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/branches/primop-branch-3/NOTES/HISTORY
ViewVC logotype

Diff of /sml/branches/primop-branch-3/NOTES/HISTORY

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 643, Fri May 12 08:29:15 2000 UTC revision 1925, Fri May 12 20:16:43 2006 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 (http://www.tti-c.org/blume/)
17    Date: 2006/05/12 15:15:00 CST
18    Tag: blume-20060512-makedepend
19    Description:
20    
21    Fixed long-standing bug with ml-makedepend where it would output a
22    spurious dependency to a non-existing file.  (This is a simple fix.
23    It might need further looking into.)
24    
25    ----------------------------------------------------------------------
26    Name: Matthias Blume (http://www.tti-c.org/blume/)
27    Date: 2006/04/20 10:30:00 CDT
28    Tag: blume-20060420-heap2exec
29    Description:
30    
31    Committed patches received from Johannes 5 Joemann (joemann@befree.free.de)
32    that enable heap2exec under Linux and FreeBSD.
33    
34    ----------------------------------------------------------------------
35    Name: John Reppy
36    Date: 2006/04/14
37    Tag:
38    Description:
39    
40    MLRISC changes:
41      1) renamed GAS_PSEUDO_OPS to AS_PSEUDO_OPS and put it in its own file.
42      2) added support for NOTB and XORB operators in pseudo-op expressions
43      3) added DarwinPseudoOp functor that supports Darwin's assembler syntax.
44      4) added support for 64-bit integer literals
45    
46    ----------------------------------------------------------------------
47    Name: Matthias Blume (http://www.tti-c.org/blume/)
48    Date: yyyy/mm/dd
49    Tag: Release_110_58
50    Description:
51    
52    New working version (110.58).  NEW BOOTFILES!
53    
54    ----------------------------------------------------------------------
55    Name: Matthias Blume (http://www.tti-c.org/blume/)
56    Date: 2006/03/01 22:15:00 CST
57    Tag: blume-20060301-lexgen
58    Description:
59    
60    Incorporated several bugfixes to lexgen.  Compiler now compiles to
61    fixpoint when using lexgen instead of ml-lex.
62    
63    ----------------------------------------------------------------------
64    Name: Matthias Blume (http://www.tti-c.org/blume/)
65    Date: 2006/02/28 22:45:00 CST
66    Tag: blume-20060228-lexgen
67    Description:
68    
69    Removed ml-flex and added lexgen instead, using Aaron Turon's newly
70    provided tarball.  The generated lexers still have problems.
71    
72    ----------------------------------------------------------------------
73    Name: Matthias Blume (http://www.tti-c.org/blume/)
74    Date: 2006/02/26 23:30:00 CST
75    Tag: blume-20060226-ml-flex
76    Description:
77    
78    Removed ml-flex' dependency on regexp-lib.cm.  Turned installation
79    of ml-flex on by default.
80    
81    ----------------------------------------------------------------------
82    Name: Matthias Blume (http://www.tti-c.org/blume/)
83    Date: 2006/02/24 16:30:00 CST
84    Tag: blume-20060224-ml-flex
85    Description:
86    
87    Added ml-flex sources.  Partially integrated, but should not be
88    turned on yet!  (Read: leave it commented-out in config/targets!)
89    
90    ----------------------------------------------------------------------
91    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
92    Date: 2006/02/23
93    Tag: jhr-20060223-x86-64
94    Description:
95    Changes to support building on x86-64 systems (using the 32-bit mode).
96    Also cleaned up signal handling on Linux.  Support for pre-2.2 Linux
97    kernels dropped.
98    
99    ----------------------------------------------------------------------
100    Name: Allen Leung (leunga (at) reservoir (dot) com)
101    Date: 2006/02/22 15:07:00 EST
102    Tag: leunga-20060222-mdl-gen
103    Description:
104    
105      CM has changed.  Updated the script for rebuilding the MLRISC generated
106    files.
107    
108    ----------------------------------------------------------------------
109    Name: Matthias Blume (http://www.tti-c.org/blume/)
110    Date: 2006/02/14 17:30:00 CST
111    Tag: blume-20060214-intel-mac
112    Description:
113    
114    Hooked code for Darwin-specific Intel ABI into main compiler.
115    (This is a temporary solution which relies on the fact that the
116    compiler itself does not use NLFFI.  Eventually we need to divorce
117    intel mac from generic x86 unix code and make separate sets of binaries.)
