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 972, Sun Oct 28 00:37:02 2001 UTC revision 1669, Wed Oct 27 22:49:25 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: 2004/10/27 17:50:00 CDT
18    Tag: blume-20041027-btrace-msg
19    Description:
20    
21    BackTrace.monitor now also reports the source of the exception that
22    triggered the trace.
23    
24    ----------------------------------------------------------------------
25    Name: Matthias Blume (blume (at) tti - c (dot) org)
26    Date: 2004/10/27 17:20:00 CDT
27    Tag: blume-20041027-x86-c-calls
28    Description:
29    
30    This is the HISTORY entry for two earlier commits, both concerning
31    the x86 c-calls code in MLRISC:
32    
33      - added a missing LOAD in the code that deals with struct arguments
34      - made sure the caller does not add the wrong number of bytes to the
35        stack pointer after a call of a function returning a struct
36        (the callee already pops the implicit argument which points to
37         the space reserved for the result)
38    
39    ----------------------------------------------------------------------
40    Name: Allen Leung (leunga (at) reservoir (dot) com)
41    Date: 2004/10/24 14:00:00 EST
42    Tag: leunga-20041024-x86-gas-fucomip
43    Description:
44    
45       John discovered a bug in the syntax of fucomip.
46       The opcodes FU?COMIP? have been changed to
47    
48           fu?comip? %st(i), %st
49    
50    ----------------------------------------------------------------------
51    Name: Matthias Blume (blume (at) tti - c (dot) org)
52    Date: 2004/10/20 15:06:00 CDT
53    Tag: blume-20041020-standalone-backtrace
54    Description:
55    
56    Added a mechanism for getting back-trace information from standalone
57    programs.  Here is how it works:
58    
59       1. The part of the program from which you want to get backtrace
60          information (usually the whole program) should be wrapped with
61          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
62          main program could be modified from something like
63    
64             fun main (pgm, args) = ...
65    
66         to
67    
68            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
69    
70       2. To be able to access BackTrace.monitor, you have to add
71          library $smlnj-tdp/plugins.cm to the .cm file that contains your
72          main function.
73    
74       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
75          CM might have created in the past for your project).
76    
77       4. Build the system using this command line:
78    
79             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
80                      myprog.cm MyProg.main myprog
81    
82          instead of the usual
83    
84             ml-build myprog.cm MyProg.main myprog
85    
86    I changed a library name:
87    
88       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
89    
90    New libraries:
91    
92       $smlnj-tdp/back-trace.cm
93          -- when loaded causes the back-trace plugin to be installed
94       $smlnj-tdp/coverage.cm
95          -- when loaded causes the coverage plugin to be installed
96    
97    ----------------------------------------------------------------------
98    Name: Matthias Blume (blume (at) tti - c (dot) org)
99    Date: 2004/10/18 16:45:00 CDT
100    Tag: blume-20041018-groupowner
101    Description:
102    
103    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
104    
105    Eliminated group owner specs from .cm files throughout the source tree.
106    
107    ----------------------------------------------------------------------
108    Name: Matthias Blume (blume (at) tti - c (dot) org)
109    Date: 2004/10/15 15:45:00 CDT
110    Tag: blume-20041015-coverage
111    Description:
112    
113    * Test coverage tool added!
114    * Further reorganization of tracing-, debugging-, and profiling support:
115    
116        - moved original BTImp -- now called BackTrace -- into a separate
117          library called $/trace-debug-profile.cm
118        - eliminated all mentions of BTrace from SMLofNJ.Internals
119        - only the instrumentation mechanism is now left in the compiler proper
120        - BackTrace module is a plugin which is NOT plugged in by default
121        - Coverage module is another such plugin
122    
123    To get the benefits of any of these plugin modules, the code in
124    question must be compiled with tdp instrumentation turned on.  This
125    can be done by setting SMLofNJ.Internals.TDP.mode to true.
126    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
127    
128    Plugins are selected at link time.  (Pre-compiled instrumented code
129    can be re-loaded with different plugins in effect.)  When an
130    instrumented module is linked, whatever plugins are at that time
131    enabled will come into effect for that module.
132    
133    To enable the back-trace plugin, load library $/trace-debug-profile.cm
134    and invoke BackTrace.install() (e.g., from the interactive prompt).
135    To enable the coverage plugin, load the same library and invoke
136    Coverage.install().
137    
138    Back-traces are generated automatically on uncaught exceptions and
139    when the code in question explicitly invokes BackTrace.trigger().
140    
141    Coverage (and execution frequency-) information must be queried
142    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
143    
144    ----------------------------------------------------------------------
145    Name: Matthias Blume (blume (at) tti - c (dot) org)
146    Date: 2004/10/14 17:40:00 CDT
147    Tag: blume-20041014-tdp-core
148    Description:
149    
150    Snapshot of a significant overhaul of how the trace/debug/profile support
151    is hooked into the system (specifically: Core and SMLofNJ.Internals).
152    
153    ----------------------------------------------------------------------
154    Name: Matthias Blume (blume (at) tti - c (dot) org)
155    Date: 2004/10/13 16:34:00 CDT
156    Tag: blume-20041013-tdp
157    Description:
158    
159    Some rationalization of names:
160    
161       structure BTrace -> structure TDPInstrument
162       etc.
163    
164    This is is preparation of using the original back-trace
165    instrumentation for other purposes.  "TDP" stands for
166    Trace/Debug/Profile.
167    
168    The control flag controlling whether instrumentation is on or off is now
169    registered under a different name, so instead of running sml as
170    
171     sml -Cinstrument.btrace-mode=true
172    
173    one has to say
174    
175     sml -Ctdp.instrument=true
176    
177    ----------------------------------------------------------------------
178    Name: Matthias Blume (blume (at) tti - c (dot) org)
179    Date: 2004/10/11 16:37:00 CDT
180    Tag: blume-20041011-regions
181    Description:
182    
183    Made some minor modifications to elabcore.sml to have source regions
184    be propagated more tightly -- resulting in better (i.e., smaller)
185    regions being reported in error- and debug messages.
186    
187    ----------------------------------------------------------------------
188    Name: Matthias Blume (blume (at) tti - c (dot) org)
189    Date: 2004/10/08 22:50:00 CDT
190    Tag: blume-20041008-cmkw
191    Description:
192    
193    Fixed handling of keywords in .cm files: After seeing "is" the lexer
194    treats subsequent occurrences of "group", "library", "source", "is",
195    "*", and "-" as ordinary identifiers rather than keywords.
196    
197    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
198    is supposed to accept a tool argument called "source", but this did
199    not work because of the clash with the keyword.
200    
201    ----------------------------------------------------------------------
202    Name: Matthias Blume (blume (at) tti - c (dot) org)
203    Date: 2004/10/07 16:00:00 CDT
204    Tag: blume-20041007-cleanup
205    Description:
206    
207    Assorted cleanup work:
208    
209      - got rid of intstrmap in favor of using the library's
210        hash table implementation
211      - threw out most of the pathnames stuff, as it was not used anyway
212      - simplified tokentable implementation
213      - fixed some minor spelling errors
214    
215    ----------------------------------------------------------------------
216    Name: Matthias Blume (blume (at) tti - c (dot) org)
217    Date: 2004/10/06 15:15:15 CDT
218    Tag: blume-20041006-handler
219    Description:
220    
221    Cleaned up the absyn to reflect the invariant that HANDLE always
222    carries a FNexp as part of the type definition.  This eliminates some
223    superfluous sanity checks at runtime down the road.
224    
225    Some minor cleanup of the btrace code.
226    
227    ----------------------------------------------------------------------
228    Name: Matthias Blume (blume (at) tti - c (dot) org)
229    Date: 2004/10/01 10:20:30 CDT
230    Tag: blume-20041001-slave
231    Description:
232    
233    Added hack to make slave mode work in the presence of the version
234    tool.  (Still, since the master does two passes over the code for
235    CMB.make, the release number gets bumped twice when slaves are
236    attached.  I don't know if this is worth fixing...)
237    
238    ----------------------------------------------------------------------
239    Name: Matthias Blume (blume (at) tti - c (dot) org)
240    Date: 2004/09/30 10:55:00 CDT
241    Tag: blume-20040930-version
242    Description:
243    
244    * Moved the "version" magic into its own little library under
245    src/system/smlnj/internal.  This avoids expensive reconstruction
246    of a stable src/compiler/core.cm.
247    
248    * At the same time, structure CompilerVersion is now known as structure
249    SMLNJVersion.
250    
251    * Arranged for the version tool to NOT kick in when rebuilding the system
252      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
253      Also, loading the versiontool does not work when rebuilding the system
254      because CM is not properly initialized at that time.
255    
256    ----------------------------------------------------------------------
257    Name: Matthias Blume (blume (at) tti - c (dot) org)
258    Date: 2004/09/29 14:00:00 CDT
259    Tag: blume-20040929-autoversion
260    Description:
261    
262    Implemented some CM magic to have
263        file src/compiler/TopLevel/main/version.sml
264    generated automagically.
265    The version is taken from two files: config/version and config/release.
266    The first is expected to contain a two-part version number such as 110.49.
267    The second should contain a single number, but it may be missing.
268    
269    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
270    time the version tool is loaded (which is the first time you say CMB.make),
271    then the tool will increment the value stored in config/release every
272    time CMB.make is invoked.
273    
274    The binfile format is now insensitive to anything beyond the first
275    two components of a version number, so bumping the release does not render
276    binfiles incompatible.  Auto-bumping can be used to keep track of versions
277    during development without invalidating existing binfiles.
278    
279    In any case, every CMB.make updates the date information in version.sml.
280    (This is the date that is printed in the banner.)
281    
282    ----------------------------------------------------------------------
283    Name: Matthias Blume (blume (at) tti - c (dot) org)
284    Date: 2004/09/28 10:53:00 CDT
285    Tag: blume-20040928-controls
286    Description:
287    
288    Some cleanup of the controls code.
289    
290    ----------------------------------------------------------------------
291    Name: Matthias Blume (blume (at) tti - c (dot) org)
292    Date: 2004/09/27 22:08:00 CDT
293    Tag: blume-20040927-controls
294    Description:
295    
296    Added two pieces of functionality to the Controls interface:
297    
298    1.   val save'restore: 'a control -> unit -> unit
299    
300       grabs the current value of the control in stage 1 and restores it
301       in stage 2.
302    
303    2.   val set' : 'a control * 'a -> unit -> unit
304    
305       stores the given value into the control in stage 2 (i.e., delayed)
306       but does all error checking in stage 1.
307       (This is for string controls that need to do parse their argument --
308        something that might fail.  In some cases, notably in CM, one
309        already knows the intended argument but wants to delay the actual
310        assignment until a time when error recovery would be more difficult.)
311    
312    Changed the handling of controls in tool arguments to classes "sml" and
313    "lazysml":
314       - use Controls.save'restore as a more robust way of restoring the
315         old value (in particular: without having to re-parse the string)
316       - use controls to handle the "overload" keyword in the init group
317         (I believe this change actually fixes a long-standing obscure bug.)
318    
319    ----------------------------------------------------------------------
320    Name: Matthias Blume (blume (at) tti - c (dot) org)
321    Date: 2004/09/27 17:00:00 CDT
322    Tag: blume-20040927-lazysml
323    Description:
324    
325    Added a new tool class called "lazysml" to CM's tool chest.  The only
326    difference to "sml" is that compilation is done with Control.lazysml
327    set to true.  A source of class "lazysml" is automatically recognized
328    by a file name suffix of ".lml".
329    
330    In addition to the above feature, the original class "sml" now also
331    supports a tool argument "lazy" which has the same effect.  As a
332    result, the following three lines are equivalent:
333    
334        foo.sml : lazysml
335        foo.sml : sml (lazy)
336        foo.sml (lazy)
337    
338    The setting goes into effect both during parsing and during
339    compilation.  The original setting is restored right after parsing and
340    after compilation, respectively.
341    
342    In addition to all the above, there is also a general mechanism to set
343    ANY of the "controls" that are available at the command line via
344    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
345    apply as well. (In fact, "lazy" is implemented as a special case of
346    the general mechanism.)
347    
348    The .cm file syntax uses a new keyword tool argument called "with".
349    There are several ways of indicating the desired settings:
350    
351        foo.sml (with:parser.quotations=true)
352        foo.sml (with:(name:parser.quotations value:true))
353        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
354        foo.sml (with:(name1=value1 name2=value2 ...))
355        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
356    
357        etc.
358    
359    Another possible abbreviation is to leave out the =v or value:v part
360    if the name refers to a boolean control (in which case the value is
361    taken to be true).  Thus, one could get lazy sml also by saying:
362    
363        foo.sml (with:parser.lazy-keyword=true)
364        foo.sml (with:parser.lazy-keyword)
365        foo.sml (with:(name:parser.lazy-keyword value:true))
366        foo.sml (with:(name:parser.lazy-keyword))
367    
368    ----------------------------------------------------------------------
369    Name: Matthias Blume (blume (at) tti - c (dot) org)
370    Date: 2004/09/24 16:22:00 CDT
371    Tag: blume-20040924-ppc-long-branch
372    Description:
373    
374    Turned message about "emiting long form of branch" off by default.
375    Added a control flag to turn it back on when desired.
376    
377    ----------------------------------------------------------------------
378    Name: Matthias Blume (blume (at) tti - c (dot) org)
379    Date: 2004/09/24 16:05:00 CDT
380    Tag: blume-20040924-rounding
381    Description:
382    
383    Applied patch for setting rounding modes under Mac OS X.  Thanks to
384    Melissa O'Neill for providing the code!
385    
386    ----------------------------------------------------------------------
387    Name: Matthias Blume (blume (at) tti - c (dot) org)
388    Date: 2004/09/23 17:30:00 CDT
389    Tag: blume-20040923-envvars
390    Description:
391    
392    1. Changed definition of type ControlRegistry.registry_tree to
393       include control_info (i.e., the name of the controlling environment
394       variable).
395    
396    2. Added command-line flags -e and -E to print the names of environment
397       variables that can be used to control internal settings.
398       (This uses the new API mentioned in 1.)
399    
400    ----------------------------------------------------------------------
401    Name: Matthias Blume (blume (at) tti - c (dot) org)
402    Date: 2004/09/13 16:50:00 CDT
403    Tag: Release_110_49
404    Description:
405    
406    New working version (110.49).  NEW BOOTFILES!
407    
408    ----------------------------------------------------------------------
409    Name: Matthias Blume
410    Date: 2004/09/13 16:20:00 CDT
411    Tag: blume-20040913-config-mlrisc
412    Description:
413    
414    Put target "mlrisc" back into the default list.
415    (There is no harm in having it, and some users have expressed their
416    wish to have "mlrisc" included by default.)
417    
418    ----------------------------------------------------------------------
419    Name: John Reppy
420    Date: 2004/09/13
421    Tag: jhr-20040913-signals
422    Description:
423    
424    Fixed the signal masking code to properly nest mask/unmask operations
425    on a per-signal basis.
426    
427    ----------------------------------------------------------------------
428    Name: Matthias Blume (blume (at) tti - c (dot) org)
429    Date: 2004/09/08 13:20:00 CDT
430    Tag: blume-20040908-heap-magic
431    Description:
432    
433    Bumped the heap macig number to 0x09082004 to account for the changed
434    layout of the ML frame under MacOS X.
435    
436    ----------------------------------------------------------------------
437    Name: Allen Leung (leunga (at) reservoir (dot) com)
438    Date: 2004/09/03 11:26:00 EST
439    Tag: leunga-20040903-cygwin-install
440    Description:
441    
442       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
443    The Cygwin runtime is turned on by setting the environment
444    variable SMLNJ_CYGWIN_RUNTIME to 1.
445    
446    ----------------------------------------------------------------------
447    Name: Matthias Blume (blume (at) tti - c (dot) org)
448    Date: 2004/08/31 17:15:00 CDT
449    Tag: blume-20040831-core
450    Description:
451    
452    Added some exports to src/compiler/core.cm upon request by J. Joemann.
453    
454    ----------------------------------------------------------------------
455    Name: Matthias Blume (blume (at) tti - c (dot) org)
456    Date: 2004/08/30 17:55:00 CDT
457    Tag: blume-20040830-installer
458    Description:
459    
460    Upon request by Johannes Joemann:
461    
462    - improved ML code of installer to fall back to coping when renaming
463      fails (i.e., when source and target are on different file systems);
464      the code compiles but has yet to be tested in anger
465    - removed mlrisc from list of default targets (config/targets)
466    
467    ----------------------------------------------------------------------
468    Name: Matthias Blume (blume (at) tti - c (dot) org)
469    Date: 2004/08/27 17:20:00 CDT
470    Tag: blume-20040827-ptreql
471    Description:
472    
473    Added ptreql primop to structure InlineT (upon request from Larry
474    Paulson).
475    
476    ----------------------------------------------------------------------
477    Name: Allen Leung (leunga (at) reservoir (dot) com)
478    Date: 2004/08/15 21:21:00 EST
479    Tag: leunga-110_48-udgraph
480    Description:
481    
482    Another bug fix from Carl Hauser:
483    
484    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
485    > 48c48
486    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
487    > e::L)
488    > ---
489    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
490    > Without this, any deletion of an edge in an undirected graph does severe
491    > violence to the graph.
