
722    ----------------------------------------------------------------------
723    Name: Matthias Blume (blume (at) tti - c (dot) org)
724    Date: 2004/08/03 14:25:00 CDT
725    Tag: blume-20040803-callingconv
726    Description:
727    
728    Added low-level support for choosing C calling conventions by
729    twiddling the type of rawccall.  (See
730    src/compiler/Semant/types/cproto.sml for details.)
731    
732    ----------------------------------------------------------------------
733    Name: Matthias Blume (blume (at) tti - c (dot) org)
734    Date: 2004/08/02 15:55:00 CDT
735    Tag: blume-20040802-backout
736    Description:
737    
738    Backed out of change to win32-filesys.c.  The earlier patch to
739    get_file_time caused CM to produce files with the wrong time stamp.
740    
741    ----------------------------------------------------------------------
742    Name: Matthias Blume (blume (at) tti - c (dot) org)
743    Date: 2004/08/02 14:45:00 CDT
744    Tag: blume-20040802-nlffi-win32
745    Description:
746    
747    Added NLFFI support for Win32, adapted from a patch provided by David
748    Hansel.  This is currently completely untested.  Also, the issue
749    concerning stdcall vs. ccall is still unresolved.
750    
751    ----------------------------------------------------------------------
752    Name: Matthias Blume (blume (at) tti - c (dot) org)
753    Date: 2004/07/30 17:55:00 CDT
754    Tag: blume-20040730-various
755    Description:
756    
757    Gearing up towards 110.47...
758    
759    - various minor bugfixes to ml-nlffigen
760    - a beginning of a manual for nlffi
761    
762    - eliminated 'export name=value' in config/install.sh as this does
763      not work with certain versions of /bin/sh
764      (Thanks to David King at Motorola for catching this.)
765    
766    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
767      - added a test for tm==NULL to gmtime.c and localtime.c
768      - applied patch for incorrect GetFileTime under win32
769      - toSeconds -> toMilliseconds in Win32/win32-process.sml
770    
771    ----------------------------------------------------------------------
772    Name: Matthias Blume (blume (at) tti - c (dot) org)
773    Date: 2004/07/21 18:20:00 CDT
774    Tag: blume-20040721-nlffigen
775    Description:
776    
777    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
778      for a typedef to an incomplete type, but leave out the "typ" member.
779      (This is just for consistency.)
780    - Started to produce what is supposed to become better (i.e., comprehensive)
781      documentation of what ml-nlffigen does and produces.
782    
783    ----------------------------------------------------------------------
784    Name: Matthias Blume (blume (at) tti - c (dot) org)
785    Date: 2004/07/14 16:25:00 CDT
786    Tag: blume-20040714-union
787    Description:
788    
789    Added C_UNION to c-calls/c-types.sml and updated the machinery
790    (ml-nlffigen, cproto.sml) that conveys C function interface
791    information to the code generator.
792    
793    However, the actual architecture-specific implementation of function
794    arguments and results that are C unions is still not implemented.
795    
796    ----------------------------------------------------------------------
797    Name: Allen Leung (leunga (at) reservoir (dot) com)
798    Date: 2004/07/14 14:38:00 EST
799    Tag: leunga-110_46_1-ppc-lwzu
800    Description:
801    
802        Added these instructions to the PowerPC architecture:
803        LBZU(X), LHZU(X), LWZU(X),
804        STWU(X), STFDU, STFSU
805    
806        etc...
807    
808        Note: I haven't added their instruction encoding into the description.
809    
810    ----------------------------------------------------------------------
811    Name: Allen Leung (leunga (at) reservoir (dot) com)
812    Date: 2004/07/13 15:04:00 EST
813    Tag: leunga-110_46_1-ppc-lwarx
814    Description:
815    
816       Added the two instructions LWARX and STWCX to the PowerPC
817    instruction set.
818    
819       A (untested) rewrite of loop-structure.sml.  The old version
820    is completely broken.
821    
822    ----------------------------------------------------------------------
823    Name: Matthias Blume (blume (at) tti - c (dot) org)
824    Date: 2004/07/13 13:50:00 CDT
825    Tag: blume-20040713-nlffi
826    Description:
827    
828    - use paramAlloc to report c-calls with too many arguments
829      (for PPC version where parameter area is pre-allocated)
830    - added ccall_maxargspace to machspec (to implement the above)
831    - made "make" commend in CM's "make" tool configurable
832    - added option (default: on) for passing the name of the SML/NJ's "bin"
833      directory to "make"; the call looks like this:
834    
835         make <options> SMLNJ_BINDIR=<dir> <target>
836    
837      This can be used by the Makefile to, e.g., pick the "right" version
838      of ml-nlffigen.
839    - minor code tweaks
840    
841    ----------------------------------------------------------------------
842    Name: Matthias Blume (blume (at) tti - c (dot) org)
843    Date: 2004/07/12 22:50:00 CDT
844    Tag: blume-110_46_1-macosx-nlffi
845    Description:
846    
847    NLFFI under Mac OS X now working (sort of).  This is largely untested,
848    though.
849    
850    Note:  1. You have to make a new, clean build of the runtime system.
851           2. There are new BOOTFILES, you have to use them!
852              (Doing the bootstrap process yourself would be *very* painful!
853               If you absolutely have to do it, build the system under
854               a different architecture and then cross-compile.)
855    
856    Version bumped to 110.46.1 to account for runtime data format changes.
857    
858    ----------------------------------------------------------------------
859    Name: Matthias Blume (blume (at) tti - c (dot) org)
860    Date: 2004/06/18 14:30:00 CDT
861    Tag: blume-20040618-unix
862    Description:
863    
864    Changed the implementation of structure Unix so that the same stream
865    is returned every time one of the {text,bin}{In,Out}streamOf functions
866    is invoked on the same proc.  This is not what the spec currently
867    says -- although IMO it arguably should.  (See discussion below.)
868    
869    ----------------------------------------------------------------------
870    Name: Matthias Blume (blume (at) tti - c (dot) org)
871    Date: 2004/06/17 18:15:00 CDT
872    Tag: Release_110_46
873    Description:
874    
875    New working version (110.46).  NEW BOOTFILES!
876    
877    ----------------------------------------------------------------------
878    Name: Matthias Blume (blume (at) tti - c (dot) org)
879    Date: 2004/06/17 17:20:00 CDT
880    Tag: blume-20040617-timer-unix
881    Description:
882    
883    Changed the interface of structures Timer and Unix to match the most
884    recent Basis spec.
885    
886    In the case of Unix there still seems to be an open/weird issue:
887    
888       The {text,bin}{In,Out}streamOf functions are supposed to create
889       fresh streams whenever they are called -- as opposed to have them
890       return the same stream every time.  This design is supposed to
891       prevent space leaks caused by proc values hanging on to streams.
892    
893       The reap function, on the other hand, is supposed to close the
894       streams.  This cannot be done without having a handle on the
895       stream in proc after all...
896    
897    I took the liberty to implement the following stopgap solution:
898    
899      The proc value hangs on to the most recently created stream(s).
900      Reap closes those.  If either or both of the two streams hadn't
901      been created at all yet, then reap will close the corresponding
902      file descriptors directly.
903    
904    PS: I don't understand the original space leak argument anymore.  If
905    a proc hangs on to the imperative stream, then I/O operations on those
906    will advance the state of the cached stream and avoid the space leak.
907    
908    ----------------------------------------------------------------------
909    Name: Matthias Blume (blume (at) tti - c (dot) org)
910    Date: 2004/05/28 16:45:00 CDT
911    Tag: blume-20040528-basis
912    Description:
913    
914    Added signature PACK_REAL and exported functor PrimIO.
915    
916    ----------------------------------------------------------------------
917    Name: Matthias Blume (blume (at) tti - c (dot) org)
918    Date: 2004/05/25 16:00:00 CDT
919    Tag: blume-20040525-group-owner
920    Description:
921    
922    CM now ignores (but still accepts) the "owner" information in group
923    descriptions.  The owner of a group is its next enclosing
924    library. Each group must have a unique owner.  (There is a virtual
925    "toplevel" library that owns groups which are not nested within a real
926    library.)  Previously, each group had to explicitly declare its owner,
927    and CM would check that such a declaration is correct.  The new scheme
928    is to have CM check that for each group there is precisely one owning
929    library.
930    
931    The advantage of the new scheme is that the programmer no longer needs
932    to maintain the somewhat annoying owner information.  The downside is
933    that CM cannot enforce the ownership rule across multiple runs of
934    CM.make.  Fortunately, enclosing the same group in two different
935    libraries A and B which are not part of the same program does not
936    cause real problems.
937    
938    ----------------------------------------------------------------------
939    Name: Matthias Blume (blume (at) tti - c (dot) org)
940    Date: 2004/05/20 16:00:00 CDT
941    Tag: blume-20040520-win32
942    Description:
943    
944    Made the win32 version work again.  (Strangely, a misplaced comma had
945    slipped into win32-process.c which prevented the runtime from being
946    compiled correctly.)
947    
948    Also, included a minor addition to ml-build.bat analogous to what was
949    done in blume-20040519-ml-build.
950    
951    ----------------------------------------------------------------------
952    Name: Matthias Blume (blume (at) tti - c (dot) org)
953    Date: 2004/05/19 22:10:00 CDT
954    Tag: blume-20040519-ml-build
955    Description:
956    
957    Arranged for ml-build to clean up after itself a little bit better.
958    The script generates a temporary SML source file and compiles it using
959    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
960    gets rid of those at the end, so they don't accumulate under .cm.
961    
962    This required a minor change to install.sh because the name of the
963    metadata directory (default: .cm) is actually configurable at
964    installation time.
965    
966    ----------------------------------------------------------------------
967    Name: Matthias Blume (blume (at) tti - c (dot) org)
968    Date: 2004/05/18 15:50:00 CDT
969    Tag: blume-20040518-mkreader
970    Description:
971    
972    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
973    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
974    
975    ----------------------------------------------------------------------
976    Name: Matthias Blume (blume (at) tti - c (dot) org)
977    Date: 2004/05/11 14:35:00 CDT
978    Tag: blume-20040511-win32sock
979    Description:
980    
981    Added previously missing support for many socket-related functions
982    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
983    for the voluminous patch!
984    
985    (I have not tested this patch under win32 yet.)
986    
987    Here is David's e-mail:
988    
989    Hi,
990    
991    Attached to this email you find a diff against sml/nj 110.45
992    that will enable socket support under Windows.
993    
994    To apply the patch (using unix or cygwin)
995    1) gunzip runtime.diff.gz
996    2) "cd" into "src/runtime" in the source tree of a fresh
997       110.45 installation.
998    3) patch -p 1 < [your/path/to]runtime.diff
999    
1000    The code compiles fine but has NOT yet been extensively tested.
1001    I only ran a few tests for basic socket client functionality
1002    (which worked fine).  Especially the functions that use ioctl
1003    are not tested at all and might not work (see below).
1004    
1005    I implemented this since we want to move to a newer version of sml/nj
1006    but need socket support in order to use it.  This is the first time I
1007    even had a look at the sml/nj source,  so please review my changes
1008    before making this part of the distribution!  Here are a few issues
1009    that I think might be better for someone to solve who is more
1010    familiar with the sml/nj source (and socket programming):
1011    
1012    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1013      exception since I could not figure out what the windows equivalent
1014      of these functions is
1015    
1016    - In sockets-osdep.h there are a some #include statements that are
1017      only used in a few files that include sockets-osdep.h
1018    
1019    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1020      does not process its return value since I don't know how to
1021      report an error upwards.
1022    
1023    - It would probably be good to have a call to WSACleanup() when
1024      the library is unloaded (if there is such a possibility).
1025      Otherwise I think Windows will take care of this automatically
1026      when the process finishes.
1027    
1028    - I used ioctlsocket() as a replacement for ioctl() but I have
1029      no idea if that is actually the proper replacement on Windows.
1030    
1031    - All these issues are marked in the code by "FIXME" comments.
1032    
1033    We use sml/nj extensively in our products and are quite happy
1034    with it.  I hope this contribution will help you.
1035    
1036    Keep up the good work!
1037    
1038    David
1039    
1040    ----------------------------------------------------------------------
1041    Name: Matthias Blume (blume (at) tti - c (dot) org)
1042    Date: 2004/05/11 14:20:00 CDT
1043    Tag: blume-20040511-installml
1044    Description:
1045    
1046    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1047    the patch.)
1048    
1049    ----------------------------------------------------------------------
1050    Name: Matthias Blume (blume (at) tti - c (dot) org)
1051    Date: 2004/05/11 14:05:00 CDT
1052    Tag: blume-20040511-nlffi-netbsd
1053    Description:
1054    
1055    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1056    the patch.)
1057    
1058    ----------------------------------------------------------------------
1059    Name: Matthias Blume (blume (at) tti - c (dot) org)
1060    Date: 2004/05/11 12:05:00 CDT
1061    Tag: blume-20040511-exports
1062    Description:
1063    
1064    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1065    export lists in compiler-related .cm-files.
1066    
1067    ----------------------------------------------------------------------
1068    Name: Matthias Blume (blume (at) tti - c (dot) org)
1069    Date: 2004/05/11 11:35:00 CDT
1070    Tag: blume-20040511-allsource
1071    Description:
1072    
1073    The installer now honors the "src-smlnj" target again, although its meaning
1074    has changed from "all sources required for the compiler" to "all sources
1075    the installer knows about".  In other words, if you enable "src-smlnj"
1076    in the "targets" file, then the installer will pull in sources for
1077    everything.  (Notice that this refers to source code only.  Compiled
1078    code is still only installed for modules that were requested explicitly
1079    or which are required for other modules that were requested explicitly.)
1080    
1081    ----------------------------------------------------------------------
1082    Name: Matthias Blume (blume (at) tti - c (dot) org)
1083    Date: 2004/04/23 17:40:00 CDT
1084    Tag: blume-20040423-ieee-scan
1085    Description:
1086    
1087    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1088    in the exponent calculation we get INF or ZERO (depending on the mantissa
1089    and the sign of the exponent).
1090    
1091    ----------------------------------------------------------------------
1092    Name: Matthias Blume (blume (at) tti - c (dot) org)
1093    Date: 2004/04/23 10:40:00 CDT
1094    Tag: blume-20040423-ml-build
1095    Description:
1096    
1097    The ml-build script now terminates with a non-0 status when something
1098    goes wrong.
1099    
1100    ----------------------------------------------------------------------
1101    Name: Matthias Blume (blume (at) tti - c (dot) org)
1102    Date: 2004/04/22 16:35:00 CDT
1103    Tag: blume-20040422-Option
1104    Description:
1105    
1106    Made exception Option to be the same as exception Option.Option
1107    (as it should be).