118    
119    ----------------------------------------------------------------------
120    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
121    Date: 2006/02/13
122    Tag: jhr-20060213-intel-mac
123    Description:
124    Changed MLRISC x86 CCalls for partial support of MacOS X ABI.
125    
126    ----------------------------------------------------------------------
127    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
128    Date: 2006/02/06
129    Tag: jhr-20060206-intel-mac
130    Description:
131    Changes to support MacOS X on Intel hardware.  The C-calls support in
132    MLRISC must also be updated to support the MacOS X ABI.
133    
134    ----------------------------------------------------------------------
135    Name: Matthias Blume (http://www.tti-c.org/blume/)
136    Date: 2005/12/16 00:15:00 CST
137    Tag: blume-20051216-cm
138    Description:
139    
140    Improved error reporting and handling in CM.
141    
142    ----------------------------------------------------------------------
143    Name: Matthias Blume (http://www.tti-c.org/blume/)
144    Date: 2005/11/21 14:00:00 CST
145    Tag: blume-20051121-html-lib
146    Description:
147    
148    Enabling $/html-lib.cm in config/preloads did not work.  This is fixed
149    now.  (Since the anchor mapping for html-lib.cm is not yet in effect
150    at the time when config/preloads is processed, the library has to
151    be referred to by another name.  In 110.57 this would be
152    $SMLNJ-LIB/HTML/html-lib.cm.  I arranged for $smlnj/smlnj-lib/html-lib.cm
153    to be valid as well.)
154    
155    Thanks to Todd Wilson (Fresno) for alerting me to this issue.
156    
157    ----------------------------------------------------------------------
158    Name: Matthias Blume (http://www.tti-c.org/blume/)
159    Date: 2005/11/19 13:45:00 CST
160    Tag: Release_110_57
161    Description:
162    
163    Moved release tag for 110.57.
164    
165    ----------------------------------------------------------------------
166    Name: Matthias Blume (http://www.tti-c.org/blume/)
167    Date: 2005/11/19 23:35:00 CST
168    Tag: blume-20051119-fish
169    Description:
170    
171    Fixed a problem in config/install.sh where it tries to "fish" the
172    name of the CM metadata directory from the wrong place (because
173    the physical location of basis.cm has changed).  Also, corrected the
174    path anchor for $/html-lib.cm.  (Thanks to M. Fluet for pointing out
175    these problems.)
176    
177    ----------------------------------------------------------------------
178    Name: Matthias Blume (http://www.tti-c.org/blume/)
179    Date: 2005/11/16 14:30:00 CST
180    Tag: blume-orig-110_57
181    Description:
182    
183    New working version (110.57).  NEW BOOTFILES!
184    
185    UPDATE: The release tag has been moved to include a fix for a brown-paper-
186            bag glitch (see above).
187    
188    ----------------------------------------------------------------------
189    Name: Matthias Blume (http://www.tti-c.org/blume/)
190    Date: 2005/11/16 10:25:00 CST
191    Tag: blume-22051116-btrace
192    Description:
193    
194    Fixed problem with bogus exception message when using back-trace
195    facility.
196    
197    ----------------------------------------------------------------------
198    Name: Matthias Blume (http://www.tti-c.org/blume/)
199    Date: 2005/11/15 22:35:00 CST
200    Tag: blume-22051115-array2copy
201    Description:
202    
203    Added simple implementation of Array2.copy.  (Warning: mostly untested.)
204    
205    ----------------------------------------------------------------------
206    Name: Matthias Blume (http://www.tti-c.org/blume/)
207    Date: 2005/11/15 21:35:00 CST
208    Tag: blume-22051115-versiontool
209    Description:
210    
211    Reversed change to src/system/smlnj/internals/versiontool.cm.  This
212    file gets loaded as a tool -- by the equivalent of CM.make during the
213    run of CMB.make.  Thus, CMB's path configuration is meaningless for
214    it.  Instead, it has the status of "user code", so it should use
215    $/basis.cm to refer to the Basis library.  (At least that's true for
216    the purpose of bootstrapping the previous change.  In the future it
217    might make sense to have versiontool.cm refer to
218    $smlnj/basis/basis.cm, i.e., the version of the Basis that the
219    compiler itself uses.)
220    
221    Also patched src/system/testml to have it activate those extra anchor
222    bindings in config/extrapathconfig.
223    
224    ----------------------------------------------------------------------
225    Name: Matthias Blume (http://www.tti-c.org/blume/)
226    Date: 2005/11/15 16:35:00 CST
227    Tag: blume-22051115-libraries
228    Description:
229    
230    This change affects the way the following libraries are tied into
231    the system:
232    
233       $/basis.cm
234       $/smlnj-lib.cm
235       $/pp-lib.cm
236       $/controls-lib.cm
237       $/html-lib.cm
238       $/ml-yacc-lib.cm
239    
240    These libraries are now internally (as seen from the source code
241    of the implementation itself) known by the following names:
242    
243       $smlnj/basis/basis.cm
244       $smlnj/smlnj-lib/smlnj-lib.cm
245       $smlnj/smlnj-lib/pp-lib.cm
246       $smlnj/smlnj-lib/controls-lib.cm
247       $smlnj/smlnj-lib/html-lib.cm
248       $smlnj/ml-yacc/ml-yacc-lib.cm
249    
250    This makes it possible to work with code that requires different
251    versions of these libraries, and which refers to these libraries using
252    their "default" names (i.e., the first set of names above).  In other
253    words, one can un-define or re-define those default names without
254    compromising the proper functioning of the compiler itself.