492    
493    ----------------------------------------------------------------------
494    Name: Allen Leung (leunga (at) reservoir (dot) com)
495    Date: 2004/08/10 23:35:00 EST
496    Tag: leunga-110_48-ppc
497    Description:
498    
499       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
500    
501    ----------------------------------------------------------------------
502    Name: Matthias Blume (blume (at) tti - c (dot) org)
503    Date: 2004/08/10 12:00:00 CDT
504    Tag: Release_110_48
505    Description:
506    
507    New working version (110.48).  NEW BOOTFILES!
508    
509    ----------------------------------------------------------------------
510    Name: Allen Leung (leunga (at) reservoir (dot) com)
511    Date: 2004/08/09 12:21:00 EST
512    Tag: leunga-110_47-dijsktra
513    Description:
514    
515    Bug fix from Carl Hauser:
516    
517    single_source_shortest_paths in dijkstra.sml was observed to get wrong
518    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
519    
520    The problem is that following the expression A.update(dist,s,Num.zero)
521    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
522    
523    ----------------------------------------------------------------------
524    Name: Matthias Blume (blume (at) tti - c (dot) org)
525    Date: 2004/08/06 18:10:00 CDT
526    Tag: blume-20040806-cmdline
527    Description:
528    
529    Fiddled with handling of command-line options:
530    
531      * sml now quits after processing the command line
532        if -H, -S, -h<n>, or -s<n> appears as the last
533        command-line argument
534      * a new option -q terminates the session when encountered on
535        the command line; subsequent arguments will be ignored
536      * bug fixes: short (erroneous) arguments are no longer ignored
537        completely
538    
539    ----------------------------------------------------------------------
540    Name: Allen Leung (leunga (at) reservoir (dot) com)
541    Date: 2004/08/04 18:17:00 EST
542    Tag: leunga-110_47-ppc-ibm-asm
543    Description:
544    
545      - Added minimal IBM assembly syntax support for PowerPC.
546    
547      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
548        been altered in the runtime.  We need an automatic way to keep the file
549        in sync.
550    
551    ----------------------------------------------------------------------
552    Name: Matthias Blume (blume (at) tti - c (dot) org)
553    Date: 2004/08/04 14:00:00 CDT
554    Tag: Release_110_47
555    Description:
556    
557    New working version (110.47).  NEW BOOTFILES!
558    
559    ----------------------------------------------------------------------
560    Name: Matthias Blume (blume (at) tti - c (dot) org)
561    Date: 2004/08/03 14:25:00 CDT
562    Tag: blume-20040803-callingconv
563    Description:
564    
565    Added low-level support for choosing C calling conventions by
566    twiddling the type of rawccall.  (See
567    src/compiler/Semant/types/cproto.sml for details.)
568    
569    ----------------------------------------------------------------------
570    Name: Matthias Blume (blume (at) tti - c (dot) org)
571    Date: 2004/08/02 15:55:00 CDT
572    Tag: blume-20040802-backout
573    Description:
574    
575    Backed out of change to win32-filesys.c.  The earlier patch to
576    get_file_time caused CM to produce files with the wrong time stamp.
577    
578    ----------------------------------------------------------------------
579    Name: Matthias Blume (blume (at) tti - c (dot) org)
580    Date: 2004/08/02 14:45:00 CDT
581    Tag: blume-20040802-nlffi-win32
582    Description:
583    
584    Added NLFFI support for Win32, adapted from a patch provided by David
585    Hansel.  This is currently completely untested.  Also, the issue
586    concerning stdcall vs. ccall is still unresolved.
587    
588    ----------------------------------------------------------------------
589    Name: Matthias Blume (blume (at) tti - c (dot) org)
590    Date: 2004/07/30 17:55:00 CDT
591    Tag: blume-20040730-various
592    Description:
593    
594    Gearing up towards 110.47...
595    
596    - various minor bugfixes to ml-nlffigen
597    - a beginning of a manual for nlffi
598    
599    - eliminated 'export name=value' in config/install.sh as this does
600      not work with certain versions of /bin/sh
601      (Thanks to David King at Motorola for catching this.)
602    
603    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
604      - added a test for tm==NULL to gmtime.c and localtime.c
605      - applied patch for incorrect GetFileTime under win32
606      - toSeconds -> toMilliseconds in Win32/win32-process.sml
607    
608    ----------------------------------------------------------------------
609    Name: Matthias Blume (blume (at) tti - c (dot) org)
610    Date: 2004/07/21 18:20:00 CDT
611    Tag: blume-20040721-nlffigen
612    Description:
613    
614    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
615      for a typedef to an incomplete type, but leave out the "typ" member.
616      (This is just for consistency.)
617    - Started to produce what is supposed to become better (i.e., comprehensive)
618      documentation of what ml-nlffigen does and produces.
619    
620    ----------------------------------------------------------------------
621    Name: Matthias Blume (blume (at) tti - c (dot) org)
622    Date: 2004/07/14 16:25:00 CDT
623    Tag: blume-20040714-union
624    Description:
625    
626    Added C_UNION to c-calls/c-types.sml and updated the machinery
627    (ml-nlffigen, cproto.sml) that conveys C function interface
628    information to the code generator.
629    
630    However, the actual architecture-specific implementation of function
631    arguments and results that are C unions is still not implemented.
632    
633    ----------------------------------------------------------------------
634    Name: Allen Leung (leunga (at) reservoir (dot) com)
635    Date: 2004/07/14 14:38:00 EST
636    Tag: leunga-110_46_1-ppc-lwzu
637    Description:
638    
639        Added these instructions to the PowerPC architecture:
640        LBZU(X), LHZU(X), LWZU(X),
641        STWU(X), STFDU, STFSU
642    
643        etc...
644    
645        Note: I haven't added their instruction encoding into the description.
646    
647    ----------------------------------------------------------------------
648    Name: Allen Leung (leunga (at) reservoir (dot) com)
649    Date: 2004/07/13 15:04:00 EST
650    Tag: leunga-110_46_1-ppc-lwarx
651    Description:
652    
653       Added the two instructions LWARX and STWCX to the PowerPC
654    instruction set.
655    
656       A (untested) rewrite of loop-structure.sml.  The old version
657    is completely broken.
658    
659    ----------------------------------------------------------------------
660    Name: Matthias Blume (blume (at) tti - c (dot) org)
661    Date: 2004/07/13 13:50:00 CDT
662    Tag: blume-20040713-nlffi
663    Description:
664    
665    - use paramAlloc to report c-calls with too many arguments
666      (for PPC version where parameter area is pre-allocated)
667    - added ccall_maxargspace to machspec (to implement the above)
668    - made "make" commend in CM's "make" tool configurable
669    - added option (default: on) for passing the name of the SML/NJ's "bin"
670      directory to "make"; the call looks like this:
671    
672         make <options> SMLNJ_BINDIR=<dir> <target>
673    
674      This can be used by the Makefile to, e.g., pick the "right" version
675      of ml-nlffigen.
676    - minor code tweaks
677    
678    ----------------------------------------------------------------------
679    Name: Matthias Blume (blume (at) tti - c (dot) org)
680    Date: 2004/07/12 22:50:00 CDT
681    Tag: blume-110_46_1-macosx-nlffi
682    Description:
683    
684    NLFFI under Mac OS X now working (sort of).  This is largely untested,
685    though.
686    
687    Note:  1. You have to make a new, clean build of the runtime system.
688           2. There are new BOOTFILES, you have to use them!
689              (Doing the bootstrap process yourself would be *very* painful!
690               If you absolutely have to do it, build the system under
691               a different architecture and then cross-compile.)
692    
693    Version bumped to 110.46.1 to account for runtime data format changes.
694    
695    ----------------------------------------------------------------------
696    Name: Matthias Blume (blume (at) tti - c (dot) org)
697    Date: 2004/06/18 14:30:00 CDT
698    Tag: blume-20040618-unix
699    Description:
700    
701    Changed the implementation of structure Unix so that the same stream
702    is returned every time one of the {text,bin}{In,Out}streamOf functions
703    is invoked on the same proc.  This is not what the spec currently
704    says -- although IMO it arguably should.  (See discussion below.)
705    
706    ----------------------------------------------------------------------
707    Name: Matthias Blume (blume (at) tti - c (dot) org)
708    Date: 2004/06/17 18:15:00 CDT
709    Tag: Release_110_46
710    Description:
711    
712    New working version (110.46).  NEW BOOTFILES!
713    
714    ----------------------------------------------------------------------
715    Name: Matthias Blume (blume (at) tti - c (dot) org)
716    Date: 2004/06/17 17:20:00 CDT
717    Tag: blume-20040617-timer-unix
718    Description:
719    
720    Changed the interface of structures Timer and Unix to match the most
721    recent Basis spec.
722    
723    In the case of Unix there still seems to be an open/weird issue:
724    
725       The {text,bin}{In,Out}streamOf functions are supposed to create
726       fresh streams whenever they are called -- as opposed to have them
727       return the same stream every time.  This design is supposed to
728       prevent space leaks caused by proc values hanging on to streams.
729    
730       The reap function, on the other hand, is supposed to close the
731       streams.  This cannot be done without having a handle on the
732       stream in proc after all...
733    
734    I took the liberty to implement the following stopgap solution:
735    
736      The proc value hangs on to the most recently created stream(s).
737      Reap closes those.  If either or both of the two streams hadn't
738      been created at all yet, then reap will close the corresponding
739      file descriptors directly.
740    
741    PS: I don't understand the original space leak argument anymore.  If
742    a proc hangs on to the imperative stream, then I/O operations on those
743    will advance the state of the cached stream and avoid the space leak.
744    
745    ----------------------------------------------------------------------
746    Name: Matthias Blume (blume (at) tti - c (dot) org)
747    Date: 2004/05/28 16:45:00 CDT
748    Tag: blume-20040528-basis
749    Description:
750    
751    Added signature PACK_REAL and exported functor PrimIO.
752    
753    ----------------------------------------------------------------------
754    Name: Matthias Blume (blume (at) tti - c (dot) org)
755    Date: 2004/05/25 16:00:00 CDT
756    Tag: blume-20040525-group-owner
757    Description:
758    
759    CM now ignores (but still accepts) the "owner" information in group
760    descriptions.  The owner of a group is its next enclosing
761    library. Each group must have a unique owner.  (There is a virtual
762    "toplevel" library that owns groups which are not nested within a real
763    library.)  Previously, each group had to explicitly declare its owner,
764    and CM would check that such a declaration is correct.  The new scheme
765    is to have CM check that for each group there is precisely one owning
766    library.
767    
768    The advantage of the new scheme is that the programmer no longer needs
769    to maintain the somewhat annoying owner information.  The downside is
770    that CM cannot enforce the ownership rule across multiple runs of
771    CM.make.  Fortunately, enclosing the same group in two different
772    libraries A and B which are not part of the same program does not
773    cause real problems.
774    
775    ----------------------------------------------------------------------
776    Name: Matthias Blume (blume (at) tti - c (dot) org)
777    Date: 2004/05/20 16:00:00 CDT
778    Tag: blume-20040520-win32
779    Description:
780    
781    Made the win32 version work again.  (Strangely, a misplaced comma had
782    slipped into win32-process.c which prevented the runtime from being
783    compiled correctly.)
784    
785    Also, included a minor addition to ml-build.bat analogous to what was
786    done in blume-20040519-ml-build.
787    
788    ----------------------------------------------------------------------
789    Name: Matthias Blume (blume (at) tti - c (dot) org)
790    Date: 2004/05/19 22:10:00 CDT
791    Tag: blume-20040519-ml-build
792    Description:
793    
794    Arranged for ml-build to clean up after itself a little bit better.
795    The script generates a temporary SML source file and compiles it using
796    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
797    gets rid of those at the end, so they don't accumulate under .cm.
798    
799    This required a minor change to install.sh because the name of the
800    metadata directory (default: .cm) is actually configurable at
801    installation time.
802    
803    ----------------------------------------------------------------------
804    Name: Matthias Blume (blume (at) tti - c (dot) org)
805    Date: 2004/05/18 15:50:00 CDT
806    Tag: blume-20040518-mkreader
807    Description:
808    
809    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
810    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
811    
812    ----------------------------------------------------------------------
813    Name: Matthias Blume (blume (at) tti - c (dot) org)
814    Date: 2004/05/11 14:35:00 CDT
815    Tag: blume-20040511-win32sock
816    Description:
817    
818    Added previously missing support for many socket-related functions
819    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
820    for the voluminous patch!
821    
822    (I have not tested this patch under win32 yet.)
823    
824    Here is David's e-mail:
825    
826    Hi,
827    
828    Attached to this email you find a diff against sml/nj 110.45
829    that will enable socket support under Windows.
830    
831    To apply the patch (using unix or cygwin)
832    1) gunzip runtime.diff.gz
833    2) "cd" into "src/runtime" in the source tree of a fresh
834       110.45 installation.
835    3) patch -p 1 < [your/path/to]runtime.diff
836    
837    The code compiles fine but has NOT yet been extensively tested.
838    I only ran a few tests for basic socket client functionality
839    (which worked fine).  Especially the functions that use ioctl
840    are not tested at all and might not work (see below).
841    
842    I implemented this since we want to move to a newer version of sml/nj
843    but need socket support in order to use it.  This is the first time I
844    even had a look at the sml/nj source,  so please review my changes
845    before making this part of the distribution!  Here are a few issues
846    that I think might be better for someone to solve who is more
847    familiar with the sml/nj source (and socket programming):
848    
849    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
850      exception since I could not figure out what the windows equivalent
851      of these functions is
852    
853    - In sockets-osdep.h there are a some #include statements that are
854      only used in a few files that include sockets-osdep.h
855    
856    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
857      does not process its return value since I don't know how to
858      report an error upwards.
859    
860    - It would probably be good to have a call to WSACleanup() when
861      the library is unloaded (if there is such a possibility).
862      Otherwise I think Windows will take care of this automatically
863      when the process finishes.
864    
865    - I used ioctlsocket() as a replacement for ioctl() but I have
866      no idea if that is actually the proper replacement on Windows.
867    
868    - All these issues are marked in the code by "FIXME" comments.
869    
870    We use sml/nj extensively in our products and are quite happy
871    with it.  I hope this contribution will help you.
872    
873    Keep up the good work!
874    
875    David
876    
877    ----------------------------------------------------------------------
878    Name: Matthias Blume (blume (at) tti - c (dot) org)
879    Date: 2004/05/11 14:20:00 CDT
880    Tag: blume-20040511-installml
881    Description:
882    
883    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
884    the patch.)
885    
886    ----------------------------------------------------------------------
887    Name: Matthias Blume (blume (at) tti - c (dot) org)
888    Date: 2004/05/11 14:05:00 CDT
889    Tag: blume-20040511-nlffi-netbsd
890    Description:
891    
892    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
893    the patch.)
894    
895    ----------------------------------------------------------------------
896    Name: Matthias Blume (blume (at) tti - c (dot) org)
897    Date: 2004/05/11 12:05:00 CDT
898    Tag: blume-20040511-exports
899    Description:
900    
901    As per request by Adam Chlipala <adam@hcoop.net>, extended various
902    export lists in compiler-related .cm-files.
903    
904    ----------------------------------------------------------------------
905    Name: Matthias Blume (blume (at) tti - c (dot) org)
906    Date: 2004/05/11 11:35:00 CDT
907    Tag: blume-20040511-allsource
908    Description:
909    
910    The installer now honors the "src-smlnj" target again, although its meaning
911    has changed from "all sources required for the compiler" to "all sources
912    the installer knows about".  In other words, if you enable "src-smlnj"
913    in the "targets" file, then the installer will pull in sources for
914    everything.  (Notice that this refers to source code only.  Compiled
915    code is still only installed for modules that were requested explicitly
916    or which are required for other modules that were requested explicitly.)
917    
918    ----------------------------------------------------------------------
919    Name: Matthias Blume (blume (at) tti - c (dot) org)
920    Date: 2004/04/23 17:40:00 CDT
921    Tag: blume-20040423-ieee-scan
922    Description:
923    
924    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
925    in the exponent calculation we get INF or ZERO (depending on the mantissa
926    and the sign of the exponent).
927    
928    ----------------------------------------------------------------------
929    Name: Matthias Blume (blume (at) tti - c (dot) org)
930    Date: 2004/04/23 10:40:00 CDT
931    Tag: blume-20040423-ml-build
932    Description:
933    
934    The ml-build script now terminates with a non-0 status when something
935    goes wrong.
936    
937    ----------------------------------------------------------------------
938    Name: Matthias Blume (blume (at) tti - c (dot) org)
939    Date: 2004/04/22 16:35:00 CDT
940    Tag: blume-20040422-Option
941    Description:
942    
943    Made exception Option to be the same as exception Option.Option
944    (as it should be).
945    
946    ----------------------------------------------------------------------
947    Name: Allen Leung (leunga (at) reservoir (dot) com)
948    Date: 2004/03/19 14:40:00 EST
949    Tag: leunga-20040319-cygwin-nlffi
950    Description:
951    
952    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
953    of SML/NJ.  The problem is that
954    
955       lib = dlopen(NULL, ...)
956       f   = dlsym(lib, "malloc");
957    
958    does not work on Windows unless we explicitly export symbols
959    such as 'malloc' during linking.  We fixed this by explicitly
960    exporting the required symbols with the magic gcc incantation:
961    
962        -Wl,--export-all cygwin.def
963    
964    where cygwin.def is a file containing all the symbols that
965    we wish to export.
966    
967    I suspect this is a Windows problem and we'll have to
968    do the same (somehow with windows compilers) when
969    we build the native win32 version with the system
970    calls LoadLibrary/GetProcAddress.