1108    
1109    ----------------------------------------------------------------------
1110    Name: Allen Leung (leunga (at) reservoir (dot) com)
1111    Date: 2004/03/19 14:40:00 EST
1112    Tag: leunga-20040319-cygwin-nlffi
1113    Description:
1114    
1115    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1116    of SML/NJ.  The problem is that
1117    
1118       lib = dlopen(NULL, ...)
1119       f   = dlsym(lib, "malloc");
1120    
1121    does not work on Windows unless we explicitly export symbols
1122    such as 'malloc' during linking.  We fixed this by explicitly
1123    exporting the required symbols with the magic gcc incantation:
1124    
1125        -Wl,--export-all cygwin.def
1126    
1127    where cygwin.def is a file containing all the symbols that
1128    we wish to export.
1129    
1130    I suspect this is a Windows problem and we'll have to
1131    do the same (somehow with windows compilers) when
1132    we build the native win32 version with the system
1133    calls LoadLibrary/GetProcAddress.
1134    
1135    ----------------------------------------------------------------------
1136    Name: Matthias Blume (blume (at) tti - c (dot) org)
1137    Date: 2004/03/04 16:35:00 CST
1138    Tag: blume-20040304-intinf-fmt
1139    Description:
1140    
1141    Fixed problem with IntInf.fmt (sign would show up on the right instead
1142    of on the left for BIN, OCT, and HEX).
1143    
1144    ----------------------------------------------------------------------
1145    Name: Matthias Blume (blume (at) tti - c (dot) org)
1146    Date: 2004/03/04 11:25:00 CST
1147    Tag: blume-20040304-symlinks
1148    Description:
1149    
1150    Fixed problem with installer script (unix only) where bin/ml-yacc and
1151    friends pointed (via symlinks) to absolute locations instead of just
1152    .run-sml.  This was reported by Vesa A Norrman.
1153    
1154    ----------------------------------------------------------------------
1155    Name: Matthias Blume (blume (at) tti - c (dot) org)
1156    Date: 2004/02/13 14:50:00 CST
1157    Tag: Release_110_45
1158    Description:
1159    
1160    New working version (110.45). New bootfiles.
1161    
1162    ----------------------------------------------------------------------
1163    Name: Matthias Blume (blume (at) tti - c (dot) org)
1164    Date: 2004/01/26 15:15:15 CST
1165    Tag: blume-20040126-toplevel
1166    Description:
1167    
1168    Improved handling of exceptions at the interactive toplevel.
1169    
1170    ----------------------------------------------------------------------
1171    Name: Matthias Blume (blume (at) tti - c (dot) org)
1172    Date: 2004/01/26 11:25:00
1173    Tag: blume-20040126-app
1174    Description:
1175    
1176    Type of top-level "app" corrected.
1177    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1178    
1179    ----------------------------------------------------------------------
1180    Name: Matthias Blume (blume (at) tti - c (dot) org)
1181    Date: 2003/11/18 17:10 CST
1182    Tag: blume-20031118-basis-fiddle
1183    Description:
1184    
1185    - changed Timer interface to what might become the spec
1186    - POSIX_FLAGS -> BIT_FLAGS according to spec
1187    - some other minor discrepancies wrt. spec eliminated
1188    
1189    ----------------------------------------------------------------------
1190    Name: Matthias Blume (blume (at) tti - c (dot) org)
1191    Date: 2003/11/06 12:00:00 CST
1192    Tag: Release_110_44
1193    Description:
1194    
1195    New working version (110.44). New bootfiles.
1196    
1197    ----------------------------------------------------------------------
1198    Name: Matthias Blume (blume (at) tti - c (dot) org)
1199    Date: 2003/11/04 11:50:00 CST
1200    Tag: blume-20031104-move-libraries
1201    Description:
1202    
1203    Eliminated the "dont_move_libraries" directive in config/targets.
1204    (The mechanism was broken and could not be fixed easily.  Moreover,
1205    there does not seem to be any reason not to move all libraries into
1206    lib during installation.  I originally implemented this directive as a
1207    backward-compatibility feature when I first introduced the new CM.
1208    Now that things have been stable for a long time and going back to the
1209    old CM is not an option, there is no reason to keep it around.)
1210    
1211    ----------------------------------------------------------------------
1212    Name: Matthias Blume (blume (at) tti - c (dot) org)
1213    Date: 2003/11/03 16:00:00 CST
1214    Tag: blume-20031103-installdir
1215    Description:
1216    
1217    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1218    Richards for pointing out the problem and providing the solution.)
1219    
1220    ----------------------------------------------------------------------
1221    Name: Matthias Blume (blume (at) tti - c (dot) org)
1222    Date: 2003/10/01 17:05:00 CDT
1223    Tag: blume-20031001-lal-mlrisc
1224    Description:
1225    
1226    MLRISC bug fix from Lal.
1227    
1228    ----------------------------------------------------------------------
1229    Name: Matthias Blume (blume (at) tti - c (dot) org)
1230    Date: 2003/09/30 16:10:00 CDT
1231    Tag: blume-20030930-primio-bat
1232    Description:
1233    
1234    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1235    2. Improved .bat files (for Win32 port) to make things work under Win95.
1236       (thanks to Aaron S. Hawley for this one)
1237    
1238    ----------------------------------------------------------------------
1239    Name: Matthias Blume (blume (at) tti - c (dot) org)
1240    Date: 2003/09/26 16:05:00 CDT
1241    Tag: blume-20030926-wrappriv
1242    Description:
1243    
1244    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1245    
1246    ----------------------------------------------------------------------
1247    Name: Matthias Blume (blume (at) tti - c (dot) org)
1248    Date: 2003/09/26 15:00:00 CDT
1249    Tag: blume-20030926-110_43_3
1250    Description:
1251    
1252    - additional cleanup
1253    - version number bump, NEW BOOTFILES
1254    
1255    ----------------------------------------------------------------------
1256    Name: Matthias Blume (blume (at) tti - c (dot) org)
1257    Date: 2003/09/26 12:00:00 CDT
1258    Tag: blume-20030926-ppautoload
1259    Description:
1260    
1261    I modified the read-eval-print loop so that the autoloader gets
1262    invoked whenever the prettyprinter tries to look up a symbol that
1263    is not currently defined in the toplevel environment but which
1264    appears in CM's autoload registry.  As a result, we see far fewer of
1265    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1266    
1267    In addition to this I tried to clean up some pieces of the Basis
1268    implementation (e.g., Socket, Word8Array) in order to prevent other
1269    instances of these ?.Foo.Bar.xxx names from being printed.
1270    
1271    The mechanism that picks names for types still needs some work, though.
1272    (Right now it seems that if there is a type A.t which is defined to
1273    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1274    "?.B.u" although the perhaps more sensible solution would be to use
1275    "A.t" in this case.  In other words, the prettyprinter should follow
1276    a chain of DEFtycs not farther than there are corresponding toplevel
1277    names in the current environment.)
1278    
1279    ----------------------------------------------------------------------
1280    Name: Matthias Blume (blume (at) tti - c (dot) org)
1281    Date: 2003/09/24 16:31:00 CDT
1282    Tag: blume-20030924-installer
1283    Description:
1284    
1285    Another installer tweak:  All the ML code for the installer is now
1286    compiled during CMB.make and put into a little library called
1287    $smlnj/installer.cm.  The installation then simply invokes
1288    
1289       sml -m $smlnj/installer.cm
1290    
1291    and everything happens automagically.
1292    
1293    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1294    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1295            and BINDIR.
1296    
1297    The new scheme guarantees that the ML code responsible for the installation
1298    is in sync with the APIs of the main system.  Also, the installer is
1299    somewhat faster because the installer script is precompiled.
1300    
1301    ----------------------------------------------------------------------
1302    Name: Matthias Blume (blume (at) tti - c (dot) org)
1303    Date: 2003/09/24 15:35:00 CDT
1304    Tag: blume-20030924-synsock
1305    Description:
1306    
1307    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1308    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1309    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1310    implementations of non-blocking operations from CML's Socket
1311    structure.
1312    
1313    ----------------------------------------------------------------------
1314    Name: Matthias Blume (blume (at) tti - c (dot) org)
1315    Date: 2003/09/24 15:10:05 CDT
1316    Tag: blume-20030924-sockets
1317    Description:
1318    
1319    1. Fixed SOCKET API and implementation to match Basis spec.
1320       This required changing the internal representation of sockets to one
1321       that remembers (for each socket file descriptor) whether it is currently
1322       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1323       call is made only if the state actually needs to change).
1324    
1325    2. OS-specific details of sockets were moved into separate files, thus
1326       making it possible to unify the bulk of the socket implementations
1327       between Unix and Win32.
1328    
1329    3. CML's socket API changed accordingly.
1330       (Note that we need to remove non-blocking functions from this API
1331       since they are redundant in the case of CML!)
1332    
1333    4. CML's socket implementation now makes use of non-blocking functions
1334       provided by Basis, thus removing all OS-dependent code from this part
1335       of CML.
1336    
1337    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1338    
1339    ----------------------------------------------------------------------
1340    Name: Matthias Blume (blume (at) tti - c (dot) org)
1341    Date: 2003/09/22 12:10:00 CDT
1342    Tag: blume-20030922-110_43_2
1343    Description:
1344    
1345    Made a new interim version and bootfiles for developer's bootstrapping
1346    convenience.
1347    
1348       110.43.2 -- NEW BOOTFILES
1349    
1350    ----------------------------------------------------------------------
1351    Name: Matthias Blume (blume (at) tti - c (dot) org)
1352    Date: 2003/09/19 15:55:00 CDT
1353    Tag: blume-20030919-cmdir
1354    Description:
1355    
1356    1. new-install.sh -> install.sh
1357    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1358    3. tweaked installer so that another name instead of .cm can be chosen
1359       at install time (by setting the CM_DIR_ARC environment variable
1360       during installation); once installation is complete, the name is
1361       fixed
1362    
1363    ----------------------------------------------------------------------
1364    Name: Matthias Blume (blume (at) tti - c (dot) org)
1365    Date: 2003/09/18 16:00:00 CDT
1366    Tag: blume-20030918-110_43_1
1367    Description:
1368    
1369    Made a new interim version and bootfiles for developer's bootstrapping
1370    convenience.
1371    
1372       110.43.1 -- NEW BOOTFILES
1373    
1374    ----------------------------------------------------------------------
1375    Name: Matthias Blume (blume (at) tti - c (dot) org)
1376    Date: 2003/09/18 15:20:00 CDT
1377    Tag: blume-20030918-misc
1378    Description:
1379    
1380    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
1381       this seems to be controversial at the moment) and used those in
1382       Posix.ProcEnv.times.
1383    
1384    2. Added Time.{from,to}Nanoseconds to Time.
1385    
1386    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
1387       For example, fromLargeInt never needs to look at more than 3 "big
1388       digits" to get its 53 bits of precision.
1389    
1390    ----------------------------------------------------------------------
1391    Name: Matthias Blume (blume (at) tti - c (dot) org)
1392    Date: 2003/09/17 16:30:00 CDT
1393    Tag: blume-20030917-real32-slices
1394    Description:
1395    
1396    Added an entry to the primitive environment
1397    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
1398    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
1399    
1400    Removed some of the "magic" constants in real64.sml and replaced them
1401    with code that generates these values from their corresponding
1402    integer counterparts.
1403    
1404    Made all(?) the slice-related changes to the Basis and made everything
1405    compile again...
1406    
1407    ----------------------------------------------------------------------
1408    Name: Matthias Blume (blume (at) tti - c (dot) org)
1409    Date: 2003/09/15 17:45:00 CDT
1410    Tag: blume-20030915-rbase
1411    Description:
1412    
1413    Fixed bug in Real.fromLargeInt.
1414    
1415    ----------------------------------------------------------------------
1416    Name: Matthias Blume (blume (at) tti - c (dot) org)
1417    Date: 2003/09/13 18:11:00 CDT
1418    Tag: blume-20030913-libinstall
1419    Description:
1420    
1421    Minor bugfix in config/libinstall (set anchor with path to
1422    standalone tool after installing it, otherwise libraries that
1423    need ml-lex or ml-yacc won't compile the first time the installer
1424    runs).
1425    
1426    ----------------------------------------------------------------------
1427    Name: Matthias Blume (blume (at) tti - c (dot) org)
1428    Date: 2003/09/12 11:45:00 CDT
1429    Tag: blume-20030912-various
1430    Description:
1431    
1432     - fixed bug in Real.toLargeInt
1433     - fixed bug in Posix.ProcEnv.times
1434     - changed inputLine functions to return an option
1435     - minor installer improvements / bugfixes
1436     - changed default @SMLalloc parameter for x86/celeron to 64k
1437    
1438    ----------------------------------------------------------------------
1439    Name: Matthias Blume (blume (at) tti - c (dot) org)
1440    Date: 2003/09/09 22:00:00 CDT
1441    Tag: Release_110_43
1442    Description:
1443    
1444        New working release 110.43.  New bootfiles.
1445    
1446    ----------------------------------------------------------------------
1447    Name: Matthias Blume (blume (at) tti - c (dot) org)
1448    Date: 2003/09/09 19:20:00 CDT
1449    Tag: blume-20030909-installer
1450    Description:
1451    
1452    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
1453    Modified config/install.bat to take advantage of it.  Also modified
1454    config/install.sh (and called it config/new-install.sh) to take advantage
1455    of it on Unix systems.  (The SML code is (supposed to be) platform-
1456    independent.)
1457    
1458    The installer can now install everything under Win32
1459    as well as under *nix as long as it compiles.
1460    
1461    Other changes:
1462    
1463      - made CML compile again under Win32
1464      - made eXene compile under Win32 (by providing a fake structure UnixSock
1465        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
1466      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
1467        same as type int;  under Win32 it isn't
1468      - fixed some slice-related problems in the win32-specific parts of CML
1469      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
1470        (under Win32, the volume name is case-insensitive, and the
1471        OS.Path code compares volume names for equality)
1472    
1473    ----------------------------------------------------------------------
1474    Name: Matthias Blume (blume (at) tti - c (dot) org)
1475    Date: 2003/09/08 11:55:00 CDT
1476    Tag: blume-20030908-fullpath
1477    Description:
1478    
1479    Made Win32 version of OS.FileSys.fullPath return current directory
1480    when given an empty string.  This is what the spec says, and incidentally,
1481    CM depends on it.  (CM otherwise goes into an infinite loop in certain
1482    cases when presented with the name of a non-existing .cm file.)