971    
972    ----------------------------------------------------------------------
973    Name: Matthias Blume (blume (at) tti - c (dot) org)
974    Date: 2004/03/04 16:35:00 CST
975    Tag: blume-20040304-intinf-fmt
976    Description:
977    
978    Fixed problem with IntInf.fmt (sign would show up on the right instead
979    of on the left for BIN, OCT, and HEX).
980    
981    ----------------------------------------------------------------------
982    Name: Matthias Blume (blume (at) tti - c (dot) org)
983    Date: 2004/03/04 11:25:00 CST
984    Tag: blume-20040304-symlinks
985    Description:
986    
987    Fixed problem with installer script (unix only) where bin/ml-yacc and
988    friends pointed (via symlinks) to absolute locations instead of just
989    .run-sml.  This was reported by Vesa A Norrman.
990    
991    ----------------------------------------------------------------------
992    Name: Matthias Blume (blume (at) tti - c (dot) org)
993    Date: 2004/02/13 14:50:00 CST
994    Tag: Release_110_45
995    Description:
996    
997    New working version (110.45). New bootfiles.
998    
999    ----------------------------------------------------------------------
1000    Name: Matthias Blume (blume (at) tti - c (dot) org)
1001    Date: 2004/01/26 15:15:15 CST
1002    Tag: blume-20040126-toplevel
1003    Description:
1004    
1005    Improved handling of exceptions at the interactive toplevel.
1006    
1007    ----------------------------------------------------------------------
1008    Name: Matthias Blume (blume (at) tti - c (dot) org)
1009    Date: 2004/01/26 11:25:00
1010    Tag: blume-20040126-app
1011    Description:
1012    
1013    Type of top-level "app" corrected.
1014    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1015    
1016    ----------------------------------------------------------------------
1017    Name: Matthias Blume (blume (at) tti - c (dot) org)
1018    Date: 2003/11/18 17:10 CST
1019    Tag: blume-20031118-basis-fiddle
1020    Description:
1021    
1022    - changed Timer interface to what might become the spec
1023    - POSIX_FLAGS -> BIT_FLAGS according to spec
1024    - some other minor discrepancies wrt. spec eliminated
1025    
1026    ----------------------------------------------------------------------
1027    Name: Matthias Blume (blume (at) tti - c (dot) org)
1028    Date: 2003/11/06 12:00:00 CST
1029    Tag: Release_110_44
1030    Description:
1031    
1032    New working version (110.44). New bootfiles.
1033    
1034    ----------------------------------------------------------------------
1035    Name: Matthias Blume (blume (at) tti - c (dot) org)
1036    Date: 2003/11/04 11:50:00 CST
1037    Tag: blume-20031104-move-libraries
1038    Description:
1039    
1040    Eliminated the "dont_move_libraries" directive in config/targets.
1041    (The mechanism was broken and could not be fixed easily.  Moreover,
1042    there does not seem to be any reason not to move all libraries into
1043    lib during installation.  I originally implemented this directive as a
1044    backward-compatibility feature when I first introduced the new CM.
1045    Now that things have been stable for a long time and going back to the
1046    old CM is not an option, there is no reason to keep it around.)
1047    
1048    ----------------------------------------------------------------------
1049    Name: Matthias Blume (blume (at) tti - c (dot) org)
1050    Date: 2003/11/03 16:00:00 CST
1051    Tag: blume-20031103-installdir
1052    Description:
1053    
1054    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1055    Richards for pointing out the problem and providing the solution.)
1056    
1057    ----------------------------------------------------------------------
1058    Name: Matthias Blume (blume (at) tti - c (dot) org)
1059    Date: 2003/10/01 17:05:00 CDT
1060    Tag: blume-20031001-lal-mlrisc
1061    Description:
1062    
1063    MLRISC bug fix from Lal.
1064    
1065    ----------------------------------------------------------------------
1066    Name: Matthias Blume (blume (at) tti - c (dot) org)
1067    Date: 2003/09/30 16:10:00 CDT
1068    Tag: blume-20030930-primio-bat
1069    Description:
1070    
1071    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1072    2. Improved .bat files (for Win32 port) to make things work under Win95.
1073       (thanks to Aaron S. Hawley for this one)
1074    
1075    ----------------------------------------------------------------------
1076    Name: Matthias Blume (blume (at) tti - c (dot) org)
1077    Date: 2003/09/26 16:05:00 CDT
1078    Tag: blume-20030926-wrappriv
1079    Description:
1080    
1081    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1082    
1083    ----------------------------------------------------------------------
1084    Name: Matthias Blume (blume (at) tti - c (dot) org)
1085    Date: 2003/09/26 15:00:00 CDT
1086    Tag: blume-20030926-110_43_3
1087    Description:
1088    
1089    - additional cleanup
1090    - version number bump, NEW BOOTFILES
1091    
1092    ----------------------------------------------------------------------
1093    Name: Matthias Blume (blume (at) tti - c (dot) org)
1094    Date: 2003/09/26 12:00:00 CDT
1095    Tag: blume-20030926-ppautoload
1096    Description:
1097    
1098    I modified the read-eval-print loop so that the autoloader gets
1099    invoked whenever the prettyprinter tries to look up a symbol that
1100    is not currently defined in the toplevel environment but which
1101    appears in CM's autoload registry.  As a result, we see far fewer of
1102    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1103    
1104    In addition to this I tried to clean up some pieces of the Basis
1105    implementation (e.g., Socket, Word8Array) in order to prevent other
1106    instances of these ?.Foo.Bar.xxx names from being printed.
1107    
1108    The mechanism that picks names for types still needs some work, though.
1109    (Right now it seems that if there is a type A.t which is defined to
1110    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1111    "?.B.u" although the perhaps more sensible solution would be to use
1112    "A.t" in this case.  In other words, the prettyprinter should follow
1113    a chain of DEFtycs not farther than there are corresponding toplevel
1114    names in the current environment.)
1115    
1116    ----------------------------------------------------------------------
1117    Name: Matthias Blume (blume (at) tti - c (dot) org)
1118    Date: 2003/09/24 16:31:00 CDT
1119    Tag: blume-20030924-installer
1120    Description:
1121    
1122    Another installer tweak:  All the ML code for the installer is now
1123    compiled during CMB.make and put into a little library called
1124    $smlnj/installer.cm.  The installation then simply invokes
1125    
1126       sml -m $smlnj/installer.cm
1127    
1128    and everything happens automagically.
1129    
1130    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1131    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1132            and BINDIR.
1133    
1134    The new scheme guarantees that the ML code responsible for the installation
1135    is in sync with the APIs of the main system.  Also, the installer is
1136    somewhat faster because the installer script is precompiled.
1137    
1138    ----------------------------------------------------------------------
1139    Name: Matthias Blume (blume (at) tti - c (dot) org)
1140    Date: 2003/09/24 15:35:00 CDT
1141    Tag: blume-20030924-synsock
1142    Description:
1143    
1144    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1145    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1146    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1147    implementations of non-blocking operations from CML's Socket
1148    structure.
1149    
1150    ----------------------------------------------------------------------
1151    Name: Matthias Blume (blume (at) tti - c (dot) org)
1152    Date: 2003/09/24 15:10:05 CDT
1153    Tag: blume-20030924-sockets
1154    Description:
1155    
1156    1. Fixed SOCKET API and implementation to match Basis spec.
1157       This required changing the internal representation of sockets to one
1158       that remembers (for each socket file descriptor) whether it is currently
1159       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1160       call is made only if the state actually needs to change).
1161    
1162    2. OS-specific details of sockets were moved into separate files, thus
1163       making it possible to unify the bulk of the socket implementations
1164       between Unix and Win32.
1165    
1166    3. CML's socket API changed accordingly.
1167       (Note that we need to remove non-blocking functions from this API
1168       since they are redundant in the case of CML!)
1169    
1170    4. CML's socket implementation now makes use of non-blocking functions
1171       provided by Basis, thus removing all OS-dependent code from this part
1172       of CML.
1173    
1174    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1175    
1176    ----------------------------------------------------------------------
1177    Name: Matthias Blume (blume (at) tti - c (dot) org)
1178    Date: 2003/09/22 12:10:00 CDT
1179    Tag: blume-20030922-110_43_2
1180    Description:
1181    
1182    Made a new interim version and bootfiles for developer's bootstrapping
1183    convenience.
1184    
1185       110.43.2 -- NEW BOOTFILES
1186    
1187    ----------------------------------------------------------------------
1188    Name: Matthias Blume (blume (at) tti - c (dot) org)
1189    Date: 2003/09/19 15:55:00 CDT
1190    Tag: blume-20030919-cmdir
1191    Description:
1192    
1193    1. new-install.sh -> install.sh
1194    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1195    3. tweaked installer so that another name instead of .cm can be chosen
1196       at install time (by setting the CM_DIR_ARC environment variable
1197       during installation); once installation is complete, the name is
1198       fixed
1199    
1200    ----------------------------------------------------------------------
1201    Name: Matthias Blume (blume (at) tti - c (dot) org)
1202    Date: 2003/09/18 16:00:00 CDT
1203    Tag: blume-20030918-110_43_1
1204    Description:
1205    
1206    Made a new interim version and bootfiles for developer's bootstrapping
1207    convenience.
1208    
1209       110.43.1 -- NEW BOOTFILES
1210    
1211    ----------------------------------------------------------------------
1212    Name: Matthias Blume (blume (at) tti - c (dot) org)
1213    Date: 2003/09/18 15:20:00 CDT
1214    Tag: blume-20030918-misc
1215    Description:
1216    
1217    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
1218       this seems to be controversial at the moment) and used those in
1219       Posix.ProcEnv.times.
1220    
1221    2. Added Time.{from,to}Nanoseconds to Time.
1222    
1223    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
1224       For example, fromLargeInt never needs to look at more than 3 "big
1225       digits" to get its 53 bits of precision.
1226    
1227    ----------------------------------------------------------------------
1228    Name: Matthias Blume (blume (at) tti - c (dot) org)
1229    Date: 2003/09/17 16:30:00 CDT
1230    Tag: blume-20030917-real32-slices
1231    Description:
1232    
1233    Added an entry to the primitive environment
1234    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
1235    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
1236    
1237    Removed some of the "magic" constants in real64.sml and replaced them
1238    with code that generates these values from their corresponding
1239    integer counterparts.
1240    
1241    Made all(?) the slice-related changes to the Basis and made everything
1242    compile again...
1243    
1244    ----------------------------------------------------------------------
1245    Name: Matthias Blume (blume (at) tti - c (dot) org)
1246    Date: 2003/09/15 17:45:00 CDT
1247    Tag: blume-20030915-rbase
1248    Description:
1249    
1250    Fixed bug in Real.fromLargeInt.
1251    
1252    ----------------------------------------------------------------------
1253    Name: Matthias Blume (blume (at) tti - c (dot) org)
1254    Date: 2003/09/13 18:11:00 CDT
1255    Tag: blume-20030913-libinstall
1256    Description:
1257    
1258    Minor bugfix in config/libinstall (set anchor with path to
1259    standalone tool after installing it, otherwise libraries that
1260    need ml-lex or ml-yacc won't compile the first time the installer
1261    runs).
1262    
1263    ----------------------------------------------------------------------
1264    Name: Matthias Blume (blume (at) tti - c (dot) org)
1265    Date: 2003/09/12 11:45:00 CDT
1266    Tag: blume-20030912-various
1267    Description:
1268    
1269     - fixed bug in Real.toLargeInt
1270     - fixed bug in Posix.ProcEnv.times
1271     - changed inputLine functions to return an option
1272     - minor installer improvements / bugfixes
1273     - changed default @SMLalloc parameter for x86/celeron to 64k
1274    
1275    ----------------------------------------------------------------------
1276    Name: Matthias Blume (blume (at) tti - c (dot) org)
1277    Date: 2003/09/09 22:00:00 CDT
1278    Tag: Release_110_43
1279    Description:
1280    
1281        New working release 110.43.  New bootfiles.
1282    
1283    ----------------------------------------------------------------------
1284    Name: Matthias Blume (blume (at) tti - c (dot) org)
1285    Date: 2003/09/09 19:20:00 CDT
1286    Tag: blume-20030909-installer
1287    Description:
1288    
1289    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
1290    Modified config/install.bat to take advantage of it.  Also modified
1291    config/install.sh (and called it config/new-install.sh) to take advantage
1292    of it on Unix systems.  (The SML code is (supposed to be) platform-
1293    independent.)
1294    
1295    The installer can now install everything under Win32
1296    as well as under *nix as long as it compiles.
1297    
1298    Other changes:
1299    
1300      - made CML compile again under Win32
1301      - made eXene compile under Win32 (by providing a fake structure UnixSock
1302        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
1303      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
1304        same as type int;  under Win32 it isn't
1305      - fixed some slice-related problems in the win32-specific parts of CML
1306      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
1307        (under Win32, the volume name is case-insensitive, and the
1308        OS.Path code compares volume names for equality)
1309    
1310    ----------------------------------------------------------------------
1311    Name: Matthias Blume (blume (at) tti - c (dot) org)
1312    Date: 2003/09/08 11:55:00 CDT
1313    Tag: blume-20030908-fullpath
1314    Description:
1315    
1316    Made Win32 version of OS.FileSys.fullPath return current directory
1317    when given an empty string.  This is what the spec says, and incidentally,
1318    CM depends on it.  (CM otherwise goes into an infinite loop in certain
1319    cases when presented with the name of a non-existing .cm file.)
1320    
1321    ----------------------------------------------------------------------
1322    Name: Matthias Blume (blume (at) tti - c (dot) org)
1323    Date: 2003/09/04 16:30:00 CDT
1324    Tag: blume-20030905-slices-etc
1325    Description:
1326    
1327      1. Changed interface to vectors and arrays in Basis to match
1328         (draft) Basis spec.
1329      2. Added signatures and implementations of slices according to
1330         Basis spec.
1331      3. Edited source code throughout the system to make it compile again
1332         under 1. and 2.  (In some cases code had to be added to have it
1333         match the new signatures.)
1334      4. MLRISC should be backward-compatible: the copies of the originals
1335         of files that needed to change under 3. were retained, the .cm files
1336         check the compiler version number and use old versions when
1337         appropriate.
1338      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
1339         dirstream -> string option (in accordance with Basis spec).
1340      6. When generating code that counts lines, ml-lex used function
1341         CharVector.foldli, taking advantage of its old interface.
1342         This has been replaced with the corresponding code from
1343         CharVectorSlice.  (html-lex must be re-lexed!)
1344      7. BitArray in smlnj-lib/Util has been extended/modified to match the
1345         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
1346      8. Removed temporary additions (fromInternal, toInternal) from the
1347         (now obsolete) IntInf in smlnj-lib/Util.
1348      9. Cleaned up structure Byte.
1349     10. Added localOffset, scan, and fromString to Date (according to spec).
1350         Cleaned/corrected implementation of Date.
1351         (Still need to check for correctness; implement better canonicalizeDate.)
1352     11. Added "scan" to signature IEEE_REAL.
1353     12. Some improvements to IntInf [in particular: efficiency-hack for
1354         mod and rem when second operand is 2 (for parity checks).]
1355     13. Changed representation of type Time.time, using a single IntInf.int
1356         value counting microseconds.  This considerably simplified the
1357         implementation of structure Time.  We now support negative time
1358         values; scan and fromString handle signs.
1359     14. Functor PrimIO now takes two additional arguments (VectorSlice and
1360         ArraySlice).
1361    
1362    ----------------------------------------------------------------------
1363    Name: Matthias Blume (blume (at) tti - c (dot) org)
1364    Date: 2003/08/28 17:00:00 CDT
1365    Tag: blume-20030828-intinf
1366    Description:
1367    
1368    This is a major update which comes with a version number bump
1369    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
1370    and an implementation of IntInf in the Basis.
1371    
1372    There are a fairly large number of related changes and updates throughout
1373    the system:
1374    
1375    Basis:
1376            - Implemented IntInf.
1377            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
1378            - Added some missing Real64 operations, most notably Real.toLargeInt.
1379            - Added FixedInt as a synonym for Int32.
1380    
1381    compiler:
1382            * Added support for a built-in intinf type.
1383              - literals
1384              - pattern matching
1385              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
1386              - overloading on literals and operations
1387    
1388            This required adding a primitive type intinf, some additional
1389            primops, and implementations for several non-trivial intinf
1390            operations in Core.  (The intinf type is completely abstract
1391            to the compiler; all operations get delegated back to the Core.)
1392    
1393            * Intinf equality is handled by polyequal.  However, the compiler
1394            does not print its usual warning in this case (since polyequal
1395            is the right thing to do there).
1396    
1397            * Improved the organization of structure InlineT.
1398    
1399            * A word about conversion primops:
1400            If conversions involving intinf do not cancel out during
1401            CPS contract, then the compiler must insert calls to Core functions.
1402            Since all core access must be resolved already during the FLINT
1403            translate phase, it would be too late a the time of CPS contract
1404            to add new Core calls.  For this reason, conversion primops
1405            for intinf carry two arguments:  1. the numeric argument that
1406            they are supposed to convert, and 2. the Core function that
1407            can help with this conversion if necessary.  If CPS contract
1408            eliminates a primop, then the associated Core function becomes
1409            dead and goes away.  Intinf conversion primops that do not get
1410            eliminated by CPS contract get rewritten into calls of their
1411            core functions by a separate, new phase.
1412    
1413    interactive system:
1414            - Control.Print.intinfDepth controls max length of intinf constants
1415            being printed. (Analogous to Control.Print.stringDepth.)
1416            - Cleanup in printutil and pputil: got rid of unused stuff and
1417            duplicates; replaced some of the code with code that makes better
1418            use of library functionality.
1419    
1420    CM:
1421            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
1422            are no longer silent.
1423    
1424    CKIT:
1425            Fixed mismatched uses of Int32 and LargeInt.  I always decided
1426            in favor of LargeInt -- which is now the same as IntInf.
1427            CKIT-knowledgable people should check whether this is what's
1428            intended and otherwise change things back to using Int32 or
1429            FixedInt.
1430    
1431    Throughout the code:
1432            Started using IntInf.int literals and built-in operations
1433            (e.g., comparison with 0) where this seems appropriate.
1434    
1435    
1436    ----------------------------------------------------------------------
1437    Name: Dave MacQueen (dbm@cs.uchicago.edu)
1438    Date: 2003/08/13 11:36:00 CDT
1439    Tag: dbm-20030813-mcz-merge1
1440    Description:
1441    
1442    Merging changes from the mcz-branch development branch into trunk.
1443    These changes involve replacement of the emulated old prettyprinter
1444    interface with direct use of the SML/NJ Lib PP library, and fixing
1445    of a couple of bugs (895, 1186) relating to error messages.  A new
1446    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
1447    has been added.
1448    
1449    ----------------------------------------------------------------------
1450    Name: Matthias Blume (blume (at) tti - c (dot) org)
1451    Date: 2003/08/11 15:45:00 CDT
1452    Tag: blume-20030811-windows
1453    Description:
1454    
1455    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
1456    
1457        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
1458    
1459    This patch restores SML/NJ's ability to run under win32.  There are a
1460    number of changes, including fixes for several bugs that had gone
1461    unnoticed until now:
1462    
1463      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
1464        trying to run the win32 version from within cygwin.)
1465    
1466      - There are a number of simple .bat scripts that substitute for their
1467        corresponding Unix shell-scripts.  (See below.)
1468    
1469      - The internals of ml-build have been modified slightly.  The main
1470        difference is that instead of calling ".link-sml" (or link-sml.bat)
1471        using OS.Process.system, the ML process delegates this task back
1472        to the script.  Otherwise problems arise in mixed environments such
1473        as Cygwin where scripts look and work like Unix scripts, but
1474        where OS.Process.system cannot run them.
1475    
1476      - In CM, the srcpath pickler used native pathname syntax -- which
1477        is incorrect in the case of cross-compilation.  The new pickle format
1478        is independent of platform-specific naming conventions.
1479    
1480      - Path configuration files (such as lib/pathconfig) can now choose
1481        between native and standard syntax.  Placing a line of the form
1482    
1483            standard!
1484    
1485        into the file causes all subsequent paths to be interpreted using
1486        CM standard pathname syntax (= Unix conventions); a line
1487    
1488            native!
1489    
1490        switches back to native style.  This was needed so that
1491        path config files can be written portably, see src/system/pathconfig.
1492    
1493      - Runtime system:
1494    
1495          - win32-filesys.c:  get_file_time and set_file_time now
1496            access modification time, not creation time.
1497    
1498          - I/O code made aware of new array representation.
1499    
1500          - Bug fixes in X86.prim.masm.
1501    
1502       - src/system/makeml made aware of win32. (For use under cygwin
1503            and other Unix-environments for windows.)
1504    
1505       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
1506         which caused BinIO.inputAll to fail consistently.
1507    
1508    .bat scripts:
1509    
1510       Windows .bat scripts assume that SMLNJ_HOME is defined.
1511    
1512       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
1513         applications (sml, ml-yacc, ml-lex).
1514       - ml-build.bat: analogous to ml-build.
1515       - config\install.bat:  Analogous to config/install.sh.  This requires
1516         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
1517         (nmake etc. must be on the path, and vcvars32 must have been run.)
1518         Moreover, sources for ml-lex and ml-yacc need to exist under src,
1519         and the bootfile hierarchy must have been unpacked under
1520         sml.boot.x86-win32.
1521         The script is very primitive and does a poor job at error checking.
1522         It only installs the base system, ml-lex, and ml-yacc.  No other
1523         libraries are being installed (i.e., you get only those that
1524         are part of the compiler.)
1525       - link-sml.bat:  analogous to .link-sml, but not currently used
1526    
1527    Unrelated bug fixes:
1528    
1529      - ml-nlffigen now exports structures ST_* corresponding to incomplete
1530        types.
1531      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
1532        otherwise.)
1533    
1534    ----------------------------------------------------------------------
1535    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
1536    Date: 2003/06/17
1537    Tag: macqueen-20030617-bug895
1538    Description:
1539    
1540    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
1541    Tag will be used for new development branch (mcz-branch) for
1542    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
1543    
1544    ----------------------------------------------------------------------
1545    Name: Matthias Blume (blume (at) tti - c (dot) org)
1546    Date: 2003/05/27 16:55:00 CDT
1547    Tag: blume-20030527-polyeq
1548    Description:
1549    
1550    Tried to eliminated most cases of polymorphic equality.
1551    
1552    ----------------------------------------------------------------------
1553    Name: Matthias Blume (blume (at) tti - c (dot) org)
1554    Date: 2003/05/21 17:45:00 CDT
1555    Tag: blume-20030517-complete
1556    Description:
1557    
1558    Two changes:
1559    
1560      1. Added a flag for controlling whether non-exhaustive bindings will
1561         be treated as errors (default is false).
1562      2. Cleaned up the *entire* source tree so that CMB.make goes through
1563         without a single non-exhaustive match- or bind warning.
1564    
1565    ----------------------------------------------------------------------
1566    Name: Matthias Blume (blume (at) tti - c (dot) org)
1567    Date: 2003/05/17 10:20:00 CDT
1568    Tag: blume-20030517-absyn
1569    Description:
1570    
1571    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
1572    
1573       This mainly affects the quality of error messages.  However, some
1574       of the code is now more straightforward than before. (Treatment of
1575       the above four constructs in translate.sml is much simpler than
1576       the "macro-expansion" that was going on before.  Plus, the mach-
1577       compiler no longer gets invoked just to be able to compile an
1578       if-expression.)
1579    
1580    2. The ErrorMsg.Error exception is now caught and absorbed by the
1581       interactive loop.
1582    
1583    ----------------------------------------------------------------------
1584    Name: Allen Leung
1585    Date: 2003/05/16 13:05:00 CDT
1586    Tag: leunga-20030516-cygwin-runtime
1587    Description:
1588    
1589        Ported the runtime system to cygwin, which uses the unix
1590    x86-unix bin files.  Missing/buggy features:
1591    
1592       o getnetbyname, getnetbyaddr: these functions seem to be missing in
1593         the Cygwin library.
1594       o Ctrl-C handling may be flaky.
1595       o Windows system calls and Windows I/O are not supported.
1596    
1597       A new set of binfiles is located at:
1598    
1599         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
1600    
1601       This is only needed for bootstrapping the cygwin version of smlnj.
1602       Other x86 versions can use the existing binfiles.
1603    
1604    ----------------------------------------------------------------------
1605    Name: Matthias Blume
1606    Date: 2003/04/08 15:42:00 CDT
1607    Tag: blume-20030408-listpair
1608    Description:
1609    
1610    1. Added a target 'mlrisc' to installer.
1611    
1612    2. Added missing elements to structure ListPair.
1613    
1614    ----------------------------------------------------------------------
1615    Name: Allen Leung
1616    Date: 2003/01/07 10:40:00 EST
1617    Tag: leunga-20030107-int-rem
1618    Description:
1619    
1620    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
1621    The arguments to the SUBL instruction were swapped.
1622    
1623    ----------------------------------------------------------------------
1624    Name: Matthias Blume
1625    Date: 2002/12/12 16:25:00 EST
1626    Tag: blume-20021212-risc-ra
1627    Description:
1628    
1629    Fixed a serious bug in the rewrite code for FP spilling/reloading that
1630    sent the RA into an infinite loop when floating point registers get
1631    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
1632    110.37 and 110.38.)
1633    There was another set of potential problems related to the handling of
1634    MLRISC annotations (but those did not yet cause real problems, apparently).
1635    
1636    ----------------------------------------------------------------------
1637    Name: Matthias Blume
1638    Date: 2002/12/06 22:40:00 EST
1639    Tag: blume-20021206-cm-fileid
1640    Description:
1641    
1642    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
1643    This fixes the problem of CM getting confused by files that suddenly
1644    change their identity (e.g., by getting unlinked and recreated by some
1645    text editor such as vi).  There might be a better/cheaper/cleaner way
1646    of doing this, but for now this will have to do.
1647    
1648    ----------------------------------------------------------------------
1649    Name: Matthias Blume
1650    Date: 2002/10/28 09:50:00 EST
1651    Tag: blume-20021028-typecheck
1652    Description:
1653    
1654    Exported structure Typecheck from $smlnj/viscomp/core.cm.
1655    
1656    ----------------------------------------------------------------------
1657    Name: Matthias Blume
1658    Date: 2002/10/17 09:10:00 EDT
1659    Tag: Release_110_42
1660    Description:
1661    
1662    In good old tradition, there has been a slight hiccup so that we have
1663    to patch 110.42 after the fact.   The old release tag has been replaced
1664    (see below).
1665    
1666    The change solves a problem with two competing approaches the
1667    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
1668    each other's way.
1669    
1670    This change only affects the runtime system code and the installer script.
1671    (No new bootfiles.)
1672    
1673    ----------------------------------------------------------------------
1674    Name: Matthias Blume
1675    Date: 2002/10/16 12:00:00 EDT
1676    Tag: Release_110_42_removed
1677    Description:
1678    
1679    New working release.  New bootfiles.
1680    
1681    ----------------------------------------------------------------------
1682    Name: Matthias Blume
1683    Date: 2002/10/10 13:10:00 EDT
1684    Tag: blume-20021010-ppc-divs
1685    Description:
1686    
1687    The mltree operator DIVS must be implemented with an overflow check on
1688    the PPC because the hardware indicates divide-by-zero using "overflow" as
1689    well.
1690    
1691    ----------------------------------------------------------------------
1692    Name: Matthias Blume
1693    Date: 2002/07/23 11:45:00
1694    Tag: blume-20020723-smlnj-home
1695    Description:
1696    
1697    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
1698    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
1699    CM_PATHCONFIG is also set), the path configuration file is assumed
1700    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
1701    the entire tree to some other place and everything will "just work".
1702    
1703    (Companion commands such as ml-build and ml-makedepend also sense this
1704    variable.)
1705    
1706    ----------------------------------------------------------------------
1707    Name: Matthias Blume
1708    Date: 2002/07/12 21:19:00 EDT
1709    Tag: blume-20020712-liveness
1710    Description:
1711    
1712    Exported two useful "step" functions from liveness module (MLRISC).
1713    
1714    ----------------------------------------------------------------------
1715    Name: Matthias Blume
1716    Date: 2002/07/05 16:00 EDT
1717    Tag: Release_110_41
1718    Description:
1719    
1720    New working release.  New bootfiles.
1721    
1722    ----------------------------------------------------------------------
1723    Name: Matthias Blume
1724    Date: 2002/07/05 10:25:00 EDT
1725    Tag: blume-20020705-btimp
1726    Description:
1727    
1728    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
1729    other clients can set up backtracing support.
1730    
1731    ----------------------------------------------------------------------
1732    Name: Matthias Blume
1733    Date: 2002/06/25 17:23:00 EDT
1734    Tag: blume-20020625-fpmax
1735    Description:
1736    
1737    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
1738    case.  (The sense of the isNaN test was reversed -- which made min and
1739    max always return their first argument.)
1740    
1741    ----------------------------------------------------------------------
1742    Name: Matthias Blume
1743    Date: 2002/06/11
1744    Tag: blume-20020611-unixpath
1745    Description:
1746    
1747    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
1748    
1749    ----------------------------------------------------------------------
1750    Name: Matthias Blume
1751    Date: 2002/06/10 16:35:00 EDT
1752    Tag: blume-20020610-ieeereal
1753    Description:
1754    
1755    I back-ported my implementation of IEEEReal.fromString from the
1756    idlbasis-devel branch so that we can test it.
1757    
1758    Another small change is that ppDec tries to give more information
1759    than just "<sig>" in the case of functors.  However, this code is
1760    broken in some mysterious way if the functor's body's signature
1761    has not been declared by ascription but gets inferred from the
1762    implementation.  This needs fixing...
1763    
1764    ----------------------------------------------------------------------
1765    Name: Matthias Blume
1766    Date: 2002/05/31
1767    Tag: blume-20020531-btrace-mode
1768    Description:
1769    
1770    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
1771    the wayside when I switched over to using Controls everywhere.)
1772    
1773    ----------------------------------------------------------------------
1774    Name: Lal George
1775    Date: 2002/05/23 12:21:40 EDT
1776    Tag: george-20020523-visual-labels
1777    Description:
1778    
1779    Labels are now displayed in the graphical output to make
1780    the fall-through and target blocks obvious.
1781    
1782    ----------------------------------------------------------------------
1783    Name: Matthias Blume
1784    Date: 2002/05/22 11:03:00 EDT
1785    Tag: blume-20020522-shrink
1786    Description:
1787    
1788    John tweaked yesterday's fix for 1131 to handle an out-of-memory
1789    situation that comes up when allocating huge arrays.
1790    
1791    ----------------------------------------------------------------------
1792    Name: Matthias Blume
1793    Date: 2002/05/21 16:00:00 EDT
1794    Tag: Release_110_40
1795    Description:
1796    
1797    New working release (110.40).  New bootfiles.
1798    
1799    [Also: John Reppy fixed GC bug 1131.]
1800    
1801    ----------------------------------------------------------------------
1802    Name: Matthias Blume
1803    Date: 2002/05/21 12:35:00 EDT
1804    Tag: blume-20020521-cmdoc
1805    Description:
1806    
1807    CM documentation update.
1808    
1809    ----------------------------------------------------------------------
1810    Name: Matthias Blume
1811    Date: 2002/05/21 10:55:00 EDT
1812    Tag: blume-20020521-misc
1813    Description:
1814    
1815    - John tweaked runtime to be silent on heap export (except when
1816      GC messages are on).
1817    - I added a few more things (cross-compiling versions of CMB) to
1818      config/preloads (as suggestions).
1819    
1820    ----------------------------------------------------------------------
1821    Name: Matthias Blume
1822    Date: 2002/05/20 22:25:00 EDT
1823    Tag: blume-20020520-controls
1824    Description:
1825    
1826    - Added ControlUtil structure to control-lib.cm.
1827    - Use it throughout.
1828    - Used Controls facility to define MLRISC controls (as opposed to
1829      registering MLRISC control ref cells with Controls after the fact)
1830    - Fixed messed-up controls priorities.
1831    
1832    * Removed again all the stuff from config/preloads that one wouldn't
1833      be able to preload at the time the initial heap image is built.
1834      (Many libraries, e.g., CML, do not exist yet at this time.  The
1835       only libraries that can be preloaded via config/preloads are those
1836       that come bundled with the bootfiles.)
1837    
1838    ----------------------------------------------------------------------
1839    Name: Matthias Blume
1840    Date: 2002/05/20 10:59:00 EDT
1841    Tag: blume-20020520-preloads
1842    Description:
1843    
1844    Added a lot of commented-out suggestions for things to be included
1845    in config/preloads.
1846    
1847    ----------------------------------------------------------------------
1848    Name: Allen Leung
1849    Date: 2002/05/18 14:20:00 EDT
1850    Tag: leunga-20020518-mdl
1851    Description:
1852    
1853    o Made the mdl tool stuff compile and run again.
1854    o I've disabled all the stuff that depends on RTL specifications; they
1855      are all badly broken anyway.
1856    
1857    ----------------------------------------------------------------------
1858    Name: Matthias Blume
1859    Date: 2002/05/17 16:49:00 EDT
1860    Tag: blume-20020517-controls
1861    Description:
1862    
1863    0. John Reppy made several modifications to the SML/NJ library.
1864       In particular, there is a shiny new controls-lib.cm.
1865    
1866    1. Pushed new controls interface through compiler so that everything
1867       compiles again.
1868    
1869    2. Added FormatComb and FORMAT_COMB to the CML version of the
1870       SML/NJ library (so that CML compiles again).
1871    
1872    3. Modified init scripts because XXX_DEFAULT environment variables
1873       are no longer with us.  (Boot-time initialization is now done
1874       using the same environment variables that are also used for
1875       startup-time initialization of controls.)
1876    
1877    ----------------------------------------------------------------------
1878    Name: Lal George
1879    Date: 2002/05/15 09:20:10 EDT
1880    Tag: george-20020515-pseudo-op-decls
1881    Description:
1882    
1883            All pseudo-ops emitted before the first segment declaration
1884    such as TEXT, DATA, and BSS directives are assumed to be global
1885    declarations and are emitted first in the assembly file. This is
1886    useful in a number of situations where one has pseudo-ops that are not
1887    specific to any segment, and also works around the constraint that one
1888    cannot have client pseudo-ops in the TEXT segment.
1889    
1890            Because no segment is associated with these declarations it is
1891    an error to allocate any space or objects before the first segment
1892    directive and an exception will be raised. However, we cannot make
1893    this check for client pseudo-ops.