1483    
1484    ----------------------------------------------------------------------
1485    Name: Matthias Blume (blume (at) tti - c (dot) org)
1486    Date: 2003/09/04 16:30:00 CDT
1487    Tag: blume-20030905-slices-etc
1488    Description:
1489    
1490      1. Changed interface to vectors and arrays in Basis to match
1491         (draft) Basis spec.
1492      2. Added signatures and implementations of slices according to
1493         Basis spec.
1494      3. Edited source code throughout the system to make it compile again
1495         under 1. and 2.  (In some cases code had to be added to have it
1496         match the new signatures.)
1497      4. MLRISC should be backward-compatible: the copies of the originals
1498         of files that needed to change under 3. were retained, the .cm files
1499         check the compiler version number and use old versions when
1500         appropriate.
1501      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
1502         dirstream -> string option (in accordance with Basis spec).
1503      6. When generating code that counts lines, ml-lex used function
1504         CharVector.foldli, taking advantage of its old interface.
1505         This has been replaced with the corresponding code from
1506         CharVectorSlice.  (html-lex must be re-lexed!)
1507      7. BitArray in smlnj-lib/Util has been extended/modified to match the
1508         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
1509      8. Removed temporary additions (fromInternal, toInternal) from the
1510         (now obsolete) IntInf in smlnj-lib/Util.
1511      9. Cleaned up structure Byte.
1512     10. Added localOffset, scan, and fromString to Date (according to spec).
1513         Cleaned/corrected implementation of Date.
1514         (Still need to check for correctness; implement better canonicalizeDate.)
1515     11. Added "scan" to signature IEEE_REAL.
1516     12. Some improvements to IntInf [in particular: efficiency-hack for
1517         mod and rem when second operand is 2 (for parity checks).]
1518     13. Changed representation of type Time.time, using a single IntInf.int
1519         value counting microseconds.  This considerably simplified the
1520         implementation of structure Time.  We now support negative time
1521         values; scan and fromString handle signs.
1522     14. Functor PrimIO now takes two additional arguments (VectorSlice and
1523         ArraySlice).
1524    
1525    ----------------------------------------------------------------------
1526    Name: Matthias Blume (blume (at) tti - c (dot) org)
1527    Date: 2003/08/28 17:00:00 CDT
1528    Tag: blume-20030828-intinf
1529    Description:
1530    
1531    This is a major update which comes with a version number bump
1532    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
1533    and an implementation of IntInf in the Basis.
1534    
1535    There are a fairly large number of related changes and updates throughout
1536    the system:
1537    
1538    Basis:
1539            - Implemented IntInf.
1540            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
1541            - Added some missing Real64 operations, most notably Real.toLargeInt.
1542            - Added FixedInt as a synonym for Int32.
1543    
1544    compiler:
1545            * Added support for a built-in intinf type.
1546              - literals
1547              - pattern matching
1548              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
1549              - overloading on literals and operations
1550    
1551            This required adding a primitive type intinf, some additional
1552            primops, and implementations for several non-trivial intinf
1553            operations in Core.  (The intinf type is completely abstract
1554            to the compiler; all operations get delegated back to the Core.)
1555    
1556            * Intinf equality is handled by polyequal.  However, the compiler
1557            does not print its usual warning in this case (since polyequal
1558            is the right thing to do there).
1559    
1560            * Improved the organization of structure InlineT.
1561    
1562            * A word about conversion primops:
1563            If conversions involving intinf do not cancel out during
1564            CPS contract, then the compiler must insert calls to Core functions.
1565            Since all core access must be resolved already during the FLINT
1566            translate phase, it would be too late a the time of CPS contract
1567            to add new Core calls.  For this reason, conversion primops
1568            for intinf carry two arguments:  1. the numeric argument that
1569            they are supposed to convert, and 2. the Core function that
1570            can help with this conversion if necessary.  If CPS contract
1571            eliminates a primop, then the associated Core function becomes
1572            dead and goes away.  Intinf conversion primops that do not get
1573            eliminated by CPS contract get rewritten into calls of their
1574            core functions by a separate, new phase.
1575    
1576    interactive system:
1577            - Control.Print.intinfDepth controls max length of intinf constants
1578            being printed. (Analogous to Control.Print.stringDepth.)
1579            - Cleanup in printutil and pputil: got rid of unused stuff and
1580            duplicates; replaced some of the code with code that makes better
1581            use of library functionality.
1582    
1583    CM:
1584            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
1585            are no longer silent.
1586    
1587    CKIT:
1588            Fixed mismatched uses of Int32 and LargeInt.  I always decided
1589            in favor of LargeInt -- which is now the same as IntInf.
1590            CKIT-knowledgable people should check whether this is what's
1591            intended and otherwise change things back to using Int32 or
1592            FixedInt.
1593    
1594    Throughout the code:
1595            Started using IntInf.int literals and built-in operations
1596            (e.g., comparison with 0) where this seems appropriate.
1597    
1598    
1599    ----------------------------------------------------------------------
1600    Name: Dave MacQueen (dbm@cs.uchicago.edu)
1601    Date: 2003/08/13 11:36:00 CDT
1602    Tag: dbm-20030813-mcz-merge1
1603    Description:
1604    
1605    Merging changes from the mcz-branch development branch into trunk.
1606    These changes involve replacement of the emulated old prettyprinter
1607    interface with direct use of the SML/NJ Lib PP library, and fixing
1608    of a couple of bugs (895, 1186) relating to error messages.  A new
1609    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
1610    has been added.
1611    
1612    ----------------------------------------------------------------------
1613    Name: Matthias Blume (blume (at) tti - c (dot) org)
1614    Date: 2003/08/11 15:45:00 CDT
1615    Tag: blume-20030811-windows
1616    Description:
1617    
1618    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
1619    
1620        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
1621    
1622    This patch restores SML/NJ's ability to run under win32.  There are a
1623    number of changes, including fixes for several bugs that had gone
1624    unnoticed until now:
1625    
1626      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
1627        trying to run the win32 version from within cygwin.)
1628    
1629      - There are a number of simple .bat scripts that substitute for their
1630        corresponding Unix shell-scripts.  (See below.)
1631    
1632      - The internals of ml-build have been modified slightly.  The main
1633        difference is that instead of calling ".link-sml" (or link-sml.bat)
1634        using OS.Process.system, the ML process delegates this task back
1635        to the script.  Otherwise problems arise in mixed environments such
1636        as Cygwin where scripts look and work like Unix scripts, but
1637        where OS.Process.system cannot run them.
1638    
1639      - In CM, the srcpath pickler used native pathname syntax -- which
1640        is incorrect in the case of cross-compilation.  The new pickle format
1641        is independent of platform-specific naming conventions.
1642    
1643      - Path configuration files (such as lib/pathconfig) can now choose
1644        between native and standard syntax.  Placing a line of the form
1645    
1646            standard!
1647    
1648        into the file causes all subsequent paths to be interpreted using
1649        CM standard pathname syntax (= Unix conventions); a line
1650    
1651            native!
1652    
1653        switches back to native style.  This was needed so that
1654        path config files can be written portably, see src/system/pathconfig.
1655    
1656      - Runtime system:
1657    
1658          - win32-filesys.c:  get_file_time and set_file_time now
1659            access modification time, not creation time.
1660    
1661          - I/O code made aware of new array representation.
1662    
1663          - Bug fixes in X86.prim.masm.
1664    
1665       - src/system/makeml made aware of win32. (For use under cygwin
1666            and other Unix-environments for windows.)
1667    
1668       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
1669         which caused BinIO.inputAll to fail consistently.
1670    
1671    .bat scripts:
1672    
1673       Windows .bat scripts assume that SMLNJ_HOME is defined.
1674    
1675       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
1676         applications (sml, ml-yacc, ml-lex).
1677       - ml-build.bat: analogous to ml-build.
1678       - config\install.bat:  Analogous to config/install.sh.  This requires
1679         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
1680         (nmake etc. must be on the path, and vcvars32 must have been run.)
1681         Moreover, sources for ml-lex and ml-yacc need to exist under src,
1682         and the bootfile hierarchy must have been unpacked under
1683         sml.boot.x86-win32.
1684         The script is very primitive and does a poor job at error checking.
1685         It only installs the base system, ml-lex, and ml-yacc.  No other
1686         libraries are being installed (i.e., you get only those that
1687         are part of the compiler.)
1688       - link-sml.bat:  analogous to .link-sml, but not currently used
1689    
1690    Unrelated bug fixes:
1691    
1692      - ml-nlffigen now exports structures ST_* corresponding to incomplete
1693        types.
1694      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
1695        otherwise.)
1696    
1697    ----------------------------------------------------------------------
1698    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
1699    Date: 2003/06/17
1700    Tag: macqueen-20030617-bug895
1701    Description:
1702    
1703    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
1704    Tag will be used for new development branch (mcz-branch) for
1705    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
1706    
1707    ----------------------------------------------------------------------
1708    Name: Matthias Blume (blume (at) tti - c (dot) org)
1709    Date: 2003/05/27 16:55:00 CDT
1710    Tag: blume-20030527-polyeq
1711    Description:
1712    
1713    Tried to eliminated most cases of polymorphic equality.
1714    
1715    ----------------------------------------------------------------------
1716    Name: Matthias Blume (blume (at) tti - c (dot) org)
1717    Date: 2003/05/21 17:45:00 CDT
1718    Tag: blume-20030517-complete
1719    Description:
1720    
1721    Two changes:
1722    
1723      1. Added a flag for controlling whether non-exhaustive bindings will
1724         be treated as errors (default is false).
1725      2. Cleaned up the *entire* source tree so that CMB.make goes through
1726         without a single non-exhaustive match- or bind warning.
1727    
1728    ----------------------------------------------------------------------
1729    Name: Matthias Blume (blume (at) tti - c (dot) org)
1730    Date: 2003/05/17 10:20:00 CDT
1731    Tag: blume-20030517-absyn
1732    Description:
1733    
1734    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
1735    
1736       This mainly affects the quality of error messages.  However, some
1737       of the code is now more straightforward than before. (Treatment of
1738       the above four constructs in translate.sml is much simpler than
1739       the "macro-expansion" that was going on before.  Plus, the mach-
1740       compiler no longer gets invoked just to be able to compile an
1741       if-expression.)
1742    
1743    2. The ErrorMsg.Error exception is now caught and absorbed by the
1744       interactive loop.
1745    
1746    ----------------------------------------------------------------------
1747    Name: Allen Leung
1748    Date: 2003/05/16 13:05:00 CDT
1749    Tag: leunga-20030516-cygwin-runtime
1750    Description:
1751    
1752        Ported the runtime system to cygwin, which uses the unix
1753    x86-unix bin files.  Missing/buggy features:
1754    
1755       o getnetbyname, getnetbyaddr: these functions seem to be missing in
1756         the Cygwin library.
1757       o Ctrl-C handling may be flaky.
1758       o Windows system calls and Windows I/O are not supported.
1759    
1760       A new set of binfiles is located at:
1761    
1762         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
1763    
1764       This is only needed for bootstrapping the cygwin version of smlnj.
1765       Other x86 versions can use the existing binfiles.
1766    
1767    ----------------------------------------------------------------------
1768    Name: Matthias Blume
1769    Date: 2003/04/08 15:42:00 CDT
1770    Tag: blume-20030408-listpair
1771    Description:
1772    
1773    1. Added a target 'mlrisc' to installer.
1774    
1775    2. Added missing elements to structure ListPair.
1776    
1777    ----------------------------------------------------------------------
1778    Name: Allen Leung
1779    Date: 2003/01/07 10:40:00 EST
1780    Tag: leunga-20030107-int-rem
1781    Description:
1782    
1783    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
1784    The arguments to the SUBL instruction were swapped.
1785    
1786    ----------------------------------------------------------------------
1787    Name: Matthias Blume
1788    Date: 2002/12/12 16:25:00 EST
1789    Tag: blume-20021212-risc-ra
1790    Description:
1791    
1792    Fixed a serious bug in the rewrite code for FP spilling/reloading that
1793    sent the RA into an infinite loop when floating point registers get
1794    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
1795    110.37 and 110.38.)
1796    There was another set of potential problems related to the handling of
1797    MLRISC annotations (but those did not yet cause real problems, apparently).
1798    
1799    ----------------------------------------------------------------------
1800    Name: Matthias Blume
1801    Date: 2002/12/06 22:40:00 EST
1802    Tag: blume-20021206-cm-fileid
1803    Description:
1804    
1805    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
1806    This fixes the problem of CM getting confused by files that suddenly
1807    change their identity (e.g., by getting unlinked and recreated by some
1808    text editor such as vi).  There might be a better/cheaper/cleaner way
1809    of doing this, but for now this will have to do.
1810    
1811    ----------------------------------------------------------------------
1812    Name: Matthias Blume
1813    Date: 2002/10/28 09:50:00 EST
1814    Tag: blume-20021028-typecheck
1815    Description:
1816    
1817    Exported structure Typecheck from $smlnj/viscomp/core.cm.
1818    
1819    ----------------------------------------------------------------------
1820    Name: Matthias Blume
1821    Date: 2002/10/17 09:10:00 EDT
1822    Tag: Release_110_42
1823    Description:
1824    
1825    In good old tradition, there has been a slight hiccup so that we have
1826    to patch 110.42 after the fact.   The old release tag has been replaced
1827    (see below).
1828    
1829    The change solves a problem with two competing approaches the
1830    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
1831    each other's way.
1832    
1833    This change only affects the runtime system code and the installer script.
1834    (No new bootfiles.)
1835    
1836    ----------------------------------------------------------------------
1837    Name: Matthias Blume
1838    Date: 2002/10/16 12:00:00 EDT
1839    Tag: Release_110_42_removed
1840    Description:
1841    
1842    New working release.  New bootfiles.
1843    
1844    ----------------------------------------------------------------------
1845    Name: Matthias Blume
1846    Date: 2002/10/10 13:10:00 EDT
1847    Tag: blume-20021010-ppc-divs
1848    Description:
1849    
1850    The mltree operator DIVS must be implemented with an overflow check on
1851    the PPC because the hardware indicates divide-by-zero using "overflow" as
1852    well.
1853    
1854    ----------------------------------------------------------------------
1855    Name: Matthias Blume
1856    Date: 2002/07/23 11:45:00
1857    Tag: blume-20020723-smlnj-home
1858    Description:
1859    
1860    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
1861    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
1862    CM_PATHCONFIG is also set), the path configuration file is assumed
1863    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
1864    the entire tree to some other place and everything will "just work".
1865    
1866    (Companion commands such as ml-build and ml-makedepend also sense this
1867    variable.)