1894    
1895            These top level declarations are a field in the CFG graph_info.
1896    In theory you can continue to add to this field after the CFG has been
1897    built -- provided you know what you are doing;-)
1898    
1899    ----------------------------------------------------------------------
1900    Name: Matthias Blume
1901    Date: 2002/05/13 16:40:00 EDT
1902    Tag: blume-20020513-pp-etc
1903    Description:
1904    
1905    A few minor bugfixes:
1906    
1907      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
1908        (Bogus printouts for redefined bindings still occur.  Compiler
1909        bug should no longer occur now.  We need to redo the prettyprinter
1910        from scratch.)
1911    
1912      - CM pathname printer now also adds escape sequences for ( and )
1913    
1914      - commend and docu fixes for ml-nlffi
1915    
1916    ----------------------------------------------------------------------
1917    Name: Matthias Blume
1918    Date: 2002/05/10 16:40:00 EDT
1919    Tag: blume-20020510-erg-textio
1920    Description:
1921    
1922    Applied the following bugfix provided by Emden Gansner:
1923    
1924        Output is corrupted when outputSubstr is used rather than output.
1925    
1926        The problem occurs when a substring
1927    
1928            ss = (s, dataStart, dataLen)
1929    
1930        where dataStart > 0, fills a stream buffer with avail bytes left.
1931        avail bytes of s, starting at index dataStart, are copied into the
1932        buffer, the buffer is flushed, and then the remaining dataLen-avail
1933        bytes of ss are copied into the beginning of the buffer. Instead of
1934        starting this copy at index dataStart+avail in s, the current code
1935        starts the copy at index avail.
1936    
1937        Fix:
1938        In text-io-fn.sml, change line 695 from
1939             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
1940        to
1941             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
1942    
1943    ----------------------------------------------------------------------
1944    Name: Matthias Blume
1945    Date: 2002/04/12 13:55:00 EDT
1946    Tag: blume-20020412-assyntax
1947    Description:
1948    
1949    1. Grabbed newer assyntax.h from the XFree86 project.
1950    2. Fiddled with how to compile X86.prim.asm without warnings.
1951    3. (Very) Minor cleanup in CM.
1952    
1953    ----------------------------------------------------------------------
1954    Name: Matthias Blume
1955    Date: 2002/04/01 (no joke!) 17:07:00 EST
1956    Tag: blume-20020401-x86div
1957    Description:
1958    
1959    Added full support for div/mod/rem/quot on the x86, using the machine
1960    instruction's two results (without clumsily recomputing the remainder)
1961    directly where appropriate.
1962    
1963    Some more extensive power-of-two support was added to the x86 instruction
1964    selector (avoiding expensive divs, mods, and muls where they can be
1965    replaced with cheaper shifts and masks).  However, this sort of thing
1966    ought to be done earlier, e.g., within the CPS optimizer so that
1967    all architectures benefit from it.
1968    
1969    The compiler compiles to a fixed point, but changes might be somewhat
1970    fragile nevertheless.  Please, report any strange things that you might
1971    see wrt. div/mod/quot/rem...
1972    
1973    ----------------------------------------------------------------------
1974    Name: Matthias Blume
1975    Date: 2002/03/29 17:22:00
1976    Tag: blume-20020329-div
1977    Description:
1978    
1979    Fixed my broken div/mod logic.  Unfortunately, this means that the
1980    inline code for div/mod now has one more comparison than before.
1981    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
1982    The problem was with quotient = 0, because that alone does not tell
1983    us which way the rounding went.  One then has to look at whether
1984    remainder and divisor have the same sign...  :(
1985    
1986    Anyway, I replaced the bootfiles with fresh ones...
1987    
1988    ----------------------------------------------------------------------
1989    Name: Matthias Blume
1990    Date: 2002/03/29 14:10:00 EST
1991    Tag: blume-20020329-inlprims
1992    Description:
1993    
1994    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
1995    
1996    Primops have changed. This means that the bin/boot-file formats have
1997    changed as well.
1998    
1999    To make sure that there is no confusion, I made a new version.
2000    
2001    
2002    CHANGES:
2003    
2004    * removed REMT from mltree (remainder should never overflow).
2005    
2006    * added primops to deal with divisions of all flavors to the frontend
2007    
2008    * handled these primops all the way through so they map to their respective
2009      MLRISC support
2010    
2011    * used these primops in the implementation of Int, Int32, Word, Word32
2012    
2013    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2014    
2015    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2016    
2017    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2018      floating point
2019    
2020    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2021      and Real (but Real.abs maps to a separate floating-point-only primop)
2022    
2023    
2024    TODO items:
2025    
2026    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2027      instructions because the machine instruction encoder cannot handle
2028      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2029      REM in the first place, and REM is supported by the x86 machine coder.)
2030    
2031    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2032      selection phase.  (The two can be streamlined because the hardware
2033      delivers both quotient and remainder at the same time anyway.)
2034    
2035    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2036      (Currently the behavior is inconsistent both across architectures and
2037      wrt. the draft Basis spec.)
2038    
2039    * Word8 should eventually be handled natively, too.
2040    
2041    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2042      as if there currently is no execution path that could trigger it in
2043      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2044      hold for things like multiplication and division.)
2045    
2046    ----------------------------------------------------------------------
2047    Name: Matthias Blume
2048    Date: 2002/03/27 16:27:00 EST
2049    Tag: blume-20020327-mlrisc-divisions
2050    Description:
2051    
2052    Added support for all four division operations (ML's div, mod, quot,
2053    and rem) to MLRISC.  In the course of doing so, I also rationalized
2054    the naming (no more annoying switch-around of DIV and QUOT), by
2055    parameterizing the operation by div_rounding_mode (which can be either
2056    DIV_TO_ZERO or DIV_TO_NEGINF).
2057    
2058    The generic MLTreeGen functor takes care of compiling all four
2059    operations down to only round-to-zero div.
2060    
2061    Missing pieces:
2062    
2063      * Doing something smarter than relying on MLTreeGen on architectures
2064        like, e.g., the x86 where hardware division delivers both quotient and
2065        remainder at the same time.  With this, the implementation of the
2066        round-to-neginf operations could be further streamlined.
2067    
2068      * Remove inlining support for div/mod/rem from the frontend and replace it
2069        with primops that get carried through to the backend.  Do this for all
2070        int and word types.
2071    
2072    ----------------------------------------------------------------------
2073    Name: Matthias Blume
2074    Date: 2002/03/25 17:25:00 EST
2075    Tag: blume-20020325-divmod
2076    Description:
2077    
2078    I improved (hopefully without breaking them) the implementation of Int.div,
2079    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2080    advantage of the following observations:
2081    
2082      Let  q = x quot y      r = x rem y
2083           d = x div  y      m = x mod y
2084    
2085    where "quot" is the round-to-zero version of integer division that
2086    hardware usually provides.  Then we have:
2087    
2088         r = x - q * y        where neither the * nor the - will overflow
2089         d = if q >= 0 orelse x = q * y then q else q - 1
2090                              where neither the * nor the - will overflow
2091         m = if q >= 0 orelse r = 0 then r else r + y
2092                              where the + will not overflow
2093    
2094    This results in substantial simplification of the generated code.
2095    The following table shows the number of CFG nodes and edges generated
2096    for
2097            fun f (x, y) = x OPER y
2098            (* with OPER \in div, mod, quot, rem *)
2099    
2100    
2101        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2102        --------------------------------------------------------
2103         div |         24 |         39 |         12 |         16
2104         mod |         41 |         71 |         12 |         16
2105        quot |          8 |         10 |          8 |         10
2106         rem |         10 |         14 |          8 |         10
2107    
2108    
2109    ----------------------------------------------------------------------
2110    Name: Matthias Blume
2111    Date: 2002/03/25 22:06:00 EST
2112    Tag: blume-20020325-cprotobug
2113    Description:
2114    
2115    Fixed a bug in cproto (c prototype decoder).
2116    
2117    ----------------------------------------------------------------------
2118    Name: Matthias Blume
2119    Date: 2002/03/25 16:00:00 EST
2120    Tag: blume-20020325-raw-primops
2121    Description:
2122    
2123    I did some cleanup to Allen's new primop code and
2124    replaced yesterday's bootfiles with new ones.
2125    (But they are stored in the same place.)
2126    
2127    ----------------------------------------------------------------------
2128    Name: Matthias Blume
2129    Date: 2002/03/24 22:40:00 EST
2130    Tag: blume-20020324-bootfiles
2131    Description:
2132    
2133    Made the bootfiles that Allen asked for.
2134    
2135    ----------------------------------------------------------------------
2136    Name: Allen Leung
2137    Date: 2002/03/23 15:50:00 EST
2138    Tag: leunga-20020323-flint-cps-rcc-primops
2139    Description:
2140    
2141      1. Changes to FLINT primops:
2142    
2143        (* make a call to a C-function;
2144         * The primop carries C function prototype information and specifies
2145         * which of its (ML-) arguments are floating point. C prototype
2146         * information is for use by the backend, ML information is for
2147         * use by the CPS converter. *)
2148      | RAW_CCALL of { c_proto: CTypes.c_proto,
2149                       ml_args: ccall_type list,
2150                       ml_res_opt: ccall_type option,
2151                       reentrant : bool
2152                     } option
2153       (* Allocate uninitialized storage on the heap.
2154        * The record is meant to hold short-lived C objects, i.e., they
2155        * are not ML pointers.  With the tag, the representation is
2156        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2157        *)
2158      | RAW_RECORD of {tag:bool,sz:int}
2159      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2160    
2161      2.  These CPS primops are now overloaded:
2162    
2163           rawload of {kind:numkind}
2164           rawstore of {kind:numkind}
2165    
2166          The one argument form is:
2167    
2168             rawload {kind} address
2169    
2170          The two argument form is:
2171    
2172             rawload {kind} [ml object, byte-offset]
2173    
2174      3. RAW_CCALL/RCC now takes two extra arguments:
2175    
2176         a. The first is whether the C call is reentrant, i.e., whether
2177            ML state should be saved and restored.
2178         b. The second argument is a string argument specifying the name of
2179            library and the C function.
2180    
2181         These things are currently not handled in the code generator, yet.
2182    
2183      4. In CProto,
2184    
2185         An encoding type of "bool" means "ml object" and is mapped into
2186         C prototype of PTR.  Note that "bool" is different than "string",
2187         even though "string" is also mapped into PTR, because "bool"
2188         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2189    
2190      5. Pickler/unpicker
2191    
2192         Changed to handle RAW_RECORD and newest RAW_CCALL
2193    
2194      6. MLRiscGen,
2195    
2196         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2197         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2198            in the file CodeGen/cpscompile/cps-c-calls.sml
2199    
2200      7. Added the conditional move operator
2201    
2202             condmove of branch
2203    
2204         to cps.  Generation of this is still buggy so it is currently
2205         disabled.
2206    
2207    ----------------------------------------------------------------------
2208    Name: Lal George
2209    Date: 2002/03/22 14:18:25 EST
2210    Tag: george-20020322-cps-branch-prob
2211    Description:
2212    
2213    Implemented the Ball-Larus branch prediction-heuristics, and
2214    incorporated graphical viewers for control flow graphs.
2215    
2216    Ball-Larus Heuristics:
2217    ---------------------
2218    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
2219    
2220    By design it uses the Dempster-Shafer theory for combining
2221    probabilities.  For example, in the function:
2222    
2223        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
2224    
2225    the ball-larus heuristics predicts that the n=0 is unlikely
2226    (OH-heuristic), and the 'then' branch is unlikely because of the
2227    RH-heuristic -- giving the 'then' branch an even lower combined
2228    probability using the Dempster-Shafer theory.
2229    
2230    Finally, John Reppy's loop analysis in MLRISC, further lowers the
2231    probability of the 'then' branch because of the loop in the else
2232    branch.
2233    
2234    
2235    Graphical Viewing:
2236    ------------------
2237    I merely plugged in Allen's graphical viewers into the compiler. The
2238    additional code is not much. At the top level, saying:
2239    
2240            Control.MLRISC.getFlag "cfg-graphical-view" := true;
2241    
2242    will display the graphical view of the control flow graph just before
2243    back-patching.  daVinci must be in your path for this to work. If
2244    daVinci is not available, then the default viewer can be changed
2245    using:
2246    
2247            Control.MLRISC.getString "viewer"
2248    
2249    which can be set to "dot" or "vcg" for the corresponding viewers. Of
2250    course, these viewers must be in your path.
2251    
2252    The above will display the compilation unit at the level of clusters,
2253    many of which are small, boring, and un-interesting. Also setting:
2254    
2255            Control.MLRISC.getInt "cfg-graphical-view_size"
2256    
2257    will display clusters that are larger than the value set by the above.
2258    
2259    
2260    ----------------------------------------------------------------------
2261    Name: Matthias Blume
2262    Date: 2002/03/21 22:20:00 EST
2263    Tag: blume-20020321-kmp-bugfix
2264    Description:
2265    
2266    Changed the interface to the KMP routine in PreString and fixed
2267    a minor bug in one place where it was used.
2268    
2269    ----------------------------------------------------------------------
2270    Name: Allen Leung
2271    Date: 2002/03/21 20:30:00 EST
2272    Tag: leunga-20020321-cfg
2273    Description:
2274    
2275      Fixed a potential problem in cfg edge splitting.
2276    
2277    ----------------------------------------------------------------------
2278    Name: Allen Leung
2279    Date: 2002/03/21 17:15:00 EST
2280    Tag: leunga-20020321-x86-fp-cfg
2281    Description:
2282    
2283      1. Recoded the buggy parts of x86-fp.
2284    
2285         a. All the block reordering code has been removed.
2286            We now depend on the block placement phases to do this work.
2287    
2288         b. Critical edge splitting code has been simplified and moved into the
2289            CFG modules, as where they belong.
2290    
2291         Both of these were quite buggy and complex.  The code is now much, much
2292         simpler.
2293    
2294      2. X86 backend.
2295    
2296         a. Added instructions for 64-bit support.  Instruction selection for
2297            64-bit has not been committed, however, since that
2298            requires changes to MLTREE which haven't been approved by
2299            Lal and John.
2300    
2301         b. Added support for FUCOMI and FUCOMIP when generating code for
2302            PentiumPro and above.  We only generate these instructions in
2303            the fast-fp mode.
2304    
2305         c. Added cases for JP and JNP in X86FreqProps.
2306    
2307      3. CFG
2308    
2309         CFG now has a bunch of methods for edge splitting and merging.
2310    
2311      4. Machine description.
2312    
2313         John's simplification of MLTREE_BASIS.fcond broke a few machine
2314         description things:
2315    
2316         rtl-build.{sig,sml} and hppa.mdl fixed.
2317    
2318         NOTE: the machine description stuff in the repository is still broken.
2319               Again, I can't put my fixes in because that involves
2320               changes to MLTREE.
2321    
2322    ----------------------------------------------------------------------
2323    Name: Matthias Blume
2324    Date: 2002/03/20 15:55:00 EST
2325    Tag: blume-20020320-kmp
2326    Description:
2327    
2328    Implemented Knuth-Morris-Pratt string matching in PreString and used
2329    it for String.isSubstring, Substring.isSubstring, and
2330    Substring.position.
2331    
2332    (Might need some stress-testing.  Simple examples worked fine.)
2333    
2334    ----------------------------------------------------------------------
2335    Name: Matthias Blume
2336    Date: 2002/03/19 16:37:00 EST
2337    Tag: blume-20020319-witnesses
2338    Description:
2339    
2340    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
2341    
2342    This implements a generic mechanism for changing constness qualifiers
2343    anywhere within big C types without resorting to outright "casts".
2344    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
2345    modify the constness at the outermost level.)
2346    The implementation of "convert" is based on the idea of "witness"
2347    values -- values that are not used by the operation but whose types
2348    "testify" to their applicability.  On the implementation side, "convert"
2349    is simply a projection (returning its second curried argument).  With
2350    cross-module inlining, it should not result in any machine code being
2351    generated.
2352    
2353    ----------------------------------------------------------------------
2354    Name: Matthias Blume
2355    Date: 2002/03/15 16:40:00 EST
2356    Tag: blume-20020315-basis
2357    Description:
2358    
2359    Provided (preliminary?) implementations for
2360    
2361      {String,Substring}.{concatWith,isSuffix,isSubstring}
2362    
2363    and
2364    
2365      Substring.full
2366    
2367    Those are in the Basis spec but they were missing in SML/NJ.
2368    
2369    ----------------------------------------------------------------------
2370    Name: Matthias Blume
2371    Date: 2002/03/14 21:30:00 EST
2372    Tag: blume-20020314-controls
2373    Description:
2374    
2375    Controls:
2376    ---------
2377    
2378    1. Factored out the recently-added Controls : CONTROLS stuff and put
2379       it into its own library $/controls-lib.cm.  The source tree for
2380       this is under src/smlnj-lib/Controls.
2381    
2382    2. Changed the names of types and functions in this interface, so they
2383       make a bit more "sense":
2384    
2385          module -> registry
2386          'a registry -> 'a group
2387    
2388    3. The interface now deals in ref cells only.  The getter/setter interface
2389       is (mostly) gone.
2390    
2391    4. Added a function that lets one register an already-existing ref cell.
2392    
2393    5. Made the corresponding modifications to the rest of the code so that
2394       everything compiles again.
2395    
2396    6. Changed the implementation of Controls.MLRISC back to something closer
2397       to the original.  In particular, this module (and therefore MLRISC)
2398       does not depend on Controls.  There now is some link-time code in
2399       int-sys.sml that registers the MLRISC controls with the Controls
2400       module.
2401    
2402    CM:
2403    ---
2404    
2405      * One can now specify the lambda-split aggressiveness in init.cmi.