1868    
1869    ----------------------------------------------------------------------
1870    Name: Matthias Blume
1871    Date: 2002/07/12 21:19:00 EDT
1872    Tag: blume-20020712-liveness
1873    Description:
1874    
1875    Exported two useful "step" functions from liveness module (MLRISC).
1876    
1877    ----------------------------------------------------------------------
1878    Name: Matthias Blume
1879    Date: 2002/07/05 16:00 EDT
1880    Tag: Release_110_41
1881    Description:
1882    
1883    New working release.  New bootfiles.
1884    
1885    ----------------------------------------------------------------------
1886    Name: Matthias Blume
1887    Date: 2002/07/05 10:25:00 EDT
1888    Tag: blume-20020705-btimp
1889    Description:
1890    
1891    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
1892    other clients can set up backtracing support.
1893    
1894    ----------------------------------------------------------------------
1895    Name: Matthias Blume
1896    Date: 2002/06/25 17:23:00 EDT
1897    Tag: blume-20020625-fpmax
1898    Description:
1899    
1900    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
1901    case.  (The sense of the isNaN test was reversed -- which made min and
1902    max always return their first argument.)
1903    
1904    ----------------------------------------------------------------------
1905    Name: Matthias Blume
1906    Date: 2002/06/11
1907    Tag: blume-20020611-unixpath
1908    Description:
1909    
1910    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
1911    
1912    ----------------------------------------------------------------------
1913    Name: Matthias Blume
1914    Date: 2002/06/10 16:35:00 EDT
1915    Tag: blume-20020610-ieeereal
1916    Description:
1917    
1918    I back-ported my implementation of IEEEReal.fromString from the
1919    idlbasis-devel branch so that we can test it.
1920    
1921    Another small change is that ppDec tries to give more information
1922    than just "<sig>" in the case of functors.  However, this code is
1923    broken in some mysterious way if the functor's body's signature
1924    has not been declared by ascription but gets inferred from the
1925    implementation.  This needs fixing...
1926    
1927    ----------------------------------------------------------------------
1928    Name: Matthias Blume
1929    Date: 2002/05/31
1930    Tag: blume-20020531-btrace-mode
1931    Description:
1932    
1933    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
1934    the wayside when I switched over to using Controls everywhere.)
1935    
1936    ----------------------------------------------------------------------
1937    Name: Lal George
1938    Date: 2002/05/23 12:21:40 EDT
1939    Tag: george-20020523-visual-labels
1940    Description:
1941    
1942    Labels are now displayed in the graphical output to make
1943    the fall-through and target blocks obvious.
1944    
1945    ----------------------------------------------------------------------
1946    Name: Matthias Blume
1947    Date: 2002/05/22 11:03:00 EDT
1948    Tag: blume-20020522-shrink
1949    Description:
1950    
1951    John tweaked yesterday's fix for 1131 to handle an out-of-memory
1952    situation that comes up when allocating huge arrays.
1953    
1954    ----------------------------------------------------------------------
1955    Name: Matthias Blume
1956    Date: 2002/05/21 16:00:00 EDT
1957    Tag: Release_110_40
1958    Description:
1959    
1960    New working release (110.40).  New bootfiles.
1961    
1962    [Also: John Reppy fixed GC bug 1131.]
1963    
1964    ----------------------------------------------------------------------
1965    Name: Matthias Blume
1966    Date: 2002/05/21 12:35:00 EDT
1967    Tag: blume-20020521-cmdoc
1968    Description:
1969    
1970    CM documentation update.
1971    
1972    ----------------------------------------------------------------------
1973    Name: Matthias Blume
1974    Date: 2002/05/21 10:55:00 EDT
1975    Tag: blume-20020521-misc
1976    Description:
1977    
1978    - John tweaked runtime to be silent on heap export (except when
1979      GC messages are on).
1980    - I added a few more things (cross-compiling versions of CMB) to
1981      config/preloads (as suggestions).
1982    
1983    ----------------------------------------------------------------------
1984    Name: Matthias Blume
1985    Date: 2002/05/20 22:25:00 EDT
1986    Tag: blume-20020520-controls
1987    Description:
1988    
1989    - Added ControlUtil structure to control-lib.cm.
1990    - Use it throughout.
1991    - Used Controls facility to define MLRISC controls (as opposed to
1992      registering MLRISC control ref cells with Controls after the fact)
1993    - Fixed messed-up controls priorities.
1994    
1995    * Removed again all the stuff from config/preloads that one wouldn't
1996      be able to preload at the time the initial heap image is built.
1997      (Many libraries, e.g., CML, do not exist yet at this time.  The
1998       only libraries that can be preloaded via config/preloads are those
1999       that come bundled with the bootfiles.)
2000    
2001    ----------------------------------------------------------------------
2002    Name: Matthias Blume
2003    Date: 2002/05/20 10:59:00 EDT
2004    Tag: blume-20020520-preloads
2005    Description:
2006    
2007    Added a lot of commented-out suggestions for things to be included
2008    in config/preloads.
2009    
2010    ----------------------------------------------------------------------
2011    Name: Allen Leung
2012    Date: 2002/05/18 14:20:00 EDT
2013    Tag: leunga-20020518-mdl
2014    Description:
2015    
2016    o Made the mdl tool stuff compile and run again.
2017    o I've disabled all the stuff that depends on RTL specifications; they
2018      are all badly broken anyway.
2019    
2020    ----------------------------------------------------------------------
2021    Name: Matthias Blume
2022    Date: 2002/05/17 16:49:00 EDT
2023    Tag: blume-20020517-controls
2024    Description:
2025    
2026    0. John Reppy made several modifications to the SML/NJ library.
2027       In particular, there is a shiny new controls-lib.cm.
2028    
2029    1. Pushed new controls interface through compiler so that everything
2030       compiles again.
2031    
2032    2. Added FormatComb and FORMAT_COMB to the CML version of the
2033       SML/NJ library (so that CML compiles again).
2034    
2035    3. Modified init scripts because XXX_DEFAULT environment variables
2036       are no longer with us.  (Boot-time initialization is now done
2037       using the same environment variables that are also used for
2038       startup-time initialization of controls.)
2039    
2040    ----------------------------------------------------------------------
2041    Name: Lal George
2042    Date: 2002/05/15 09:20:10 EDT
2043    Tag: george-20020515-pseudo-op-decls
2044    Description:
2045    
2046            All pseudo-ops emitted before the first segment declaration
2047    such as TEXT, DATA, and BSS directives are assumed to be global
2048    declarations and are emitted first in the assembly file. This is
2049    useful in a number of situations where one has pseudo-ops that are not
2050    specific to any segment, and also works around the constraint that one
2051    cannot have client pseudo-ops in the TEXT segment.
2052    
2053            Because no segment is associated with these declarations it is
2054    an error to allocate any space or objects before the first segment
2055    directive and an exception will be raised. However, we cannot make
2056    this check for client pseudo-ops.
2057    
2058            These top level declarations are a field in the CFG graph_info.
2059    In theory you can continue to add to this field after the CFG has been
2060    built -- provided you know what you are doing;-)
2061    
2062    ----------------------------------------------------------------------
2063    Name: Matthias Blume
2064    Date: 2002/05/13 16:40:00 EDT
2065    Tag: blume-20020513-pp-etc
2066    Description:
2067    
2068    A few minor bugfixes:
2069    
2070      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2071        (Bogus printouts for redefined bindings still occur.  Compiler
2072        bug should no longer occur now.  We need to redo the prettyprinter
2073        from scratch.)
2074    
2075      - CM pathname printer now also adds escape sequences for ( and )
2076    
2077      - commend and docu fixes for ml-nlffi
2078    
2079    ----------------------------------------------------------------------
2080    Name: Matthias Blume
2081    Date: 2002/05/10 16:40:00 EDT
2082    Tag: blume-20020510-erg-textio
2083    Description:
2084    
2085    Applied the following bugfix provided by Emden Gansner:
2086    
2087        Output is corrupted when outputSubstr is used rather than output.
2088    
2089        The problem occurs when a substring
2090    
2091            ss = (s, dataStart, dataLen)
2092    
2093        where dataStart > 0, fills a stream buffer with avail bytes left.
2094        avail bytes of s, starting at index dataStart, are copied into the
2095        buffer, the buffer is flushed, and then the remaining dataLen-avail
2096        bytes of ss are copied into the beginning of the buffer. Instead of
2097        starting this copy at index dataStart+avail in s, the current code
2098        starts the copy at index avail.
2099    
2100        Fix:
2101        In text-io-fn.sml, change line 695 from
2102             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2103        to
2104             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2105    
2106    ----------------------------------------------------------------------
2107    Name: Matthias Blume
2108    Date: 2002/04/12 13:55:00 EDT
2109    Tag: blume-20020412-assyntax
2110    Description:
2111    
2112    1. Grabbed newer assyntax.h from the XFree86 project.
2113    2. Fiddled with how to compile X86.prim.asm without warnings.
2114    3. (Very) Minor cleanup in CM.
2115    
2116    ----------------------------------------------------------------------
2117    Name: Matthias Blume
2118    Date: 2002/04/01 (no joke!) 17:07:00 EST
2119    Tag: blume-20020401-x86div
2120    Description:
2121    
2122    Added full support for div/mod/rem/quot on the x86, using the machine
2123    instruction's two results (without clumsily recomputing the remainder)
2124    directly where appropriate.
2125    
2126    Some more extensive power-of-two support was added to the x86 instruction
2127    selector (avoiding expensive divs, mods, and muls where they can be
2128    replaced with cheaper shifts and masks).  However, this sort of thing
2129    ought to be done earlier, e.g., within the CPS optimizer so that
2130    all architectures benefit from it.
2131    
2132    The compiler compiles to a fixed point, but changes might be somewhat
2133    fragile nevertheless.  Please, report any strange things that you might
2134    see wrt. div/mod/quot/rem...
2135    
2136    ----------------------------------------------------------------------
2137    Name: Matthias Blume
2138    Date: 2002/03/29 17:22:00
2139    Tag: blume-20020329-div
2140    Description:
2141    
2142    Fixed my broken div/mod logic.  Unfortunately, this means that the
2143    inline code for div/mod now has one more comparison than before.
2144    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2145    The problem was with quotient = 0, because that alone does not tell
2146    us which way the rounding went.  One then has to look at whether
2147    remainder and divisor have the same sign...  :(
2148    
2149    Anyway, I replaced the bootfiles with fresh ones...
2150    
2151    ----------------------------------------------------------------------
2152    Name: Matthias Blume
2153    Date: 2002/03/29 14:10:00 EST
2154    Tag: blume-20020329-inlprims
2155    Description:
2156    
2157    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2158    
2159    Primops have changed. This means that the bin/boot-file formats have
2160    changed as well.
2161    
2162    To make sure that there is no confusion, I made a new version.
2163    
2164    
2165    CHANGES:
2166    
2167    * removed REMT from mltree (remainder should never overflow).
2168    
2169    * added primops to deal with divisions of all flavors to the frontend
2170    
2171    * handled these primops all the way through so they map to their respective
2172      MLRISC support
2173    
2174    * used these primops in the implementation of Int, Int32, Word, Word32
2175    
2176    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2177    
2178    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2179    
2180    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2181      floating point
2182    
2183    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2184      and Real (but Real.abs maps to a separate floating-point-only primop)
2185    
2186    
2187    TODO items:
2188    
2189    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2190      instructions because the machine instruction encoder cannot handle
2191      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2192      REM in the first place, and REM is supported by the x86 machine coder.)
2193    
2194    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2195      selection phase.  (The two can be streamlined because the hardware
2196      delivers both quotient and remainder at the same time anyway.)
2197    
2198    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2199      (Currently the behavior is inconsistent both across architectures and
2200      wrt. the draft Basis spec.)
2201    
2202    * Word8 should eventually be handled natively, too.
2203    
2204    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2205      as if there currently is no execution path that could trigger it in
2206      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2207      hold for things like multiplication and division.)
2208    
2209    ----------------------------------------------------------------------
2210    Name: Matthias Blume
2211    Date: 2002/03/27 16:27:00 EST
2212    Tag: blume-20020327-mlrisc-divisions
2213    Description:
2214    
2215    Added support for all four division operations (ML's div, mod, quot,
2216    and rem) to MLRISC.  In the course of doing so, I also rationalized
2217    the naming (no more annoying switch-around of DIV and QUOT), by
2218    parameterizing the operation by div_rounding_mode (which can be either
2219    DIV_TO_ZERO or DIV_TO_NEGINF).
2220    
2221    The generic MLTreeGen functor takes care of compiling all four
2222    operations down to only round-to-zero div.
2223    
2224    Missing pieces:
2225    
2226      * Doing something smarter than relying on MLTreeGen on architectures
2227        like, e.g., the x86 where hardware division delivers both quotient and
2228        remainder at the same time.  With this, the implementation of the
2229        round-to-neginf operations could be further streamlined.
2230    
2231      * Remove inlining support for div/mod/rem from the frontend and replace it
2232        with primops that get carried through to the backend.  Do this for all
2233        int and word types.
2234    
2235    ----------------------------------------------------------------------
2236    Name: Matthias Blume
2237    Date: 2002/03/25 17:25:00 EST
2238    Tag: blume-20020325-divmod
2239    Description:
2240    
2241    I improved (hopefully without breaking them) the implementation of Int.div,
2242    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2243    advantage of the following observations:
2244    
2245      Let  q = x quot y      r = x rem y
2246           d = x div  y      m = x mod y
2247    
2248    where "quot" is the round-to-zero version of integer division that
2249    hardware usually provides.  Then we have:
2250    
2251         r = x - q * y        where neither the * nor the - will overflow
2252         d = if q >= 0 orelse x = q * y then q else q - 1
2253                              where neither the * nor the - will overflow
2254         m = if q >= 0 orelse r = 0 then r else r + y
2255                              where the + will not overflow
2256    
2257    This results in substantial simplification of the generated code.
2258    The following table shows the number of CFG nodes and edges generated
2259    for
2260            fun f (x, y) = x OPER y
2261            (* with OPER \in div, mod, quot, rem *)
2262    
2263    
2264        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2265        --------------------------------------------------------
2266         div |         24 |         39 |         12 |         16
2267         mod |         41 |         71 |         12 |         16
2268        quot |          8 |         10 |          8 |         10
2269         rem |         10 |         14 |          8 |         10
2270    
2271    
2272    ----------------------------------------------------------------------
2273    Name: Matthias Blume
2274    Date: 2002/03/25 22:06:00 EST
2275    Tag: blume-20020325-cprotobug
2276    Description:
2277    
2278    Fixed a bug in cproto (c prototype decoder).
2279    
2280    ----------------------------------------------------------------------
2281    Name: Matthias Blume
2282    Date: 2002/03/25 16:00:00 EST
2283    Tag: blume-20020325-raw-primops
2284    Description:
2285    
2286    I did some cleanup to Allen's new primop code and
2287    replaced yesterday's bootfiles with new ones.
2288    (But they are stored in the same place.)
2289    
2290    ----------------------------------------------------------------------
2291    Name: Matthias Blume
2292    Date: 2002/03/24 22:40:00 EST
2293    Tag: blume-20020324-bootfiles
2294    Description:
2295    
2296    Made the bootfiles that Allen asked for.
2297    
2298    ----------------------------------------------------------------------
2299    Name: Allen Leung
2300    Date: 2002/03/23 15:50:00 EST
2301    Tag: leunga-20020323-flint-cps-rcc-primops
2302    Description:
2303    
2304      1. Changes to FLINT primops:
2305    
2306        (* make a call to a C-function;
2307         * The primop carries C function prototype information and specifies
2308         * which of its (ML-) arguments are floating point. C prototype
2309         * information is for use by the backend, ML information is for
2310         * use by the CPS converter. *)
2311      | RAW_CCALL of { c_proto: CTypes.c_proto,
2312                       ml_args: ccall_type list,
2313                       ml_res_opt: ccall_type option,
2314                       reentrant : bool
2315                     } option
2316       (* Allocate uninitialized storage on the heap.
2317        * The record is meant to hold short-lived C objects, i.e., they
2318        * are not ML pointers.  With the tag, the representation is
2319        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2320        *)
2321      | RAW_RECORD of {tag:bool,sz:int}
2322      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2323    
2324      2.  These CPS primops are now overloaded:
2325    
2326           rawload of {kind:numkind}
2327           rawstore of {kind:numkind}
2328    
2329          The one argument form is:
2330    
2331             rawload {kind} address
2332    
2333          The two argument form is:
2334    
2335             rawload {kind} [ml object, byte-offset]
2336    
2337      3. RAW_CCALL/RCC now takes two extra arguments:
2338    
2339         a. The first is whether the C call is reentrant, i.e., whether
2340            ML state should be saved and restored.
2341         b. The second argument is a string argument specifying the name of
2342            library and the C function.
2343    
2344         These things are currently not handled in the code generator, yet.
2345    
2346      4. In CProto,
2347    
2348         An encoding type of "bool" means "ml object" and is mapped into
2349         C prototype of PTR.  Note that "bool" is different than "string",
2350         even though "string" is also mapped into PTR, because "bool"
2351         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2352    
2353      5. Pickler/unpicker
2354    
2355         Changed to handle RAW_RECORD and newest RAW_CCALL
2356    
2357      6. MLRiscGen,
2358    
2359         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2360         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2361            in the file CodeGen/cpscompile/cps-c-calls.sml
2362    
2363      7. Added the conditional move operator
2364    
2365             condmove of branch
2366    
2367         to cps.  Generation of this is still buggy so it is currently
2368         disabled.
2369    
2370    ----------------------------------------------------------------------
2371    Name: Lal George
2372    Date: 2002/03/22 14:18:25 EST
2373    Tag: george-20020322-cps-branch-prob
2374    Description:
2375    
2376    Implemented the Ball-Larus branch prediction-heuristics, and
2377    incorporated graphical viewers for control flow graphs.
2378    
2379    Ball-Larus Heuristics:
2380    ---------------------
2381    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
2382    
2383    By design it uses the Dempster-Shafer theory for combining
2384    probabilities.  For example, in the function:
2385    
2386        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
2387    
2388    the ball-larus heuristics predicts that the n=0 is unlikely
2389    (OH-heuristic), and the 'then' branch is unlikely because of the
2390    RH-heuristic -- giving the 'then' branch an even lower combined
2391    probability using the Dempster-Shafer theory.
2392    
2393    Finally, John Reppy's loop analysis in MLRISC, further lowers the
2394    probability of the 'then' branch because of the loop in the else
2395    branch.
2396    
2397    
2398    Graphical Viewing:
2399    ------------------
2400    I merely plugged in Allen's graphical viewers into the compiler. The
2401    additional code is not much. At the top level, saying:
2402    
2403            Control.MLRISC.getFlag "cfg-graphical-view" := true;
2404    
2405    will display the graphical view of the control flow graph just before
2406    back-patching.  daVinci must be in your path for this to work. If
2407    daVinci is not available, then the default viewer can be changed
2408    using:
2409    
2410            Control.MLRISC.getString "viewer"
2411    
2412    which can be set to "dot" or "vcg" for the corresponding viewers. Of
2413    course, these viewers must be in your path.
2414    
2415    The above will display the compilation unit at the level of clusters,
2416    many of which are small, boring, and un-interesting. Also setting:
2417    
2418            Control.MLRISC.getInt "cfg-graphical-view_size"
2419    
2420    will display clusters that are larger than the value set by the above.
2421    
2422    
2423    ----------------------------------------------------------------------
2424    Name: Matthias Blume
2425    Date: 2002/03/21 22:20:00 EST
2426    Tag: blume-20020321-kmp-bugfix
2427    Description:
2428    
2429    Changed the interface to the KMP routine in PreString and fixed
2430    a minor bug in one place where it was used.
2431    
2432    ----------------------------------------------------------------------
2433    Name: Allen Leung
2434    Date: 2002/03/21 20:30:00 EST
2435    Tag: leunga-20020321-cfg
2436    Description:
2437    
2438      Fixed a potential problem in cfg edge splitting.
2439    
2440    ----------------------------------------------------------------------
2441    Name: Allen Leung
2442    Date: 2002/03/21 17:15:00 EST
2443    Tag: leunga-20020321-x86-fp-cfg
2444    Description:
2445    
2446      1. Recoded the buggy parts of x86-fp.
2447    
2448         a. All the block reordering code has been removed.
2449            We now depend on the block placement phases to do this work.
2450    
2451         b. Critical edge splitting code has been simplified and moved into the
2452            CFG modules, as where they belong.
2453    
2454         Both of these were quite buggy and complex.  The code is now much, much
2455         simpler.
2456    
2457      2. X86 backend.
2458    
2459         a. Added instructions for 64-bit support.  Instruction selection for
2460            64-bit has not been committed, however, since that
2461            requires changes to MLTREE which haven't been approved by
2462            Lal and John.
2463    
2464         b. Added support for FUCOMI and FUCOMIP when generating code for
2465            PentiumPro and above.  We only generate these instructions in
2466            the fast-fp mode.
2467    
2468         c. Added cases for JP and JNP in X86FreqProps.
2469    
2470      3. CFG
2471    
2472         CFG now has a bunch of methods for edge splitting and merging.
2473    
2474      4. Machine description.
2475    
2476         John's simplification of MLTREE_BASIS.fcond broke a few machine
2477         description things:
2478    
2479         rtl-build.{sig,sml} and hppa.mdl fixed.
2480    
2481         NOTE: the machine description stuff in the repository is still broken.
2482               Again, I can't put my fixes in because that involves
2483               changes to MLTREE.
2484    
2485    ----------------------------------------------------------------------
2486    Name: Matthias Blume
2487    Date: 2002/03/20 15:55:00 EST
2488    Tag: blume-20020320-kmp
2489    Description:
2490    
2491    Implemented Knuth-Morris-Pratt string matching in PreString and used
2492    it for String.isSubstring, Substring.isSubstring, and
2493    Substring.position.
2494    
2495    (Might need some stress-testing.  Simple examples worked fine.)
2496    
2497    ----------------------------------------------------------------------
2498    Name: Matthias Blume
2499    Date: 2002/03/19 16:37:00 EST
2500    Tag: blume-20020319-witnesses
2501    Description:
2502    
2503    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
2504    
2505    This implements a generic mechanism for changing constness qualifiers
2506    anywhere within big C types without resorting to outright "casts".
2507    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
2508    modify the constness at the outermost level.)
2509    The implementation of "convert" is based on the idea of "witness"
2510    values -- values that are not used by the operation but whose types
2511    "testify" to their applicability.  On the implementation side, "convert"
2512    is simply a projection (returning its second curried argument).  With
2513    cross-module inlining, it should not result in any machine code being
2514    generated.
2515    
2516    ----------------------------------------------------------------------
2517    Name: Matthias Blume
2518    Date: 2002/03/15 16:40:00 EST
2519    Tag: blume-20020315-basis
2520    Description:
2521    
2522    Provided (preliminary?) implementations for
2523    
2524      {String,Substring}.{concatWith,isSuffix,isSubstring}
2525    
2526    and
2527    
2528      Substring.full
2529    
2530    Those are in the Basis spec but they were missing in SML/NJ.
2531    
2532    ----------------------------------------------------------------------
2533    Name: Matthias Blume
2534    Date: 2002/03/14 21:30:00 EST
2535    Tag: blume-20020314-controls
2536    Description:
2537    
2538    Controls:
2539    ---------
2540    
2541    1. Factored out the recently-added Controls : CONTROLS stuff and put
2542       it into its own library $/controls-lib.cm.  The source tree for
2543       this is under src/smlnj-lib/Controls.
2544    
2545    2. Changed the names of types and functions in this interface, so they
2546       make a bit more "sense":
2547    
2548          module -> registry
2549          'a registry -> 'a group
2550    
2551    3. The interface now deals in ref cells only.  The getter/setter interface
2552       is (mostly) gone.
2553    
2554    4. Added a function that lets one register an already-existing ref cell.
2555    
2556    5. Made the corresponding modifications to the rest of the code so that
2557       everything compiles again.
2558    
2559    6. Changed the implementation of Controls.MLRISC back to something closer
2560       to the original.  In particular, this module (and therefore MLRISC)
2561       does not depend on Controls.  There now is some link-time code in
2562       int-sys.sml that registers the MLRISC controls with the Controls
2563       module.
2564    
2565    CM:
2566    ---
2567    
2568      * One can now specify the lambda-split aggressiveness in init.cmi.
2569    
2570    ----------------------------------------------------------------------
2571    Name: Allen Leung
2572    Date: 2002/03/13 17:30:00 EST
2573    Tag: leunga-20020313-x86-fp-unary
2574    Description:
2575    
2576    Bug fix for:
2577    
2578    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
2579    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
2580    > - fun f(x,(y,z)) = Real.~ y;
2581    > [autoloading]
2582    > [autoloading done]
2583    >       fchsl   (%eax), 184(%esp)
2584    > Error: MLRisc bug: X86MCEmitter.emitInstr
2585    >
2586    > uncaught exception Error
2587    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
2588    
2589    The problem was that the code generator did not generate any fp registers
2590    in this case, and the ra didn't know that it needed to run the X86FP phase to
2591    translate the pseudo fp instruction.   This only happened with unary fp
2592    operators in certain situations.
2593    
2594    ----------------------------------------------------------------------
2595    Name: Matthias Blume
2596    Date: 2002/03/13 14:00:00 EST
2597    Tag: blume-20020313-overload-etc
2598    Description:
2599    
2600    1. Added _overload as a synonym for overload for backward compatibility.
2601       (Control.overloadKW must be true for either version to be accepted.)
2602    
2603    2. Fixed bug in install script that caused more things to be installed
2604       than what was requested in config/targets.
2605    
2606    3. Made CM aware of the (_)overload construct so that autoloading
2607       works.
2608    
2609    ----------------------------------------------------------------------
2610    Name: Matthias Blume
2611    Date: 2002/03/12 22:03:00 EST
2612    Tag: blume-20020312-url
2613    Description:
2614    
2615    Forgot to update BOOT and srcarchiveurl.
2616    
2617    ----------------------------------------------------------------------
2618    Name: Matthias Blume
2619    Date: 2002/03/12 17:30:00 EST
2620    Tag: blume-20020312-version110392
2621    Description:
2622    
2623    Yet another version number bump (because of small changes to the
2624    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
2625    
2626    Changes:
2627    
2628      The new pid generation scheme described a few weeks ago was overly
2629      complicated.  I implemented a new mechanism that is simpler and
2630      provides a bit more "stability":  Once CM has seen a compilation
2631      unit, it keeps its identity constant (as long as you do not delete
2632      those crucial CM/GUID/* files).  This means that when you change
2633      an interface, compile, then go back to the old interface, and
2634      compile again, you arrive at the original pid.
2635    
2636      There now also is a mechanism that instructs CM to use the plain
2637      environment hash as a module's pid (effectively making its GUID
2638      the empty string).  For this, "noguid" must be specified as an
2639      option to the .sml file in question within its .cm file.
2640      This is most useful for code that is being generated by tools such
2641      as ml-nlffigen (because during development programmers tend to
2642      erase the tool's entire output directory tree including CM's cached
2643      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
2644      revert to the old, broken behavior of SML/NJ, but in specific cases
2645      where there is no danger of interface confusion, its use is ok
2646      (I think).
2647    
2648      ml-nlffigen by default generates "noguid" annotations.  They can be
2649      turned off by specifying -guid in its command line.
2650    
2651    ----------------------------------------------------------------------
2652    Name: Lal George
2653    Date: 2002/03/12 12 14:42:36 EST
2654    Tag: george-20020312-frequency-computation
2655    Description:
2656    
2657    Integrated jump chaining and static block frequency into the
2658    compiler. More details and numbers later.
2659    
2660    ----------------------------------------------------------------------
2661    Name: Lal George
2662    Date: 2002/03/11 11 22:38:53 EST
2663    Tag: george-20020311-jump-chain-elim
2664    Description:
2665    
2666    Tested the jump chain elimination on all architectures (except the
2667    hppa).  This is on by default right now and is profitable for the
2668    alpha and x86, however, it may not be profitable for the sparc and ppc
2669    when compiling the compiler.
2670    
2671    The gc test will typically jump to a label at the end of the cluster,
2672    where there is another jump to an external cluster containing the actual
2673    code to invoke gc. This is to allow factoring of common gc invocation
2674    sequences. That is to say, we generate:
2675    
2676            f:
2677               testgc
2678               ja   L1      % jump if above to L1
2679    
2680            L1:
2681               jmp L2
2682    
2683    
2684    After jump chain elimination the 'ja L1' instructions is converted to
2685    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
2686    up being implemented in their long form (if L2 is far away) using:
2687    
2688            jbe     L3      % jump if below or equal to L3
2689            jmp     L2
2690         L3:
2691            ...
2692    
2693    
2694    For large compilation units L2  may be far away.
2695    
2696    
2697    ----------------------------------------------------------------------
2698    Name: Matthias Blume
2699    Date: 2002/03/11 13:30:00 EST
2700    Tag: blume-20020311-mltreeeval
2701    Description:
2702    
2703    A functor parameter was missing.
2704    
2705    ----------------------------------------------------------------------
2706    Name: Allen Leung
2707    Date: 2002/03/11 10:30:00 EST
2708    Tag: leunga-20020311-runtime-string0
2709    Description:
2710    
2711       The representation of the empty string now points to a
2712    legal null terminated C string instead of unit.  It is now possible
2713    to convert an ML string into C string with InlineT.CharVector.getData.