2406    
2407    ----------------------------------------------------------------------
2408    Name: Allen Leung
2409    Date: 2002/03/13 17:30:00 EST
2410    Tag: leunga-20020313-x86-fp-unary
2411    Description:
2412    
2413    Bug fix for:
2414    
2415    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
2416    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
2417    > - fun f(x,(y,z)) = Real.~ y;
2418    > [autoloading]
2419    > [autoloading done]
2420    >       fchsl   (%eax), 184(%esp)
2421    > Error: MLRisc bug: X86MCEmitter.emitInstr
2422    >
2423    > uncaught exception Error
2424    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
2425    
2426    The problem was that the code generator did not generate any fp registers
2427    in this case, and the ra didn't know that it needed to run the X86FP phase to
2428    translate the pseudo fp instruction.   This only happened with unary fp
2429    operators in certain situations.
2430    
2431    ----------------------------------------------------------------------
2432    Name: Matthias Blume
2433    Date: 2002/03/13 14:00:00 EST
2434    Tag: blume-20020313-overload-etc
2435    Description:
2436    
2437    1. Added _overload as a synonym for overload for backward compatibility.
2438       (Control.overloadKW must be true for either version to be accepted.)
2439    
2440    2. Fixed bug in install script that caused more things to be installed
2441       than what was requested in config/targets.
2442    
2443    3. Made CM aware of the (_)overload construct so that autoloading
2444       works.
2445    
2446    ----------------------------------------------------------------------
2447    Name: Matthias Blume
2448    Date: 2002/03/12 22:03:00 EST
2449    Tag: blume-20020312-url
2450    Description:
2451    
2452    Forgot to update BOOT and srcarchiveurl.
2453    
2454    ----------------------------------------------------------------------
2455    Name: Matthias Blume
2456    Date: 2002/03/12 17:30:00 EST
2457    Tag: blume-20020312-version110392
2458    Description:
2459    
2460    Yet another version number bump (because of small changes to the
2461    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
2462    
2463    Changes:
2464    
2465      The new pid generation scheme described a few weeks ago was overly
2466      complicated.  I implemented a new mechanism that is simpler and
2467      provides a bit more "stability":  Once CM has seen a compilation
2468      unit, it keeps its identity constant (as long as you do not delete
2469      those crucial CM/GUID/* files).  This means that when you change
2470      an interface, compile, then go back to the old interface, and
2471      compile again, you arrive at the original pid.
2472    
2473      There now also is a mechanism that instructs CM to use the plain
2474      environment hash as a module's pid (effectively making its GUID
2475      the empty string).  For this, "noguid" must be specified as an
2476      option to the .sml file in question within its .cm file.
2477      This is most useful for code that is being generated by tools such
2478      as ml-nlffigen (because during development programmers tend to
2479      erase the tool's entire output directory tree including CM's cached
2480      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
2481      revert to the old, broken behavior of SML/NJ, but in specific cases
2482      where there is no danger of interface confusion, its use is ok
2483      (I think).
2484    
2485      ml-nlffigen by default generates "noguid" annotations.  They can be
2486      turned off by specifying -guid in its command line.
2487    
2488    ----------------------------------------------------------------------
2489    Name: Lal George
2490    Date: 2002/03/12 12 14:42:36 EST
2491    Tag: george-20020312-frequency-computation
2492    Description:
2493    
2494    Integrated jump chaining and static block frequency into the
2495    compiler. More details and numbers later.
2496    
2497    ----------------------------------------------------------------------
2498    Name: Lal George
2499    Date: 2002/03/11 11 22:38:53 EST
2500    Tag: george-20020311-jump-chain-elim
2501    Description:
2502    
2503    Tested the jump chain elimination on all architectures (except the
2504    hppa).  This is on by default right now and is profitable for the
2505    alpha and x86, however, it may not be profitable for the sparc and ppc
2506    when compiling the compiler.
2507    
2508    The gc test will typically jump to a label at the end of the cluster,
2509    where there is another jump to an external cluster containing the actual
2510    code to invoke gc. This is to allow factoring of common gc invocation
2511    sequences. That is to say, we generate:
2512    
2513            f:
2514               testgc
2515               ja   L1      % jump if above to L1
2516    
2517            L1:
2518               jmp L2
2519    
2520    
2521    After jump chain elimination the 'ja L1' instructions is converted to
2522    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
2523    up being implemented in their long form (if L2 is far away) using:
2524    
2525            jbe     L3      % jump if below or equal to L3
2526            jmp     L2
2527         L3:
2528            ...
2529    
2530    
2531    For large compilation units L2  may be far away.
2532    
2533    
2534    ----------------------------------------------------------------------
2535    Name: Matthias Blume
2536    Date: 2002/03/11 13:30:00 EST
2537    Tag: blume-20020311-mltreeeval
2538    Description:
2539    
2540    A functor parameter was missing.
2541    
2542    ----------------------------------------------------------------------
2543    Name: Allen Leung
2544    Date: 2002/03/11 10:30:00 EST
2545    Tag: leunga-20020311-runtime-string0
2546    Description:
2547    
2548       The representation of the empty string now points to a
2549    legal null terminated C string instead of unit.  It is now possible
2550    to convert an ML string into C string with InlineT.CharVector.getData.
2551    This compiles into one single machine instruction.
2552    
2553    ----------------------------------------------------------------------
2554    Name: Allen Leung
2555    Date: 2002/03/10 23:55:00 EST
2556    Tag: leunga-20020310-x86-call
2557    Description:
2558    
2559       Added machine generation for CALL instruction (relative displacement mode)
2560    
2561    ----------------------------------------------------------------------
2562    Name: Matthias Blume
2563    Date: 2002/03/08 16:05:00
2564    Tag: blume-20020308-entrypoints
2565    Description:
2566    
2567    Version number bumped to 110.39.1.  NEW BOOTFILES!
2568    
2569    Entrypoints: non-zero offset into a code object where execution should begin.
2570    
2571    - Added the notion of an entrypoint to CodeObj.
2572    - Added reading/writing of entrypoint info to Binfile.
2573    - Made runtime system bootloader aware of entrypoints.
2574    - Use the address of the label of the first function given to mlriscGen
2575      as the entrypoint.  This address is currently always 0, but it will
2576      not be 0 once we turn on block placement.
2577    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
2578      with entry points) from mlriscGen.
2579    
2580    ----------------------------------------------------------------------
2581    Name: Allen Leung
2582    Date: 2002/03/07 20:45:00 EST
2583    Tag: leunga-20020307-x86-cmov
2584    Description:
2585    
2586       Bug fixes for CMOVcc on x86.
2587    
2588       1. Added machine code generation for CMOVcc
2589       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
2590       3. CMOVcc cannot have an immediate operand as argument.
2591    
2592    ----------------------------------------------------------------------
2593    Name: Matthias Blume
2594    Date: 2002/03/07 16:15:00 EST
2595    Tag: blume-20020307-controls
2596    Description:
2597    
2598    This is a very large but mostly boring patch which makes (almost)
2599    every tuneable compiler knob (i.e., pretty much everything under
2600    Control.* plus a few other things) configurable via both the command
2601    line and environment variables in the style CM did its configuration
2602    until now.
2603    
2604    Try starting sml with '-h' (or, if you are brave, '-H')
2605    
2606    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
2607    implements the underlying generic mechanism.
2608    
2609    The interface to some of the existing such facilities has changed somewhat.
2610    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
2611    (The getFoo interface is still there for backward-compatibility, but its
2612    use is deprecated.)
2613    
2614    The ml-build script passes -Cxxx=yyy command-line arguments through so
2615    that one can now twiddle the compiler settings when using this "batch"
2616    compiler.
2617    
2618    TODO items:
2619    
2620    We should go through and throw out all controls that are no longer
2621    connected to anything.  Moreover, we should go through and provide
2622    meaningful (and correct!) documentation strings for those controls
2623    that still are connected.
2624    
2625    Currently, multiple calls to Controls.new are accepted (only the first
2626    has any effect).  Eventually we should make sure that every control
2627    is being made (via Controls.new) exactly once.  Future access can then
2628    be done using Controls.acc.
2629    
2630    Finally, it would probably be a good idea to use the getter-setter
2631    interface to controls rather than ref cells.  For the time being, both
2632    styles are provided by the Controls module, but getter-setter pairs are
2633    better if thread-safety is of any concern because they can be wrapped.
2634    
2635    *****************************************
2636    
2637    One bug fix: The function blockPlacement in three of the MLRISC
2638    backpatch files used to be hard-wired to one of two possibilities at
2639    link time (according to the value of the placementFlag).  But (I
2640    think) it should rather sense the flag every time.
2641    
2642    *****************************************
2643    
2644    Other assorted changes (by other people who did not supply a HISTORY entry):
2645    
2646    1. the cross-module inliner now works much better (Monnier)
2647    2. representation of weights, frequencies, and probabilities in MLRISC
2648       changed in preparation of using those for weighted block placement
2649       (Reppy, George)
2650    
2651    ----------------------------------------------------------------------
2652    Name: Lal George
2653    Date: 2002/03/07 14:44:24 EST 2002
2654    Tag: george-20020307-weighted-block-placement
2655    
2656    Tested the weighted block placement optimization on all architectures
2657    (except the hppa) using AMPL to generate the block and edge frequencies.
2658    Changes were required in the machine properties to correctly
2659    categorize trap instructions. There is an MLRISC flag
2660    "weighted-block-placement" that can be used to enable weighted block
2661    placement, but this will be ineffective without block/edge
2662    frequencies (coming soon).
2663    
2664    
2665    ----------------------------------------------------------------------
2666    Name: Lal George
2667    Date: 2002/03/05 17:24:48 EST
2668    Tag: george-20020305-linkage-cluster
2669    
2670    In order to support the block placement optimization, a new cluster
2671    is generated as the very first cluster (called the linkage cluster).
2672    It contains a single jump to the 'real' entry point for the compilation
2673    unit. Block placement has no effect on the linkage cluster itself, but
2674    all the other clusters  have full freedom in the manner in which they
2675    reorder blocks or functions.
2676    
2677    On the x86 the typical linkage code that is generated is:
2678       ----------------------
2679            .align 2
2680       L0:
2681            addl    $L1-L0, 72(%esp)
2682            jmp     L1
2683    
2684    
2685            .align  2
2686       L1:
2687       ----------------------
2688    
2689    72(%esp) is the memory location for the stdlink register. This
2690    must contain the address of the CPS function being called. In the
2691    above example, it contains the address of  L0; before
2692    calling L1 (the real entry point for the compilation unit), it
2693    must contain the address for L1, and hence
2694    
2695            addl $L1-L0, 72(%esp)
2696    
2697    I have tested this on all architectures except the hppa.The increase
2698    in code size is of course negligible
2699    
2700    ----------------------------------------------------------------------
2701    Name: Allen Leung
2702    Date: 2002/03/03 13:20:00 EST
2703    Tag: leunga-20020303-mlrisc-tools
2704    
2705      Added #[ ... ] expressions to mlrisc tools
2706    
2707    ----------------------------------------------------------------------
2708    Name: Matthias Blume
2709    Date: 2002/02/27 12:29:00 EST
2710    Tag: blume-20020227-cdebug
2711    Description:
2712    
2713    - made types in structure C and C_Debug to be equal
2714    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
2715    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
2716    
2717    ----------------------------------------------------------------------
2718    Name: Matthias Blume
2719    Date: 2002/02/26 12:00:00 EST
2720    Tag: blume-20020226-ffi
2721    Description:
2722    
2723    1. Fixed a minor bug in CM's "noweb" tool:
2724       If numbering is turned off, then truly don't number (i.e., do not
2725       supply the -L option to noweb).  The previous behavior was to supply
2726       -L'' -- which caused noweb to use the "default" line numbering scheme.
2727       Thanks to Chris Richards for pointing this out (and supplying the fix).
2728    
2729    2. Once again, I reworked some aspects of the FFI:
2730    
2731       A. The incomplete/complete type business:
2732    
2733       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
2734         gone!
2735       - ML types representing an incomplete type are now *equal* to
2736         ML types representing their corresponding complete types (just like
2737         in C).  This is still safe because ml-nlffigen will not generate
2738         RTTI for incomplete types, nor will it generate functions that
2739         require access to such RTTI.   But when ML code generated from both
2740         incomplete and complete versions of the C type meet, the ML types
2741         are trivially interoperable.
2742    
2743         NOTE:  These changes restore the full generality of the translation
2744         (which was previously lost when I eliminated functorization)!
2745    
2746       B. Enum types:
2747    
2748       - Structure C now has a type constructor "enum" that is similar to
2749         how the "su" constructor works.  However, "enum" is not a phantom
2750         type because each "T enum" has values (and is isomorphic to
2751         MLRep.Signed.int).
2752       - There are generic access operations for enum objects (using
2753         MLRep.Signed.int).
2754       - ml-nlffigen will generate a structure E_foo for each "enum foo".
2755         * The structure contains the definition of type "mlrep" (the ML-side
2756         representation type of the enum).  Normally, mlrep is the same
2757         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
2758         then mlrep will be defined as a datatype -- thus facilitating
2759         pattern matching on mlrep values.
2760         ("-ec" will be suppressed if there are duplicate values in an
2761          enumeration.)
2762         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
2763         will be generated for each C enum constant xxx.
2764         * Conversion functions m2i and i2m convert between mlrep and
2765         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
2766         * Coversion functions c and ml convert between mlrep and "tag enum".
2767         * Access functions (get/set) fetch and store mlrep values.
2768       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
2769         enumerations are merged into one single enumeration represented by
2770         structure E_'.
2771    
2772    ----------------------------------------------------------------------
2773    Name: Allen Leung
2774    Date: 2002/02/25 04:45:00 EST
2775    Tag: leunga-20020225-cps-spill
2776    
2777    This is a new implementation of the CPS spill phase.
2778    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
2779    In case of problems, replace it with the old file spill.sml
2780    
2781    The current compiler runs into some serious performance problems when
2782    constructing a large record.  This can happen when we try to compile a
2783    structure with many items.  Even a very simple structure like the following
2784    makes the compiler slow down.
2785    
2786        structure Foo = struct
2787           val x_1 = 0w1 : Word32.int
2788           val x_2 = 0w2 : Word32.int
2789           val x_3 = 0w3 : Word32.int
2790           ...
2791           val x_N = 0wN : Word32.int
2792        end
2793    
2794    The following table shows the compile time, from N=1000 to N=4000,
2795    with the old compiler:
2796    
2797    N
2798    1000   CPS 100 spill                           0.04u  0.00s  0.00g
2799           MLRISC ra                               0.06u  0.00s  0.05g
2800              (spills = 0 reloads = 0)
2801           TOTAL                                   0.63u  0.07s  0.21g
2802    
2803    1100   CPS 100 spill                           8.25u  0.32s  0.64g
2804           MLRISC ra                               5.68u  0.59s  3.93g
2805              (spills = 0 reloads = 0)
2806           TOTAL                                   14.71u  0.99s  4.81g
2807    
2808    1500   CPS 100 spill                           58.55u  2.34s  1.74g
2809           MLRISC ra                               5.54u  0.65s  3.91g
2810              (spills = 543 reloads = 1082)
2811           TOTAL                                   65.40u  3.13s  6.00g
2812    
2813    2000   CPS 100 spill                           126.69u  4.84s  3.08g
2814           MLRISC ra                               0.80u  0.10s  0.55g
2815              (spills = 42 reloads = 84)
2816           TOTAL                                   129.42u  5.10s  4.13g
2817    
2818    3000   CPS 100 spill                           675.59u  19.03s  11.64g
2819           MLRISC ra                               2.69u  0.27s  1.38g
2820              (spills = 62 reloads = 124)
2821           TOTAL                                   682.48u  19.61s  13.99g
2822    
2823    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
2824           MLRISC ra                               4.96u  0.27s  2.72g
2825              (spills = 85 reloads = 170)
2826           TOTAL                                   2375.26u  57.21s  48.00g
2827    
2828    As you can see the old cps spill module suffers from some serious
2829    performance problem.  But since I cannot decipher the old code fully,
2830    instead of patching the problems up, I'm reimplementing it
2831    with a different algorithm.  The new code is more modular,
2832    smaller when compiled, and substantially faster
2833    (O(n log n) time and O(n) space).  Timing of the new spill module:
2834    
2835    4000  CPS 100 spill                           0.02u  0.00s  0.00g
2836          MLRISC ra                               0.25u  0.02s  0.15g
2837             (spills=1 reloads=3)
2838          TOTAL                                   7.74u  0.34s  1.62g
2839    
2840    Implementation details:
2841    
2842    As far as I can tell, the purpose of the CPS spill module is to make sure the
2843    number of live variables at any program point (the bandwidth)
2844    does not exceed a certain limit, which is determined by the
2845    size of the spill area.
2846    
2847    When the bandwidth is too large, we decrease the register pressure by
2848    packing live variables into spill records.  How we achieve this is
2849    completely different than what we did in the old code.
2850    
2851    First, there is something about the MLRiscGen code generator
2852    that we should be aware of:
2853    
2854    o MLRiscGen performs code motion!
2855    
2856       In particular, it will move floating point computations and
2857       address computations involving only the heap pointer to
2858       their use sites (if there is only a single use).
2859       What this means is that if we have a CPS record construction
2860       statement
2861    
2862           RECORD(k,vl,w,e)
2863    
2864       we should never count the new record address w as live if w
2865       has only one use (which is often the case).
2866    
2867       We should do something similar to floating point, but the transformation
2868       there is much more complex, so I won't deal with that.