2714    This compiles into one single machine instruction.
2715    
2716    ----------------------------------------------------------------------
2717    Name: Allen Leung
2718    Date: 2002/03/10 23:55:00 EST
2719    Tag: leunga-20020310-x86-call
2720    Description:
2721    
2722       Added machine generation for CALL instruction (relative displacement mode)
2723    
2724    ----------------------------------------------------------------------
2725    Name: Matthias Blume
2726    Date: 2002/03/08 16:05:00
2727    Tag: blume-20020308-entrypoints
2728    Description:
2729    
2730    Version number bumped to 110.39.1.  NEW BOOTFILES!
2731    
2732    Entrypoints: non-zero offset into a code object where execution should begin.
2733    
2734    - Added the notion of an entrypoint to CodeObj.
2735    - Added reading/writing of entrypoint info to Binfile.
2736    - Made runtime system bootloader aware of entrypoints.
2737    - Use the address of the label of the first function given to mlriscGen
2738      as the entrypoint.  This address is currently always 0, but it will
2739      not be 0 once we turn on block placement.
2740    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
2741      with entry points) from mlriscGen.
2742    
2743    ----------------------------------------------------------------------
2744    Name: Allen Leung
2745    Date: 2002/03/07 20:45:00 EST
2746    Tag: leunga-20020307-x86-cmov
2747    Description:
2748    
2749       Bug fixes for CMOVcc on x86.
2750    
2751       1. Added machine code generation for CMOVcc
2752       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
2753       3. CMOVcc cannot have an immediate operand as argument.
2754    
2755    ----------------------------------------------------------------------
2756    Name: Matthias Blume
2757    Date: 2002/03/07 16:15:00 EST
2758    Tag: blume-20020307-controls
2759    Description:
2760    
2761    This is a very large but mostly boring patch which makes (almost)
2762    every tuneable compiler knob (i.e., pretty much everything under
2763    Control.* plus a few other things) configurable via both the command
2764    line and environment variables in the style CM did its configuration
2765    until now.
2766    
2767    Try starting sml with '-h' (or, if you are brave, '-H')
2768    
2769    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
2770    implements the underlying generic mechanism.
2771    
2772    The interface to some of the existing such facilities has changed somewhat.
2773    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
2774    (The getFoo interface is still there for backward-compatibility, but its
2775    use is deprecated.)
2776    
2777    The ml-build script passes -Cxxx=yyy command-line arguments through so
2778    that one can now twiddle the compiler settings when using this "batch"
2779    compiler.
2780    
2781    TODO items:
2782    
2783    We should go through and throw out all controls that are no longer
2784    connected to anything.  Moreover, we should go through and provide
2785    meaningful (and correct!) documentation strings for those controls
2786    that still are connected.
2787    
2788    Currently, multiple calls to Controls.new are accepted (only the first
2789    has any effect).  Eventually we should make sure that every control
2790    is being made (via Controls.new) exactly once.  Future access can then
2791    be done using Controls.acc.
2792    
2793    Finally, it would probably be a good idea to use the getter-setter
2794    interface to controls rather than ref cells.  For the time being, both
2795    styles are provided by the Controls module, but getter-setter pairs are
2796    better if thread-safety is of any concern because they can be wrapped.
2797    
2798    *****************************************
2799    
2800    One bug fix: The function blockPlacement in three of the MLRISC
2801    backpatch files used to be hard-wired to one of two possibilities at
2802    link time (according to the value of the placementFlag).  But (I
2803    think) it should rather sense the flag every time.
2804    
2805    *****************************************
2806    
2807    Other assorted changes (by other people who did not supply a HISTORY entry):
2808    
2809    1. the cross-module inliner now works much better (Monnier)
2810    2. representation of weights, frequencies, and probabilities in MLRISC
2811       changed in preparation of using those for weighted block placement
2812       (Reppy, George)
2813    
2814    ----------------------------------------------------------------------
2815    Name: Lal George
2816    Date: 2002/03/07 14:44:24 EST 2002
2817    Tag: george-20020307-weighted-block-placement
2818    
2819    Tested the weighted block placement optimization on all architectures
2820    (except the hppa) using AMPL to generate the block and edge frequencies.
2821    Changes were required in the machine properties to correctly
2822    categorize trap instructions. There is an MLRISC flag
2823    "weighted-block-placement" that can be used to enable weighted block
2824    placement, but this will be ineffective without block/edge
2825    frequencies (coming soon).
2826    
2827    
2828    ----------------------------------------------------------------------
2829    Name: Lal George
2830    Date: 2002/03/05 17:24:48 EST
2831    Tag: george-20020305-linkage-cluster
2832    
2833    In order to support the block placement optimization, a new cluster
2834    is generated as the very first cluster (called the linkage cluster).
2835    It contains a single jump to the 'real' entry point for the compilation
2836    unit. Block placement has no effect on the linkage cluster itself, but
2837    all the other clusters  have full freedom in the manner in which they
2838    reorder blocks or functions.
2839    
2840    On the x86 the typical linkage code that is generated is:
2841       ----------------------
2842            .align 2
2843       L0:
2844            addl    $L1-L0, 72(%esp)
2845            jmp     L1
2846    
2847    
2848            .align  2
2849       L1:
2850       ----------------------
2851    
2852    72(%esp) is the memory location for the stdlink register. This
2853    must contain the address of the CPS function being called. In the
2854    above example, it contains the address of  L0; before
2855    calling L1 (the real entry point for the compilation unit), it
2856    must contain the address for L1, and hence
2857    
2858            addl $L1-L0, 72(%esp)
2859    
2860    I have tested this on all architectures except the hppa.The increase
2861    in code size is of course negligible
2862    
2863    ----------------------------------------------------------------------
2864    Name: Allen Leung
2865    Date: 2002/03/03 13:20:00 EST
2866    Tag: leunga-20020303-mlrisc-tools
2867    
2868      Added #[ ... ] expressions to mlrisc tools
2869    
2870    ----------------------------------------------------------------------
2871    Name: Matthias Blume
2872    Date: 2002/02/27 12:29:00 EST
2873    Tag: blume-20020227-cdebug
2874    Description:
2875    
2876    - made types in structure C and C_Debug to be equal
2877    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
2878    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
2879    
2880    ----------------------------------------------------------------------
2881    Name: Matthias Blume
2882    Date: 2002/02/26 12:00:00 EST
2883    Tag: blume-20020226-ffi
2884    Description:
2885    
2886    1. Fixed a minor bug in CM's "noweb" tool:
2887       If numbering is turned off, then truly don't number (i.e., do not
2888       supply the -L option to noweb).  The previous behavior was to supply
2889       -L'' -- which caused noweb to use the "default" line numbering scheme.
2890       Thanks to Chris Richards for pointing this out (and supplying the fix).
2891    
2892    2. Once again, I reworked some aspects of the FFI:
2893    
2894       A. The incomplete/complete type business:
2895    
2896       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
2897         gone!
2898       - ML types representing an incomplete type are now *equal* to
2899         ML types representing their corresponding complete types (just like
2900         in C).  This is still safe because ml-nlffigen will not generate
2901         RTTI for incomplete types, nor will it generate functions that
2902         require access to such RTTI.   But when ML code generated from both
2903         incomplete and complete versions of the C type meet, the ML types
2904         are trivially interoperable.
2905    
2906         NOTE:  These changes restore the full generality of the translation
2907         (which was previously lost when I eliminated functorization)!
2908    
2909       B. Enum types:
2910    
2911       - Structure C now has a type constructor "enum" that is similar to
2912         how the "su" constructor works.  However, "enum" is not a phantom
2913         type because each "T enum" has values (and is isomorphic to
2914         MLRep.Signed.int).
2915       - There are generic access operations for enum objects (using
2916         MLRep.Signed.int).
2917       - ml-nlffigen will generate a structure E_foo for each "enum foo".
2918         * The structure contains the definition of type "mlrep" (the ML-side
2919         representation type of the enum).  Normally, mlrep is the same
2920         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
2921         then mlrep will be defined as a datatype -- thus facilitating
2922         pattern matching on mlrep values.
2923         ("-ec" will be suppressed if there are duplicate values in an
2924          enumeration.)
2925         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
2926         will be generated for each C enum constant xxx.
2927         * Conversion functions m2i and i2m convert between mlrep and
2928         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
2929         * Coversion functions c and ml convert between mlrep and "tag enum".
2930         * Access functions (get/set) fetch and store mlrep values.
2931       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
2932         enumerations are merged into one single enumeration represented by
2933         structure E_'.
2934    
2935    ----------------------------------------------------------------------
2936    Name: Allen Leung
2937    Date: 2002/02/25 04:45:00 EST
2938    Tag: leunga-20020225-cps-spill
2939    
2940    This is a new implementation of the CPS spill phase.
2941    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
2942    In case of problems, replace it with the old file spill.sml
2943    
2944    The current compiler runs into some serious performance problems when
2945    constructing a large record.  This can happen when we try to compile a
2946    structure with many items.  Even a very simple structure like the following
2947    makes the compiler slow down.
2948    
2949        structure Foo = struct
2950           val x_1 = 0w1 : Word32.int
2951           val x_2 = 0w2 : Word32.int
2952           val x_3 = 0w3 : Word32.int
2953           ...
2954           val x_N = 0wN : Word32.int
2955        end
2956    
2957    The following table shows the compile time, from N=1000 to N=4000,
2958    with the old compiler:
2959    
2960    N
2961    1000   CPS 100 spill                           0.04u  0.00s  0.00g
2962           MLRISC ra                               0.06u  0.00s  0.05g
2963              (spills = 0 reloads = 0)
2964           TOTAL                                   0.63u  0.07s  0.21g
2965    
2966    1100   CPS 100 spill                           8.25u  0.32s  0.64g
2967           MLRISC ra                               5.68u  0.59s  3.93g
2968              (spills = 0 reloads = 0)
2969           TOTAL                                   14.71u  0.99s  4.81g
2970    
2971    1500   CPS 100 spill                           58.55u  2.34s  1.74g
2972           MLRISC ra                               5.54u  0.65s  3.91g
2973              (spills = 543 reloads = 1082)
2974           TOTAL                                   65.40u  3.13s  6.00g
2975    
2976    2000   CPS 100 spill                           126.69u  4.84s  3.08g
2977           MLRISC ra                               0.80u  0.10s  0.55g
2978              (spills = 42 reloads = 84)
2979           TOTAL                                   129.42u  5.10s  4.13g
2980    
2981    3000   CPS 100 spill                           675.59u  19.03s  11.64g
2982           MLRISC ra                               2.69u  0.27s  1.38g
2983              (spills = 62 reloads = 124)
2984           TOTAL                                   682.48u  19.61s  13.99g
2985    
2986    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
2987           MLRISC ra                               4.96u  0.27s  2.72g
2988              (spills = 85 reloads = 170)
2989           TOTAL                                   2375.26u  57.21s  48.00g
2990    
2991    As you can see the old cps spill module suffers from some serious
2992    performance problem.  But since I cannot decipher the old code fully,
2993    instead of patching the problems up, I'm reimplementing it
2994    with a different algorithm.  The new code is more modular,
2995    smaller when compiled, and substantially faster
2996    (O(n log n) time and O(n) space).  Timing of the new spill module:
2997    
2998    4000  CPS 100 spill                           0.02u  0.00s  0.00g
2999          MLRISC ra                               0.25u  0.02s  0.15g
3000             (spills=1 reloads=3)
3001          TOTAL                                   7.74u  0.34s  1.62g
3002    
3003    Implementation details:
3004    
3005    As far as I can tell, the purpose of the CPS spill module is to make sure the
3006    number of live variables at any program point (the bandwidth)
3007    does not exceed a certain limit, which is determined by the
3008    size of the spill area.
3009    
3010    When the bandwidth is too large, we decrease the register pressure by
3011    packing live variables into spill records.  How we achieve this is
3012    completely different than what we did in the old code.
3013    
3014    First, there is something about the MLRiscGen code generator
3015    that we should be aware of:
3016    
3017    o MLRiscGen performs code motion!
3018    
3019       In particular, it will move floating point computations and
3020       address computations involving only the heap pointer to
3021       their use sites (if there is only a single use).
3022       What this means is that if we have a CPS record construction
3023       statement
3024    
3025           RECORD(k,vl,w,e)
3026    
3027       we should never count the new record address w as live if w
3028       has only one use (which is often the case).
3029    
3030       We should do something similar to floating point, but the transformation
3031       there is much more complex, so I won't deal with that.
3032    
3033    Secondly, there are now two new cps primops at our disposal:
3034    
3035     1. rawrecord of record_kind option
3036        This pure operator allocates some uninitialized storage from the heap.
3037        There are two forms:
3038    
3039         rawrecord NONE [INT n]  allocates a tagless record of length n
3040         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
3041                                     and initializes the tag.
3042    
3043     2. rawupdate of cty
3044          rawupdate cty (v,i,x)
3045          Assigns to x to the ith component of record v.
3046          The storelist is not updated.
3047    
3048    We use these new primops for both spilling and increment record construction.
3049    
3050     1. Spilling.
3051    
3052        This is implemented with a linear scan algorithm (but generalized
3053        to trees).  The algorithm will create a single spill record at the
3054        beginning of the cps function and use rawupdate to spill to it,
3055        and SELECT or SELp to reload from it.  So both spills and reloads
3056        are fine-grain operations.  In contrast, in the old algorithm
3057        "spills" have to be bundled together in records.
3058    
3059        Ideally, we should sink the spill record construction to where
3060        it is needed.  We can even split the spill record into multiple ones
3061        at the places where they are needed.  But CPS is not a good
3062        representation for global code motion, so I'll keep it simple and
3063        am not attempting this.
3064    
3065     2. Incremental record construction (aka record splitting).
3066    
3067        Long records with many component values which are simulatenously live
3068        (recall that single use record addresses are not considered to
3069         be live) are constructed with rawrecord and rawupdate.
3070        We allocate space on the heap with rawrecord first, then gradually
3071        fill it in with rawupdate.  This is the technique suggested to me
3072        by Matthias.
3073    
3074        Some restrictions on when this is applicable:
3075        1. It is not a VECTOR record.  The code generator currently does not handle
3076           this case. VECTOR record uses double indirection like arrays.
3077        2. All the record component values are defined in the same "basic block"
3078           as the record constructor.  This is to prevent speculative
3079           record construction.
3080    
3081    ----------------------------------------------------------------------
3082    Name: Allen Leung
3083    Date: 2002/02/22 01:02:00 EST
3084    Tag: leunga-20020222-mlrisc-tools
3085    
3086    Minor bug fixes in the parser and rewriter
3087    
3088    ----------------------------------------------------------------------
3089    Name: Allen Leung
3090    Date: 2002/02/21 20:20:00 EST
3091    Tag: leunga-20020221-peephole
3092    
3093    Regenerated the peephole files.  Some contained typos in the specification
3094    and some didn't compile because of pretty printing bugs in the old version
3095    of 'nowhere'.
3096    
3097    ----------------------------------------------------------------------
3098    Name: Allen Leung
3099    Date: 2002/02/19 20:20:00 EST
3100    Tag: leunga-20020219-mlrisc-tools
3101    Description:
3102    
3103       Minor bug fixes to the mlrisc-tools library:
3104    
3105       1.  Fixed up parsing colon suffixed keywords
3106       2.  Added the ability to shut the error messages up
3107       3.  Reimplemented the pretty printer and fixed up/improved
3108           the pretty printing of handle and -> types.
3109       4.  Fixed up generation of literal symbols in the nowhere tool.
3110       5.  Added some SML keywords to to sml.sty
3111    
3112    ----------------------------------------------------------------------
3113    Name: Matthias Blume
3114    Date: 2002/02/19 16:20:00 EST
3115    Tag: blume-20020219-cmffi
3116    Description:
3117    
3118    A wild mix of changes, some minor, some major:
3119    
3120    * All C FFI-related libraries are now anchored under $c:
3121        $/c.cm      --> $c/c.cm
3122        $/c-int.cm  --> $c/internals/c-int.cm
3123        $/memory.cm --> $c/memory/memory.cm
3124    
3125    * "make" tool (in CM) now treats its argument pathname slightly
3126      differently:
3127        1. If the native expansion is an absolute name, then before invoking
3128           the "make" command on it, CM will apply OS.Path.mkRelative
3129           (with relativeTo = OS.FileSys.getDir()) to it.
3130        2. The argument will be passed through to subsequent phases of CM
3131           processing without "going native".  In particular, if the argument
3132           was an anchored path, then "make" will not lose track of that anchor.
3133    
3134    * Compiler backends now "know" their respective C calling conventions
3135      instead of having to be told about it by ml-nlffigen.  This relieves
3136      ml-nlffigen from one of its burdens.
3137    
3138    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
3139    
3140    * Export C_DEBUG and C_Debug from $c/c.cm.
3141    
3142    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
3143      subtyping relationship between incomplete pointers and their complete
3144      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
3145      with the convention of instantiating 'o with ('t, 'c) obj whenever
3146      the pointer target type is complete.  In the incomplete case, 'o
3147      will be instantiated with some "'c iobj" -- a type obtained by
3148      using one of the functors PointerToIncompleteType or PointerToCompleteType.
3149    
3150      Operations that work on both incomplete and complete pointer types are
3151      typed as taking an 'o ptr while operations that require the target to
3152      be known are typed as taking some ('t, 'c) obj ptr.
3153    
3154      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
3155      where void is an eqtype without any values.  This makes it possible
3156      to work on voidptr values using functions meant to operate on light
3157      incomplete pointers.
3158    
3159    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
3160      been vastly simplified.
3161    
3162    ----------------------------------------------------------------------
3163    Name: Matthias Blume
3164    Date: 2002/02/19 10:48:00 EST
3165    Tag: blume-20020219-pqfix
3166    Description:
3167    
3168    Applied Chris Okasaki's bug fix for priority queues.
3169    
3170    ----------------------------------------------------------------------
3171    Name: Matthias Blume
3172    Date: 2002/02/15 17:05:00
3173    Tag: Release_110_39
3174    Description:
3175    
3176    Last-minute retagging is becoming a tradition... :-(
3177    
3178    This is the working release 110.39.
3179    
3180    ----------------------------------------------------------------------
3181    Name: Matthias Blume
3182    Date: 2002/02/15 16:00:00 EST
3183    Tag: Release_110_39-orig
3184    Description:
3185    
3186    Working release 110.39.  New bootfiles.
3187    
3188    (Update: There was a small bug in the installer so it wouldn't work
3189    with all shells.  So I retagged. -Matthias)
3190    
3191    ----------------------------------------------------------------------
3192    Name: Matthias Blume
3193    Date: 2002/02/15 14:17:00 EST
3194    Tag: blume-20020215-showbindings
3195    Description:
3196    
3197    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
3198    the latter can be useful for exploring what bindings are available at
3199    the interactive prompt.  (The first function returns only the list
3200    of symbols that are really bound, the second prints those but also the
3201    ones that CM's autoloading mechanism knows about.)
3202    
3203    ----------------------------------------------------------------------
3204    Name: Matthias Blume
3205    Date: 2002/02/15 12:08:00 EST
3206    Tag: blume-20020215-iptrs
3207    Description:
3208    
3209    Two improvements to ml-nlffigen:
3210    
3211      1. Write files only if they do not exist or if their current contents
3212         do not coincide with what's being written.  (That is, avoid messing
3213         with the time stamps unless absolutely necessary.)
3214    
3215      2. Implement a "repository" mechanism for generated files related
3216         to "incomplete pointer types".   See the README file for details.
3217    
3218    ----------------------------------------------------------------------
3219    Name: Matthias Blume
3220    Date: 2002/02/14 11:50:00 EST
3221    Tag: blume-20020214-quote
3222    Description:
3223    
3224    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
3225    because of the new and improved tag generation scheme.  (Thanks to Allen
3226    Leung for pointing it out.)
3227    
3228    ----------------------------------------------------------------------
3229    Name: Lal George
3230    Date: 2002/02/14 09:55:27 EST 2002
3231    Tag: george-20020214-isabelle-bug
3232    Description:
3233    
3234    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
3235    of Isabelle on the x86.
3236    
3237    From Allen:
3238    -----------
3239     I've found the problem:
3240    
3241         in ra-core.sml, I use the counter "blocked" to keep track of the
3242         true number of elements in the freeze queue.  When the counter goes
3243         to zero, I skip examining the queue.  But I've messed up the
3244         bookkeeping in combine():
3245    
3246             else ();
3247             case !ucol of
3248               PSEUDO => (if !cntv > 0 then
3249                     (if !cntu > 0 then blocked := !blocked - 1 else ();
3250                                        ^^^^^^^^^^^^^^^^^^^^^^^
3251                      moveu := mergeMoveList(!movev, !moveu)
3252                     )
3253                  else ();
3254    
3255         combine() is called to coalesce two nodes u and v.
3256         I think I was thinking that if the move counts of u and v are both
3257         greater than zero then after they are coalesced then one node is
3258         removed from the freeze queue.  Apparently I was thinking that
3259         both u and v are of low degree, but that's clearly not necessarily true.
3260    
3261    
3262    02/12/2002:
3263        Here's the patch.  HOL now compiles.
3264    
3265        I don't know how this impact on performance (compile
3266        time or runtime).  This bug caused the RA (especially on the x86)
3267        to go thru the potential spill phase when there are still nodes on the
3268        freeze queue.
3269    
3270    
3271    
3272    
3273    ----------------------------------------------------------------------
3274    Name: Matthias Blume
3275    Date: 2002/02/13 22:40:00 EST
3276    Tag: blume-20020213-fptr-rtti
3277    Description:
3278    
3279    Fixed a bug in ml-nlffigen that was introduced with one of the previous
3280    updates.
3281    
3282    ----------------------------------------------------------------------
3283    Name: Matthias Blume
3284    Date: 2002/02/13 16:41:00 EST
3285    Tag: blume-20020213-cmlpq
3286    Description:
3287    
3288    Added new priority queue export symbols (which have just been added to
3289    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
3290    not compile and the installer would choke.)
3291    
3292    ----------------------------------------------------------------------
3293    Name: Matthias Blume
3294    Date: 2002/02/13 16:15:00 EST
3295    Tag: blume-20020213-various
3296    Description:
3297    
3298    1. More tweaks to ml-nlffigen:
3299    
3300       - better internal datastructures (resulting in slight speedup)
3301       - "-match" option requires exact match
3302       - "localized" gensym counters (untagged structs/unions nested within
3303         other structs/unions or within typedefs get a fresh counter; their
3304         tag will be prefixed by a concatenation of their parents' tags)
3305       - bug fixes (related to calculation of transitive closure of types
3306         to be included in the output)
3307    
3308    2. Minor Basis updates:
3309    
3310       - added implementations for List.collate and Option.app
3311    
3312    ----------------------------------------------------------------------
3313    Name: Matthias Blume
3314    Date: 2002/02/11 15:55:00 EST
3315    Tag: blume-20020211-gensym
3316    Description:
3317    
3318    Added a "-gensym" option to command line of ml-nlffigen.  This can be
3319    used to specify a "stem" -- a string that is inserted in all "gensym'd"
3320    names (ML structure names that correspond to unnamed C structs, unions,
3321    and enums), so that separate runs of ml-nlffigen do not clash.
3322    
3323    ----------------------------------------------------------------------
3324    Name: Matthias Blume
3325    Date: 2002/02/11 12:05:00 EST
3326    Tag: blume-20020211-gensml
3327    Description:
3328    
3329    A quick fix for a problem with GenSML (in the pgraph-util library):
3330    Make generation of toplevel "local" optional.  (Strictly speaking,
3331    signature definitions within "local" are not legal SML.)
3332    
3333    Other than that: updates to INSTALL and cm/TODO.
3334    
3335    ----------------------------------------------------------------------
3336    Name: Matthias Blume
3337    Date: 2002/02/08 15:00:00 EST
3338    Tag: blume-20020208-uniquepid
3339    Description:
3340    
3341    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
3342    
3343    1. The installer (config/install.sh) has gotten smarter:
3344    
3345         - Configuration options are a bit easier to specify now
3346           (in config/targets).
3347         - Bug in recognizing .tar.bz2 files fixed.
3348         - Installer automatically resolves dependencies between
3349           configuration options (e.g., if you ask for eXene, you will
3350           also get cml -- regardless whether you asked for it or not).
3351         - Installer can run in "quieter mode" by setting the environment
3352           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
3353           "completely silent", though.
3354         - Build HashCons library as part of smlnj-lib.
3355    
3356    2. A new scheme for assigning persistent identifiers to compilation
3357       units (and, by extension, to types etc.) has been put into place.
3358       This fixes a long-standing bug where types and even dynamic values
3359       can get internally confused, thereby compromising type safety
3360       (abstraction) and dynamic correctness.  See
3361    
3362         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
3363    
3364       for an example of how things could go wrong until now.
3365    
3366       The downside of the new scheme is that pids are not quite as
3367       persistent as they used to be: CM will generate a fresh pid
3368       for every compilation unit that it thinks it sees for the first
3369       time.  That means that if you compile starting from a clean, fresh
3370       source tree at two different times, you end up with different
3371       binaries.
3372    
3373       Cutoff recompilation, however, has not been compromised because
3374       CM keeps pid information in special caches between runs.
3375    
3376    ----------------------------------------------------------------------
3377    Name: Lal George
3378    Date: 2002/02/07 15:34:13 EST 2002
3379    Tag: <none>
3380    Description:
3381    
3382    Compilers that generate assembly code may produce  global labels
3383    whose value is resolved at link time. The various peephole optimization
3384    modules did not take this in account.
3385    
3386    TODO. The Labels.addrOf function should really return an option
3387    type so that clients are forced to deal with this issue, rather
3388    than an exception being raised.
3389    
3390    ----------------------------------------------------------------------
3391    Name: Lal George
3392    Date: 2002/02/06 13:55:02 EST
3393    Tag: george-20020206-ra-breakup
3394    Description:
3395    
3396    1. A bug fix from Allen.
3397    
3398        A typo causes extra fstp %st(0)'s to be generated at compensation
3399        edges, which might cause stack underflow traps at runtime.  This
3400        occurs in fft where there are extraneous fstps right before the 'into'
3401        trap instruction (in this case they are harmless since none of the
3402        integers overflow.)
3403    
3404    2. Pulled out various utility modules that were embedded in the modules
3405       of the register allocator. I need these modules for other purposes, but
3406       they are not complete enough to put into a library (just yet).
3407    ----------------------------------------------------------------------
3408    Name: Matthias Blume
3409    Date: 2002/01/31 16:05:00 EST
3410    Tag: blume-20020131-sparc-ccalls
3411    Description:
3412    
3413    1. C-calls on Sparc needlessly allocated a huge chunk (96 bytes)
3414       of extra stack space by mistake.  Fixed.
3415    
3416    2. Bug in logic of handling of command-line options in ml-nlffigen fixed.
3417    
3418    ----------------------------------------------------------------------
3419    Name: Allen Leung
3420    Date: 2002/01/30
3421    Tag: leunga-20020130-nowhere-bug-fix
3422    Description:
3423    
3424       MLRISC bug fixes:
3425       1. Fixed a bindings computation bug in the 'nowhere' program generator tool.
3426       2. MachineInt.fromString was negating its value.
3427    
3428    ----------------------------------------------------------------------
3429    Name: Matthias Blume
3430    Date: 2002/01/29
3431    Tag: blume-20020129-INSTALL
3432    Description:
3433    
3434    - Added somewhat detailed installation instructions (file INSTALL).
3435    - Fixed curl-detection bug in config/install.sh.
3436    - It is now possible to select the URL getter using the URLGETTER
3437      environment variable:
3438    
3439          not set / "unknown"      --> automatic detection (script tries wget,
3440                                       curl, and lynx)
3441          "wget" / "curl" / "lynx" --> use the specified program (script "knows"
3442                                       how to properly invoke them)
3443          other                    --> use $URLGETTER directly, it must take
3444                                       precisely two command-line arguments
3445                                       (source URL and destination file name)
3446    
3447    ----------------------------------------------------------------------
3448    Name: Matthias Blume
3449    Date: 2002/01/28
3450    Tag: blume-20020128-sparc-ccalls
3451    Description:
3452    
3453    - Fixed problem with calculation of "used" registers in sparc-c-calls.
3454    - Make use of the allocParam argument in sparc-c-calls.
3455    
3456    ----------------------------------------------------------------------
3457    Name: Matthias Blume
3458    Date: 2002/01/28
3459    Tag: blume-20020128-allocParam
3460    Description:
3461    
3462    John Reppy:  Changes c-calls API to accept client-callback for
3463    allocating extra stack space.
3464    me: Corresponding changes to mlriscGen (using a dummy argument that
3465        does not change the current behavior).