2869    
2870    Secondly, there are now two new cps primops at our disposal:
2871    
2872     1. rawrecord of record_kind option
2873        This pure operator allocates some uninitialized storage from the heap.
2874        There are two forms:
2875    
2876         rawrecord NONE [INT n]  allocates a tagless record of length n
2877         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
2878                                     and initializes the tag.
2879    
2880     2. rawupdate of cty
2881          rawupdate cty (v,i,x)
2882          Assigns to x to the ith component of record v.
2883          The storelist is not updated.
2884    
2885    We use these new primops for both spilling and increment record construction.
2886    
2887     1. Spilling.
2888    
2889        This is implemented with a linear scan algorithm (but generalized
2890        to trees).  The algorithm will create a single spill record at the
2891        beginning of the cps function and use rawupdate to spill to it,
2892        and SELECT or SELp to reload from it.  So both spills and reloads
2893        are fine-grain operations.  In contrast, in the old algorithm
2894        "spills" have to be bundled together in records.
2895    
2896        Ideally, we should sink the spill record construction to where
2897        it is needed.  We can even split the spill record into multiple ones
2898        at the places where they are needed.  But CPS is not a good
2899        representation for global code motion, so I'll keep it simple and
2900        am not attempting this.
2901    
2902     2. Incremental record construction (aka record splitting).
2903    
2904        Long records with many component values which are simulatenously live
2905        (recall that single use record addresses are not considered to
2906         be live) are constructed with rawrecord and rawupdate.
2907        We allocate space on the heap with rawrecord first, then gradually
2908        fill it in with rawupdate.  This is the technique suggested to me
2909        by Matthias.
2910    
2911        Some restrictions on when this is applicable:
2912        1. It is not a VECTOR record.  The code generator currently does not handle
2913           this case. VECTOR record uses double indirection like arrays.
2914        2. All the record component values are defined in the same "basic block"
2915           as the record constructor.  This is to prevent speculative
2916           record construction.
2917    
2918    ----------------------------------------------------------------------
2919    Name: Allen Leung
2920    Date: 2002/02/22 01:02:00 EST
2921    Tag: leunga-20020222-mlrisc-tools
2922    
2923    Minor bug fixes in the parser and rewriter
2924    
2925    ----------------------------------------------------------------------
2926    Name: Allen Leung
2927    Date: 2002/02/21 20:20:00 EST
2928    Tag: leunga-20020221-peephole
2929    
2930    Regenerated the peephole files.  Some contained typos in the specification
2931    and some didn't compile because of pretty printing bugs in the old version
2932    of 'nowhere'.
2933    
2934    ----------------------------------------------------------------------
2935    Name: Allen Leung
2936    Date: 2002/02/19 20:20:00 EST
2937    Tag: leunga-20020219-mlrisc-tools
2938    Description:
2939    
2940       Minor bug fixes to the mlrisc-tools library:
2941    
2942       1.  Fixed up parsing colon suffixed keywords
2943       2.  Added the ability to shut the error messages up
2944       3.  Reimplemented the pretty printer and fixed up/improved
2945           the pretty printing of handle and -> types.
2946       4.  Fixed up generation of literal symbols in the nowhere tool.
2947       5.  Added some SML keywords to to sml.sty
2948    
2949    ----------------------------------------------------------------------
2950    Name: Matthias Blume
2951    Date: 2002/02/19 16:20:00 EST
2952    Tag: blume-20020219-cmffi
2953    Description:
2954    
2955    A wild mix of changes, some minor, some major:
2956    
2957    * All C FFI-related libraries are now anchored under $c:
2958        $/c.cm      --> $c/c.cm
2959        $/c-int.cm  --> $c/internals/c-int.cm
2960        $/memory.cm --> $c/memory/memory.cm
2961    
2962    * "make" tool (in CM) now treats its argument pathname slightly
2963      differently:
2964        1. If the native expansion is an absolute name, then before invoking
2965           the "make" command on it, CM will apply OS.Path.mkRelative
2966           (with relativeTo = OS.FileSys.getDir()) to it.
2967        2. The argument will be passed through to subsequent phases of CM
2968           processing without "going native".  In particular, if the argument
2969           was an anchored path, then "make" will not lose track of that anchor.
2970    
2971    * Compiler backends now "know" their respective C calling conventions
2972      instead of having to be told about it by ml-nlffigen.  This relieves
2973      ml-nlffigen from one of its burdens.
2974    
2975    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
2976    
2977    * Export C_DEBUG and C_Debug from $c/c.cm.
2978    
2979    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
2980      subtyping relationship between incomplete pointers and their complete
2981      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
2982      with the convention of instantiating 'o with ('t, 'c) obj whenever
2983      the pointer target type is complete.  In the incomplete case, 'o
2984      will be instantiated with some "'c iobj" -- a type obtained by
2985      using one of the functors PointerToIncompleteType or PointerToCompleteType.
2986    
2987      Operations that work on both incomplete and complete pointer types are
2988      typed as taking an 'o ptr while operations that require the target to
2989      be known are typed as taking some ('t, 'c) obj ptr.
2990    
2991      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
2992      where void is an eqtype without any values.  This makes it possible
2993      to work on voidptr values using functions meant to operate on light
2994      incomplete pointers.
2995    
2996    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
2997      been vastly simplified.
2998    
2999    ----------------------------------------------------------------------
3000    Name: Matthias Blume
3001    Date: 2002/02/19 10:48:00 EST
3002    Tag: blume-20020219-pqfix
3003    Description:
3004    
3005    Applied Chris Okasaki's bug fix for priority queues.
3006    
3007    ----------------------------------------------------------------------
3008    Name: Matthias Blume
3009    Date: 2002/02/15 17:05:00
3010    Tag: Release_110_39
3011    Description:
3012    
3013    Last-minute retagging is becoming a tradition... :-(
3014    
3015    This is the working release 110.39.
3016    
3017    ----------------------------------------------------------------------
3018    Name: Matthias Blume
3019    Date: 2002/02/15 16:00:00 EST
3020    Tag: Release_110_39-orig
3021    Description:
3022    
3023    Working release 110.39.  New bootfiles.
3024    
3025    (Update: There was a small bug in the installer so it wouldn't work
3026    with all shells.  So I retagged. -Matthias)
3027    
3028    ----------------------------------------------------------------------
3029    Name: Matthias Blume
3030    Date: 2002/02/15 14:17:00 EST
3031    Tag: blume-20020215-showbindings
3032    Description:
3033    
3034    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
3035    the latter can be useful for exploring what bindings are available at
3036    the interactive prompt.  (The first function returns only the list
3037    of symbols that are really bound, the second prints those but also the
3038    ones that CM's autoloading mechanism knows about.)
3039    
3040    ----------------------------------------------------------------------
3041    Name: Matthias Blume
3042    Date: 2002/02/15 12:08:00 EST
3043    Tag: blume-20020215-iptrs
3044    Description:
3045    
3046    Two improvements to ml-nlffigen:
3047    
3048      1. Write files only if they do not exist or if their current contents
3049         do not coincide with what's being written.  (That is, avoid messing
3050         with the time stamps unless absolutely necessary.)
3051    
3052      2. Implement a "repository" mechanism for generated files related
3053         to "incomplete pointer types".   See the README file for details.
3054    
3055    ----------------------------------------------------------------------
3056    Name: Matthias Blume
3057    Date: 2002/02/14 11:50:00 EST
3058    Tag: blume-20020214-quote
3059    Description:
3060    
3061    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
3062    because of the new and improved tag generation scheme.  (Thanks to Allen
3063    Leung for pointing it out.)
3064    
3065    ----------------------------------------------------------------------
3066    Name: Lal George
3067    Date: 2002/02/14 09:55:27 EST 2002
3068    Tag: george-20020214-isabelle-bug
3069    Description:
3070    
3071    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
3072    of Isabelle on the x86.
3073    
3074    From Allen:
3075    -----------
3076     I've found the problem:
3077    
3078         in ra-core.sml, I use the counter "blocked" to keep track of the
3079         true number of elements in the freeze queue.  When the counter goes
3080         to zero, I skip examining the queue.  But I've messed up the
3081         bookkeeping in combine():
3082    
3083             else ();
3084             case !ucol of
3085               PSEUDO => (if !cntv > 0 then
3086                     (if !cntu > 0 then blocked := !blocked - 1 else ();
3087                                        ^^^^^^^^^^^^^^^^^^^^^^^
3088                      moveu := mergeMoveList(!movev, !moveu)
3089                     )
3090                  else ();
3091    
3092         combine() is called to coalesce two nodes u and v.
3093         I think I was thinking that if the move counts of u and v are both
3094         greater than zero then after they are coalesced then one node is
3095         removed from the freeze queue.  Apparently I was thinking that
3096         both u and v are of low degree, but that's clearly not necessarily true.
3097    
3098    
3099    02/12/2002:
3100        Here's the patch.  HOL now compiles.
3101    
3102        I don't know how this impact on performance (compile
3103        time or runtime).  This bug caused the RA (especially on the x86)
3104        to go thru the potential spill phase when there are still nodes on the
3105        freeze queue.
3106    
3107    
3108    
3109    
3110    ----------------------------------------------------------------------
3111    Name: Matthias Blume
3112    Date: 2002/02/13 22:40:00 EST
3113    Tag: blume-20020213-fptr-rtti
3114    Description:
3115    
3116    Fixed a bug in ml-nlffigen that was introduced with one of the previous
3117    updates.
3118    
3119    ----------------------------------------------------------------------
3120    Name: Matthias Blume
3121    Date: 2002/02/13 16:41:00 EST
3122    Tag: blume-20020213-cmlpq
3123    Description:
3124    
3125    Added new priority queue export symbols (which have just been added to
3126    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
3127    not compile and the installer would choke.)
3128    
3129    ----------------------------------------------------------------------
3130    Name: Matthias Blume
3131    Date: 2002/02/13 16:15:00 EST
3132    Tag: blume-20020213-various
3133    Description:
3134    
3135    1. More tweaks to ml-nlffigen:
3136    
3137       - better internal datastructures (resulting in slight speedup)
3138       - "-match" option requires exact match
3139       - "localized" gensym counters (untagged structs/unions nested within
3140         other structs/unions or within typedefs get a fresh counter; their
3141         tag will be prefixed by a concatenation of their parents' tags)
3142       - bug fixes (related to calculation of transitive closure of types
3143         to be included in the output)
3144    
3145    2. Minor Basis updates:
3146    
3147       - added implementations for List.collate and Option.app
3148    
3149    ----------------------------------------------------------------------
3150    Name: Matthias Blume
3151    Date: 2002/02/11 15:55:00 EST
3152    Tag: blume-20020211-gensym
3153    Description:
3154    
3155    Added a "-gensym" option to command line of ml-nlffigen.  This can be
3156    used to specify a "stem" -- a string that is inserted in all "gensym'd"
3157    names (ML structure names that correspond to unnamed C structs, unions,
3158    and enums), so that separate runs of ml-nlffigen do not clash.
3159    
3160    ----------------------------------------------------------------------
3161    Name: Matthias Blume
3162    Date: 2002/02/11 12:05:00 EST
3163    Tag: blume-20020211-gensml
3164    Description:
3165    
3166    A quick fix for a problem with GenSML (in the pgraph-util library):
3167    Make generation of toplevel "local" optional.  (Strictly speaking,
3168    signature definitions within "local" are not legal SML.)
3169    
3170    Other than that: updates to INSTALL and cm/TODO.
3171    
3172    ----------------------------------------------------------------------
3173    Name: Matthias Blume
3174    Date: 2002/02/08 15:00:00 EST
3175    Tag: blume-20020208-uniquepid
3176    Description:
3177    
3178    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
3179    
3180    1. The installer (config/install.sh) has gotten smarter:
3181    
3182         - Configuration options are a bit easier to specify now
3183           (in config/targets).
3184         - Bug in recognizing .tar.bz2 files fixed.
3185         - Installer automatically resolves dependencies between
3186           configuration options (e.g., if you ask for eXene, you will
3187           also get cml -- regardless whether you asked for it or not).
3188         - Installer can run in "quieter mode" by setting the environment
3189           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
3190           "completely silent", though.
3191         - Build HashCons library as part of smlnj-lib.
3192    
3193    2. A new scheme for assigning persistent identifiers to compilation
3194       units (and, by extension, to types etc.) has been put into place.
3195       This fixes a long-standing bug where types and even dynamic values
3196       can get internally confused, thereby compromising type safety
3197       (abstraction) and dynamic correctness.  See
3198    
3199         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
3200    
3201       for an example of how things could go wrong until now.
3202    
3203       The downside of the new scheme is that pids are not quite as
3204       persistent as they used to be: CM will generate a fresh pid
3205       for every compilation unit that it thinks it sees for the first
3206       time.  That means that if you compile starting from a clean, fresh
3207       source tree at two different times, you end up with different
3208       binaries.
3209    
3210       Cutoff recompilation, however, has not been compromised because
3211       CM keeps pid information in special caches between runs.
3212    
3213    ----------------------------------------------------------------------
3214    Name: Lal George
3215    Date: 2002/02/07 15:34:13 EST 2002
3216    Tag: <none>
3217    Description:
3218    
3219    Compilers that generate assembly code may produce  global labels
3220    whose value is resolved at link time. The various peephole optimization
3221    modules did not take this in account.
3222    
3223    TODO. The Labels.addrOf function should really return an option
3224    type so that clients are forced to deal with this issue, rather
3225    than an exception being raised.
3226    
3227    ----------------------------------------------------------------------
3228    Name: Lal George
3229    Date: 2002/02/06 13:55:02 EST
3230    Tag: george-20020206-ra-breakup
3231    Description:
3232    
3233    1. A bug fix from Allen.
3234    
3235        A typo causes extra fstp %st(0)'s to be generated at compensation
3236        edges, which might cause stack underflow traps at runtime.  This
3237        occurs in fft where there are extraneous fstps right before the 'into'
3238        trap instruction (in this case they are harmless since none of the
3239        integers overflow.)
3240    
3241    2. Pulled out various utility modules that were embedded in the modules
3242       of the register allocator. I need these modules for other purposes, but
3243       they are not complete enough to put into a library (just yet).
3244    ----------------------------------------------------------------------
3245    Name: Matthias Blume
3246    Date: 2002/01/31 16:05:00 EST
3247    Tag: blume-20020131-sparc-ccalls
3248    Description:
3249    
3250    1. C-calls on Sparc needlessly allocated a huge chunk (96 bytes)
3251       of extra stack space by mistake.  Fixed.
3252    
3253    2. Bug in logic of handling of command-line options in ml-nlffigen fixed.
3254    
3255    ----------------------------------------------------------------------
3256    Name: Allen Leung
3257    Date: 2002/01/30
3258    Tag: leunga-20020130-nowhere-bug-fix
3259    Description:
3260    
3261       MLRISC bug fixes:
3262       1. Fixed a bindings computation bug in the 'nowhere' program generator tool.
3263       2. MachineInt.fromString was negating its value.
3264    
3265    ----------------------------------------------------------------------
3266    Name: Matthias Blume
3267    Date: 2002/01/29
3268    Tag: blume-20020129-INSTALL
3269    Description:
3270    
3271    - Added somewhat detailed installation instructions (file INSTALL).
3272    - Fixed curl-detection bug in config/install.sh.
3273    - It is now possible to select the URL getter using the URLGETTER
3274      environment variable:
3275    
3276          not set / "unknown"      --> automatic detection (script tries wget,
3277                                       curl, and lynx)
3278          "wget" / "curl" / "lynx" --> use the specified program (script "knows"
3279                                       how to properly invoke them)
3280          other                    --> use $URLGETTER directly, it must take
3281                                       precisely two command-line arguments
3282                                       (source URL and destination file name)
3283    
3284    ----------------------------------------------------------------------
3285    Name: Matthias Blume
3286    Date: 2002/01/28
3287    Tag: blume-20020128-sparc-ccalls
3288    Description:
3289    
3290    - Fixed problem with calculation of "used" registers in sparc-c-calls.
3291    - Make use of the allocParam argument in sparc-c-calls.
3292    
3293    ----------------------------------------------------------------------
3294    Name: Matthias Blume
3295    Date: 2002/01/28
3296    Tag: blume-20020128-allocParam
3297    Description:
3298    
3299    John Reppy:  Changes c-calls API to accept client-callback for
3300    allocating extra stack space.
3301    me: Corresponding changes to mlriscGen (using a dummy argument that
3302        does not change the current behavior).
3303    
3304    ----------------------------------------------------------------------
3305    Name: Matthias Blume
3306    Date: 2002/01/28 12:00:00
3307    Tag: Release_110_38
3308    Description:
3309    
3310    This time for real!!!
3311    
3312    ----------------------------------------------------------------------
3313    Name: Matthias Blume
3314    Date: 2002/01/28 10:56:00 EST
3315    Tag: blume-20020128-retraction
3316    Description:
3317    
3318    0. Retracted earlier 110.38.  (The Release_110_38 tag has been replaced
3319       with blume-Release_110_38-retracted.)
3320    
3321    1. Fixed a problem with incorrect rounding modes in real64.sml.
3322       (Thanks to Andrew Mccreight <andrew.mccreight@yale.edu>.)
3323    
3324    2. A bug in ml-nlffigen related to the handling of unnamed structs, unions,
3325       and enums fixed.  The naming of corresponding ML identifiers should
3326       now be consistent again.