3466    
3467    ----------------------------------------------------------------------
3468    Name: Matthias Blume
3469    Date: 2002/01/28 12:00:00
3470    Tag: Release_110_38
3471    Description:
3472    
3473    This time for real!!!
3474    
3475    ----------------------------------------------------------------------
3476    Name: Matthias Blume
3477    Date: 2002/01/28 10:56:00 EST
3478    Tag: blume-20020128-retraction
3479    Description:
3480    
3481    0. Retracted earlier 110.38.  (The Release_110_38 tag has been replaced
3482       with blume-Release_110_38-retracted.)
3483    
3484    1. Fixed a problem with incorrect rounding modes in real64.sml.
3485       (Thanks to Andrew Mccreight <andrew.mccreight@yale.edu>.)
3486    
3487    2. A bug in ml-nlffigen related to the handling of unnamed structs, unions,
3488       and enums fixed.  The naming of corresponding ML identifiers should
3489       now be consistent again.
3490    
3491    ----------------------------------------------------------------------
3492    Name: Allen Leung
3493    Date: 2002/01/27
3494    Tag: leunga-20020127-nowhere
3495    Description:
3496    
3497       Added a target called nowhere in the configuration scripts.
3498       Enabling this will build the MLRISC 'nowhere' tool (for translating
3499       programs with where-clauses into legal SML code) during installation.
3500    
3501    ----------------------------------------------------------------------
3502    Name: Matthias Blume
3503    Date: 2002/01/25 21:27:00 EST
3504    Tag: blume-Release_110_38-retracted
3505    Description:
3506    
3507    Call it a (working) release!  Version is 110.38. Bootfiles are ready.
3508    
3509    README will be added later.
3510    
3511    !!! NOTE:  Re-tagged as blume-Release_110_38-retracted. Original tag
3512    (Release_110_38) removed.  Reason: Last-minute bug fixes.
3513    
3514    ----------------------------------------------------------------------
3515    Name: Matthias Blume
3516    Date: 2002/01/25
3517    Tag: blume-20020125-ffi
3518    Description:
3519    
3520    A large number of tweaks and improvements to ml-nlffi-lib and
3521    ml-nlffigen:
3522    
3523       - ML represenation types have been streamlined
3524       - getter and setter functions work with concrete values, not abstract
3525         ones where possible
3526       - ml-nlffigen command line more flexible (see README file there)
3527       - some bugs have been fixed (hopefully)
3528    
3529    ----------------------------------------------------------------------
3530    Name: Lal George
3531    Date: 2002/01/24
3532    Tag: george-20020124-risc-ra-interface
3533    Description:
3534    
3535       There is a dramatic simplification in the interface to the
3536       register allocator for RISC architectures as a result of making
3537       parallel copy instructions explicit.
3538    
3539    ----------------------------------------------------------------------
3540    Name: Matthias Blume
3541    Date: 2002/01/22
3542    Tag: blume-20020122-x86-ccalls
3543    Description:
3544    
3545    Bug fix for c-calls on x86 (having to do with how char- and
3546    short-arguments are being handled).
3547    
3548    ----------------------------------------------------------------------
3549    Name: Matthias Blume
3550    Date: 2002/01/21
3551    Tag: blume-20020121-ff
3552    Description:
3553    
3554    Another day of fiddling with the FFI...
3555    
3556    1. Bug fix/workaround:  CKIT does not complain about negative array
3557       dimensions, so ml-nlffigen has to guard itself against this possibility.
3558       (Otherwise a negative dimension would send it into an infinite loop.)
3559    
3560    2. Some of the abstract types (light objects, light pointers, most "base"
3561       types) in structure C are now eqtypes.
3562    
3563    3. Added constructors and test functions for NULL function pointers.
3564    
3565    ----------------------------------------------------------------------
3566    Name: Matthias Blume
3567    Date: 2002/01/18
3568    Tag: blume-20020118-ready-for-new-release
3569    Description:
3570    
3571    Made config/srcarchiveurl point to a new place.  (Will provide boot
3572    files shortly.)
3573    
3574    Maybe we christen this to be 110.38?
3575    
3576    ----------------------------------------------------------------------
3577    Name: Matthias Blume
3578    Date: 2002/01/18
3579    Tag: blume-20020118-more-ffifiddle
3580    Description:
3581    
3582    Today's FFI fiddling:
3583    
3584      - Provided a structure CGetSet with "convenient" versions of C.Get.* and
3585        C.Set.* that use concrete (MLRep.*) arguments and results instead
3586        of abstract ones.
3587    
3588      - Provided word-style bit operations etc. for "int" representation
3589        types in MLRep.S<Foo>Bitops where <Foo> ranges over Char, Int, Short,
3590        and Long.
3591    
3592    ----------------------------------------------------------------------
3593    Name: Matthias Blume
3594    Date: 2002/01/18
3595    Tag: blume-20020118-use-x86-fp
3596    Description:
3597    
3598    Now that x86-fast-fp seems to be working, I turned it back on again
3599    by default.  (Seems to work fine now, even with the FFI.)
3600    
3601    Other than that, I added some documentation about the FFI to
3602    src/ml-nlffigen/README and updated the FFI test examples in
3603    src/ml-nlffi-lib/Tests/*.
3604    
3605    ----------------------------------------------------------------------
3606    Name: Allen Leung
3607    Date: 2002/01/17
3608    Tag: leunga-20020117-x86-fast-fp-call
3609    Description:
3610    
3611       1. Fixed a problem with handling return fp values when x86's fast fp
3612          mode is turned on.
3613    
3614       2. Minor pretty printing fix for cellset.  Print %st(0) as %st(0) instead
3615          of %f32.
3616    
3617       3. Added a constructor INT32lit to the ast of MLRISC tools.
3618    
3619    ----------------------------------------------------------------------
3620    Name: Matthias Blume
3621    Date: 2002/01/16
3622    Tag: blume-20020116-ffifiddle
3623    Description:
3624    
3625    More fiddling with the FFI interface:
3626    
3627     - Make constness 'c instead of rw wherever possible.  This eliminates
3628       the need for certain explicit coercions.  (However, due to ML's
3629       value polymorphism, there will still be many cases where explicit
3630       coercions are necessary.  Phantom types are not the whole answer
3631       to modeling a subtyping relationship in ML.)
3632    
3633     - ro/rw coersions for pointers added.  (Avoids the detour through */&.)
3634    
3635     - "printf" test example added to src/ml-nlffi-lib/Tests.  (Demonstrates
3636       clumsy workaround for varargs problem.)
3637    
3638    ----------------------------------------------------------------------
3639    Name: Lal George
3640    Date: 2002/01/15
3641    Tag: <none>
3642    Description:
3643    
3644    1. Since COPY instructions are no longer native to the architecture,
3645       a generic functor can be used to implement the expandCopies function.
3646    
3647    2. Allowed EXPORT and IMPORT pseudo-op declarations to appear inside a
3648       TEXT segment.
3649    
3650    ----------------------------------------------------------------------
3651    Name: Matthias Blume
3652    Date: 2002/01/15
3653    Tag: blume-20020115-ffiupdates
3654    Description:
3655    
3656    1. Fix for bug resulting in single-precision float values being returned
3657       incorrectly from FFI calls.
3658    
3659    2. Small modifications to C FFI API:
3660    
3661        - memory-allocation routines return straight objects (no options)
3662          and raise an exception in out-of-memory situations
3663        - unsafe extensions to cast between function pointers and pointers
3664          from/to ints
3665        - added structure C_Debug as an alternative to structure C where
3666          pointer-dereferencing (|*| and |*!) always check for null-pointers
3667        - added open_lib' to DynLinkage;  open_lib' works like open_lib
3668          but also takes a (possibly empty) list of existing library handles
3669          that the current library depends on
3670    
3671    ----------------------------------------------------------------------
3672    Name: Matthias Blume
3673    Date: 2002/01/10
3674    Tag: blume-20020110-newffigen
3675    Description:
3676    
3677    1. Updates to portable graph code.
3678    
3679    2. Major update to ml-nlffigen and ml-nlffi-lib.  Things are much
3680       more scalable now so that even huge interfaces such as the one
3681       for GTK compile in finite time and space. :-)
3682       See src/ml-nlffigen/README for details on what's new.
3683    
3684    ----------------------------------------------------------------------
3685    Name: Lal George
3686    Date: 2001/01/09 14:31:35 EST 2002
3687    Tag: george-20011206-rm-native-copy
3688    Description:
3689    
3690            Removed the native COPY and FCOPY instructions
3691            from all the architectures and replaced it with the
3692            explicit COPY instruction from the previous commit.
3693    
3694            It is now possible to simplify many of the optimizations
3695            modules that manipulate copies. This has not been
3696            done in this change.
3697    
3698    ----------------------------------------------------------------------
3699    Name: Lal George
3700    Date: 2001/12/06 16:50:13 EST 2001
3701    Tag: george-20011206-mlrisc-instruction
3702    Description:
3703    
3704    Changed the representation of instructions from being fully abstract
3705    to being partially concrete. That is to say:
3706    
3707      from
3708            type instruction
3709    
3710      to
3711            type instr                              (* machine instruction *)
3712    
3713            datatype instruction =
3714                LIVE of {regs: C.cellset, spilled: C.cellset}
3715              | KILL of {regs: C.cellset, spilled: C.cellset}
3716              | COPYXXX of {k: CB.cellkind, dst: CB.cell list, src: CB.cell list}
3717              | ANNOTATION of {i: instruction, a: Annotations.annotation}
3718              | INSTR of instr
3719    
3720    This makes the handling of certain special instructions that appear on
3721    all architectures easier and uniform.
3722    
3723    LIVE and KILL say that a list of registers are live or killed at the
3724    program point where they appear. No spill code is generated when an
3725    element of the 'regs' field is spilled, but the register is moved to
3726    the 'spilled' (which is present, more for debugging than anything else).
3727    
3728    LIVE replaces the (now deprecated) DEFFREG instruction on the alpha.
3729    We used to generate:
3730    
3731            DEFFREG f1
3732            f1 := f2 + f3
3733            trapb
3734    
3735    but now generate:
3736    
3737            f1 := f2 + f3
3738            trapb
3739            LIVE {regs=[f1,f2,f3], spilled=[]}
3740    
3741    Furthermore, the DEFFREG (hack) required that all floating point instruction
3742    use all registers mentioned in the instruction. Therefore f1 := f2 + f3,
3743    defines f1 and uses [f1,f2,f3]! This hack is no longer required resulting
3744    in a cleaner alpha implementation. (Hopefully, intel will not get rid of
3745    this architecture).
3746    
3747    COPYXXX is intended to replace the parallel COPY and FCOPY  available on
3748    all the architectures. This will result in further simplification of the
3749    register allocator that must be aware of them for coalescing purposes, and
3750    will also simplify certain aspects of the machine description that provides
3751    callbacks related to parallel copies.
3752    
3753    ANNOTATION should be obvious, and now INSTR represents the honest to God
3754    machine instruction set!
3755    
3756    The <arch>/instructions/<arch>Instr.sml files define certain utility
3757    functions for making porting easier -- essentially converting upper case
3758    to lower case. All machine instructions (of type instr) are in upper case,
3759    and the lower case form generates an MLRISC instruction. For example on
3760    the alpha we have:
3761    
3762      datatype instr =
3763         LDA of {r:cell, b:cell, d:operand}
3764       | ...
3765    
3766      val lda : {r:cell, b:cell, d:operand} -> instruction
3767        ...
3768    
3769    where lda is just (INSTR o LDA), etc.
3770    
3771    ----------------------------------------------------------------------
3772    Name: Matthias Blume
3773    Date: 2001/11/22 21:40:00 EST
3774    Tag: Release_110_37
3775    Description:
3776    
3777    Release 110.37.  This time for real.
3778    
3779    ----------------------------------------------------------------------
3780    Name: Matthias Blume
3781    Date: 2001/11/21 16:35:00 EST
3782    Tag: blume-20011121-foot-in-mouth
3783    Description:
3784    
3785    Removed the "Release_110_37" tag because of a serious bug.
3786    This will be re-tagged once the bug is fixed.
3787    
3788    ----------------------------------------------------------------------
3789    Name: Matthias Blume
3790    Date: 2001/11/21 16:14:00 EST
3791    Tag: blume-20011121-forgottenfile
3792    Description:
3793    
3794    Forgot to add a file.  (Just a .tex-file -- part of
3795    the CM manual source.)
3796    
3797    ----------------------------------------------------------------------
3798    Name: Matthias Blume
3799    Date: 2001/11/21 16:10:00 EST
3800    Tag: blume-20011121-invalid_110_37
3801    Description:
3802    
3803    Note: I removed the original tag "Release_110_37" from this commit
3804          because we found a serious bug in all non-x86 backends.
3805          - Matthias
3806    
3807    1. Modifications to the SML/NJ code generator and to the runtime system
3808       so that code object name strings are directly inserted into code
3809       objects at code generation time.  The only business the runtime system
3810       has with this is now to read the name strings on occasions.
3811       (The encoding of the name string has also changed somewhat.)
3812    
3813    2. CM now implements a simple "set calculus" for specifying export lists.
3814       In particular, it is now possible to refer to the export lists of
3815       other libraries/groups/sources and form unions as well as differences.
3816       See the latest CM manual for details.
3817    
3818    3. An separate notion of "proxy" libraries has again be eliminated from
3819       CM's model.  (Proxy libraries are now simply a special case of using
3820       the export list calculus.)
3821    
3822    4. Some of the existing libraries now take advantage of the new set
3823       calculus.
3824       (Notice that not all libraries have been converted because some
3825       of the existing .cm-files are supposed to be backward compatible
3826       with 110.0.x.)
3827    
3828    5. Some cleanup in stand-alone programs.  (Don't use "exnMessage" -- use
3829       "General.exnMessage"!  The former relies on a certain hook to be
3830       initialized, and that often does not happen in the stand-alone case.)
3831    
3832    ----------------------------------------------------------------------
3833    Name: Lal George
3834    Date: 2001/11/21  13:56:18 EST
3835    Tag: george-2001121-pseudo-ops
3836    Description:
3837    
3838      Implemented a complete redesign of MLRISC pseudo-ops. Now there
3839      ought to never be any question of incompatabilities with
3840      pseudo-op syntax expected by host assemblers.
3841    
3842      For now, only modules supporting GAS syntax are implemented
3843      but more should follow, such as MASM, and vendor assembler
3844      syntax, e.g. IBM as, Sun as, etc.
3845    
3846    ----------------------------------------------------------------------
3847  Name: Matthias Blume  Name: Matthias Blume
3848  Date: 2001/11/14 11:52:00 EST  Date: 2001/11/14 11:52:00 EST
3849  Tag: blume-20011114-srcname  Tag: blume-20011114-srcname