3327    
3328    ----------------------------------------------------------------------
3329    Name: Allen Leung
3330    Date: 2002/01/27
3331    Tag: leunga-20020127-nowhere
3332    Description:
3333    
3334       Added a target called nowhere in the configuration scripts.
3335       Enabling this will build the MLRISC 'nowhere' tool (for translating
3336       programs with where-clauses into legal SML code) during installation.
3337    
3338    ----------------------------------------------------------------------
3339    Name: Matthias Blume
3340    Date: 2002/01/25 21:27:00 EST
3341    Tag: blume-Release_110_38-retracted
3342    Description:
3343    
3344    Call it a (working) release!  Version is 110.38. Bootfiles are ready.
3345    
3346    README will be added later.
3347    
3348    !!! NOTE:  Re-tagged as blume-Release_110_38-retracted. Original tag
3349    (Release_110_38) removed.  Reason: Last-minute bug fixes.
3350    
3351    ----------------------------------------------------------------------
3352    Name: Matthias Blume
3353    Date: 2002/01/25
3354    Tag: blume-20020125-ffi
3355    Description:
3356    
3357    A large number of tweaks and improvements to ml-nlffi-lib and
3358    ml-nlffigen:
3359    
3360       - ML represenation types have been streamlined
3361       - getter and setter functions work with concrete values, not abstract
3362         ones where possible
3363       - ml-nlffigen command line more flexible (see README file there)
3364       - some bugs have been fixed (hopefully)
3365    
3366    ----------------------------------------------------------------------
3367    Name: Lal George
3368    Date: 2002/01/24
3369    Tag: george-20020124-risc-ra-interface
3370    Description:
3371    
3372       There is a dramatic simplification in the interface to the
3373       register allocator for RISC architectures as a result of making
3374       parallel copy instructions explicit.
3375    
3376    ----------------------------------------------------------------------
3377    Name: Matthias Blume
3378    Date: 2002/01/22
3379    Tag: blume-20020122-x86-ccalls
3380    Description:
3381    
3382    Bug fix for c-calls on x86 (having to do with how char- and
3383    short-arguments are being handled).
3384    
3385    ----------------------------------------------------------------------
3386    Name: Matthias Blume
3387    Date: 2002/01/21
3388    Tag: blume-20020121-ff
3389    Description:
3390    
3391    Another day of fiddling with the FFI...
3392    
3393    1. Bug fix/workaround:  CKIT does not complain about negative array
3394       dimensions, so ml-nlffigen has to guard itself against this possibility.
3395       (Otherwise a negative dimension would send it into an infinite loop.)
3396    
3397    2. Some of the abstract types (light objects, light pointers, most "base"
3398       types) in structure C are now eqtypes.
3399    
3400    3. Added constructors and test functions for NULL function pointers.
3401    
3402    ----------------------------------------------------------------------
3403    Name: Matthias Blume
3404    Date: 2002/01/18
3405    Tag: blume-20020118-ready-for-new-release
3406    Description:
3407    
3408    Made config/srcarchiveurl point to a new place.  (Will provide boot
3409    files shortly.)
3410    
3411    Maybe we christen this to be 110.38?
3412    
3413    ----------------------------------------------------------------------
3414    Name: Matthias Blume
3415    Date: 2002/01/18
3416    Tag: blume-20020118-more-ffifiddle
3417    Description:
3418    
3419    Today's FFI fiddling:
3420    
3421      - Provided a structure CGetSet with "convenient" versions of C.Get.* and
3422        C.Set.* that use concrete (MLRep.*) arguments and results instead
3423        of abstract ones.
3424    
3425      - Provided word-style bit operations etc. for "int" representation
3426        types in MLRep.S<Foo>Bitops where <Foo> ranges over Char, Int, Short,
3427        and Long.
3428    
3429    ----------------------------------------------------------------------
3430    Name: Matthias Blume
3431    Date: 2002/01/18
3432    Tag: blume-20020118-use-x86-fp
3433    Description:
3434    
3435    Now that x86-fast-fp seems to be working, I turned it back on again
3436    by default.  (Seems to work fine now, even with the FFI.)
3437    
3438    Other than that, I added some documentation about the FFI to
3439    src/ml-nlffigen/README and updated the FFI test examples in
3440    src/ml-nlffi-lib/Tests/*.
3441    
3442    ----------------------------------------------------------------------
3443    Name: Allen Leung
3444    Date: 2002/01/17
3445    Tag: leunga-20020117-x86-fast-fp-call
3446    Description:
3447    
3448       1. Fixed a problem with handling return fp values when x86's fast fp
3449          mode is turned on.
3450    
3451       2. Minor pretty printing fix for cellset.  Print %st(0) as %st(0) instead
3452          of %f32.
3453    
3454       3. Added a constructor INT32lit to the ast of MLRISC tools.
3455    
3456    ----------------------------------------------------------------------
3457    Name: Matthias Blume
3458    Date: 2002/01/16
3459    Tag: blume-20020116-ffifiddle
3460    Description:
3461    
3462    More fiddling with the FFI interface:
3463    
3464     - Make constness 'c instead of rw wherever possible.  This eliminates
3465       the need for certain explicit coercions.  (However, due to ML's
3466       value polymorphism, there will still be many cases where explicit
3467       coercions are necessary.  Phantom types are not the whole answer
3468       to modeling a subtyping relationship in ML.)
3469    
3470     - ro/rw coersions for pointers added.  (Avoids the detour through */&.)
3471    
3472     - "printf" test example added to src/ml-nlffi-lib/Tests.  (Demonstrates
3473       clumsy workaround for varargs problem.)
3474    
3475    ----------------------------------------------------------------------
3476    Name: Lal George
3477    Date: 2002/01/15
3478    Tag: <none>
3479    Description:
3480    
3481    1. Since COPY instructions are no longer native to the architecture,
3482       a generic functor can be used to implement the expandCopies function.
3483    
3484    2. Allowed EXPORT and IMPORT pseudo-op declarations to appear inside a
3485       TEXT segment.
3486    
3487    ----------------------------------------------------------------------
3488    Name: Matthias Blume
3489    Date: 2002/01/15
3490    Tag: blume-20020115-ffiupdates
3491    Description:
3492    
3493    1. Fix for bug resulting in single-precision float values being returned
3494       incorrectly from FFI calls.
3495    
3496    2. Small modifications to C FFI API:
3497    
3498        - memory-allocation routines return straight objects (no options)
3499          and raise an exception in out-of-memory situations
3500        - unsafe extensions to cast between function pointers and pointers
3501          from/to ints
3502        - added structure C_Debug as an alternative to structure C where
3503          pointer-dereferencing (|*| and |*!) always check for null-pointers
3504        - added open_lib' to DynLinkage;  open_lib' works like open_lib
3505          but also takes a (possibly empty) list of existing library handles
3506          that the current library depends on
3507    
3508    ----------------------------------------------------------------------
3509    Name: Matthias Blume
3510    Date: 2002/01/10
3511    Tag: blume-20020110-newffigen
3512    Description:
3513    
3514    1. Updates to portable graph code.
3515    
3516    2. Major update to ml-nlffigen and ml-nlffi-lib.  Things are much
3517       more scalable now so that even huge interfaces such as the one
3518       for GTK compile in finite time and space. :-)
3519       See src/ml-nlffigen/README for details on what's new.
3520    
3521    ----------------------------------------------------------------------
3522    Name: Lal George
3523    Date: 2001/01/09 14:31:35 EST 2002
3524    Tag: george-20011206-rm-native-copy
3525    Description:
3526    
3527            Removed the native COPY and FCOPY instructions
3528            from all the architectures and replaced it with the
3529            explicit COPY instruction from the previous commit.
3530    
3531            It is now possible to simplify many of the optimizations
3532            modules that manipulate copies. This has not been
3533            done in this change.
3534    
3535    ----------------------------------------------------------------------
3536    Name: Lal George
3537    Date: 2001/12/06 16:50:13 EST 2001
3538    Tag: george-20011206-mlrisc-instruction
3539    Description:
3540    
3541    Changed the representation of instructions from being fully abstract
3542    to being partially concrete. That is to say:
3543    
3544      from
3545            type instruction
3546    
3547      to
3548            type instr                              (* machine instruction *)
3549    
3550            datatype instruction =
3551                LIVE of {regs: C.cellset, spilled: C.cellset}
3552              | KILL of {regs: C.cellset, spilled: C.cellset}
3553              | COPYXXX of {k: CB.cellkind, dst: CB.cell list, src: CB.cell list}
3554              | ANNOTATION of {i: instruction, a: Annotations.annotation}
3555              | INSTR of instr
3556    
3557    This makes the handling of certain special instructions that appear on
3558    all architectures easier and uniform.
3559    
3560    LIVE and KILL say that a list of registers are live or killed at the
3561    program point where they appear. No spill code is generated when an
3562    element of the 'regs' field is spilled, but the register is moved to
3563    the 'spilled' (which is present, more for debugging than anything else).
3564    
3565    LIVE replaces the (now deprecated) DEFFREG instruction on the alpha.
3566    We used to generate:
3567    
3568            DEFFREG f1
3569            f1 := f2 + f3
3570            trapb
3571    
3572    but now generate:
3573    
3574            f1 := f2 + f3
3575            trapb
3576            LIVE {regs=[f1,f2,f3], spilled=[]}
3577    
3578    Furthermore, the DEFFREG (hack) required that all floating point instruction
3579    use all registers mentioned in the instruction. Therefore f1 := f2 + f3,
3580    defines f1 and uses [f1,f2,f3]! This hack is no longer required resulting
3581    in a cleaner alpha implementation. (Hopefully, intel will not get rid of
3582    this architecture).
3583    
3584    COPYXXX is intended to replace the parallel COPY and FCOPY  available on
3585    all the architectures. This will result in further simplification of the
3586    register allocator that must be aware of them for coalescing purposes, and
3587    will also simplify certain aspects of the machine description that provides
3588    callbacks related to parallel copies.
3589    
3590    ANNOTATION should be obvious, and now INSTR represents the honest to God
3591    machine instruction set!
3592    
3593    The <arch>/instructions/<arch>Instr.sml files define certain utility
3594    functions for making porting easier -- essentially converting upper case
3595    to lower case. All machine instructions (of type instr) are in upper case,
3596    and the lower case form generates an MLRISC instruction. For example on
3597    the alpha we have:
3598    
3599      datatype instr =
3600         LDA of {r:cell, b:cell, d:operand}
3601       | ...
3602    
3603      val lda : {r:cell, b:cell, d:operand} -> instruction
3604        ...
3605    
3606    where lda is just (INSTR o LDA), etc.
3607    
3608    ----------------------------------------------------------------------
3609    Name: Matthias Blume
3610    Date: 2001/11/22 21:40:00 EST
3611    Tag: Release_110_37
3612    Description:
3613    
3614    Release 110.37.  This time for real.
3615    
3616    ----------------------------------------------------------------------
3617    Name: Matthias Blume
3618    Date: 2001/11/21 16:35:00 EST
3619    Tag: blume-20011121-foot-in-mouth
3620    Description:
3621    
3622    Removed the "Release_110_37" tag because of a serious bug.
3623    This will be re-tagged once the bug is fixed.
3624    
3625    ----------------------------------------------------------------------
3626    Name: Matthias Blume
3627    Date: 2001/11/21 16:14:00 EST
3628    Tag: blume-20011121-forgottenfile
3629    Description:
3630    
3631    Forgot to add a file.  (Just a .tex-file -- part of
3632    the CM manual source.)
3633    
3634    ----------------------------------------------------------------------
3635    Name: Matthias Blume
3636    Date: 2001/11/21 16:10:00 EST
3637    Tag: blume-20011121-invalid_110_37
3638    Description:
3639    
3640    Note: I removed the original tag "Release_110_37" from this commit
3641          because we found a serious bug in all non-x86 backends.
3642          - Matthias
3643    
3644    1. Modifications to the SML/NJ code generator and to the runtime system
3645       so that code object name strings are directly inserted into code
3646       objects at code generation time.  The only business the runtime system
3647       has with this is now to read the name strings on occasions.
3648       (The encoding of the name string has also changed somewhat.)
3649    
3650    2. CM now implements a simple "set calculus" for specifying export lists.
3651       In particular, it is now possible to refer to the export lists of
3652       other libraries/groups/sources and form unions as well as differences.
3653       See the latest CM manual for details.
3654    
3655    3. An separate notion of "proxy" libraries has again be eliminated from
3656       CM's model.  (Proxy libraries are now simply a special case of using
3657       the export list calculus.)
3658    
3659    4. Some of the existing libraries now take advantage of the new set
3660       calculus.
3661       (Notice that not all libraries have been converted because some
3662       of the existing .cm-files are supposed to be backward compatible
3663       with 110.0.x.)
3664    
3665    5. Some cleanup in stand-alone programs.  (Don't use "exnMessage" -- use
3666       "General.exnMessage"!  The former relies on a certain hook to be
3667       initialized, and that often does not happen in the stand-alone case.)
3668    
3669    ----------------------------------------------------------------------
3670    Name: Lal George
3671    Date: 2001/11/21  13:56:18 EST
3672    Tag: george-2001121-pseudo-ops
3673    Description:
3674    
3675      Implemented a complete redesign of MLRISC pseudo-ops. Now there
3676      ought to never be any question of incompatabilities with
3677      pseudo-op syntax expected by host assemblers.
3678    
3679      For now, only modules supporting GAS syntax are implemented
3680      but more should follow, such as MASM, and vendor assembler
3681      syntax, e.g. IBM as, Sun as, etc.
3682    
3683    ----------------------------------------------------------------------
3684    Name: Matthias Blume
3685    Date: 2001/11/14 11:52:00 EST
3686    Tag: blume-20011114-srcname
3687    Description:
3688    
3689    1. Routed the name of the current source file to mlriscgen where it
3690       should be directly emitted into the code object.  (This last part
3691       is yet to be done.)
3692    
3693    2. Some cleanup of the pgraph code to make it match the proposal that
3694       I put out the other day.  (The proposal notwithstanding, things are
3695       still in flux here.)
3696    
3697    ----------------------------------------------------------------------
3698    Name: Lal George
3699    Date: 2001/11/14 09:44:04 EST
3700    Tag:
3701    Description:
3702    
3703      Fix for a backpatching bug reported by Allen.
3704    
3705      Because the boundary between short and long span-dependent
3706      instructions is +/- 128, there are an astounding number of
3707      span-dependent instructions whose size is over estimated.
3708    
3709      Allen came up with the idea of letting the size of span
3710      dependent instructions be non-monotonic, for a maxIter
3711      number of times, after which the size must be monotonically
3712      increasing.
3713    
3714      This table shows the number of span-dependent instructions
3715      whose size was over-estimated as a function of maxIter, for the
3716      file Parse/parse/ml.grm.sml:
3717    
3718         maxIter            # of instructions:
3719            10                      687
3720            20                      438
3721            30                      198
3722            40                        0
3723    
3724      In compiling the compiler, there is no significant difference in
3725      compilation speed between maxIter=10 and maxIter=40. Actually,
3726      my measurements showed that maxIter=40 was a tad faster than
3727      maxIter=10! Also 96% of the  files in the compiler reach a fix
3728      point within 13 iterations, so fixing maxIter at 40, while high,
3729      is okay.
3730    
3731    ----------------------------------------------------------------------
3732    Name: Matthias Blume
3733    Date: 2001/10/31 15:25:00 EST
3734    Tag: blume-20011031-pgraph
3735    Description:
3736    
3737    CKIT:
3738    * Changed the "Function" constructor of type Ast.ctype to carry optional
3739      argument identifiers.
3740    * Changed the return type of TypeUtil.getFunction accordingly.
3741    * Type equality ignores the argument names.
3742    * TypeUtil.composite tries to preserve argument names but gives up quickly
3743      if there is a mismatch.
3744    
3745    installation script:
3746    * attempts to use "curl" if available (unless "wget" is available as well)
3747    
3748    CM:
3749    * has an experimental implementation of "portable graphs" which I will
3750      soon propose as an implementation-independent library format
3751    * there are also new libraries $/pgraph.cm and $/pgraph-util.cm
3752    
3753    NLFFI-LIB:
3754    * some cleanup (all cosmetic)
3755    
3756    NLFFIGEN:
3757    * temporarily disabled the mechanism that suppresses ML output for
3758      C definitions whose identifiers start with an underscore character
3759    * generate val bindings for enum constants
3760    * user can request that only one style (light or heavy) is being used;
3761      default is to use both (command-line arguments: -heavy and -light)
3762    * fixed bug in handling of function types involving incomplete pointers
3763    * generate ML entry points that take record arguments (i.e., using
3764      named arguments) for C functions that have a prototype with named
3765      arguments
3766      (see changes to CKIT)
3767    
3768    ----------------------------------------------------------------------
3769  Name: Allen Leung  Name: Allen Leung
3770  Date: 2001/10/27 20:34:00 EDT  Date: 2001/10/27 20:34:00 EDT
3771  Tag: leunga-20011027-x86-fast-fp-call  Tag: leunga-20011027-x86-fast-fp-call
# Line 109  Line 3862 
3862    
3863  ----------------------------------------------------------------------  ----------------------------------------------------------------------
3864  Name: Matthias Blume  Name: Matthias Blume
 >>>>>>> 1.169  
3865  Date: 2001/09/18 15:35:00 EDT  Date: 2001/09/18 15:35:00 EDT
3866  Tag: blume-20010918-readme11036  Tag: blume-20010918-readme11036
3867  Description:  Description:

Legend:
Removed from v.972  
changed lines
  Added in v.1669

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