255    
256    A similar procedure had already been performed for several of the
257    MLRISC libraries that are linked into the compiler.  I did some
258    cleanup on this code.
259    
260    A new file in the config directory (named extrapathconfig) is
261    responsible for setting up path anchors that the compiler itself does
262    not need, but that are typically required by user code.
263    
264    ----------------------------------------------------------------------
265    Name: Matthias Blume
266    Date: 2005/11/07 17:05:00 CST
267    Tag: blume-20051107-slice
268    Description:
269    
270    Fixed erroneous out-of-bounds test in the "update" function of
271    various *ArraySlice modules.  (Thanks to Vesa A. Norrman for pointing
272    out the problem.)
273    
274    Pushed some Basis changes through ML-Lex, CML, and eXene.
275    
276    ----------------------------------------------------------------------
277    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
278    Date: 2005/11/07
279    Description:
280    Fixed a Basis incompatibility: The depreciated function Substring.all
281    was removed (use Substring.full instead).
282    
283    ----------------------------------------------------------------------
284    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
285    Date: 2005/11/05
286    Description:
287    Tweaked interval set API in SML/NJ library; see the CHANGES file for details.
288    
289    ----------------------------------------------------------------------
290    Name: Matthias Blume (http://www.tti-c.org/blume/)
291    Date: 2005/11/02 16:25:00 CST
292    Tag: blume-20051102-hex
293    Description:
294    
295    * Runtime system bootstrap code now accepts hex digits in BOOTLIST
296      in either upper case or lower case format.
297    * Pushed changes to names of Pack<N>{Big,Little} structures through
298      CML and eXene.
299    
300    ----------------------------------------------------------------------
301    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
302    Date: 2005/11/02
303    Description:
304    Fixed a Basis incompatibility: Pack<N>{Big,Little} structures should
305    be called PackWord<N>{Big,Little}.
306    
307    ----------------------------------------------------------------------
308    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
309    Date: 2005/10/28
310    Description:
311    Fixed a minor Basis incompatibility: hex digits should be upper case.
312    
313    ----------------------------------------------------------------------
314    Name: Matthias Blume (blume (at) tti - c (dot) org)
315    Date: 2005/10/25 16:00:00 CDT
316    Tag: Release_110_56
317    Description:
318    
319    New working version (110.56).  NEW BOOTFILES!
320    
321    ----------------------------------------------------------------------
322    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
323    Date: 2005/10/25
324    Description:
325    Added interval sets to utility library (signatures INTERVAL_DOMAIN
326    and INTERVAL_SET, and functor IntervalSetFn).
327    
328    ----------------------------------------------------------------------
329    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
330    Date: 2005/10/14
331    Description:
332    Add Zhong Shao's fix for datatype equality functions.
333    
334    ----------------------------------------------------------------------
335    Name: Allen Leung (leunga (at) reservoir (dot) com)
336    Date: 2005/10/14 12:30:00 EST
337    Tag: leunga-20051014-x86-spill
338    Description:
339    
340    Bug fix a bug found by Carl Hauser.
341    There was a typo in the reload code for FCMP in x86SpillInstr.sml.
342    
343    ----------------------------------------------------------------------
344    Name: Allen Leung (leunga (at) reservoir (dot) com)
345    Date: 2005/10/14 11:15:00 EST
346    Tag: leunga-20051014-x86-asm
347    Description:
348    
349    Removed some debugging code in file x86Asm.sml.
350    The function emit_operand was printing out debugging output.
351    
352    ----------------------------------------------------------------------
353    Name: John Reppy
354    Date: 2005/07/27
355    Tag:
356    Description:
357    
358    Fixed ml-lex to recognize "\r" as representing carriage return.
359    
360    ----------------------------------------------------------------------
361    Name: John Reppy
362    Date: 2005/07/27
363    Tag:
364    Description:
365    
366    Fixed ml-yacc to work on files with non-native end-of-line
367    encodings (e.g., Windows text file on a Unix system).
368    
369    ----------------------------------------------------------------------
370    Name: John Reppy
371    Date: 2005/07/20
372    Tag:
373    Description:
374    
375    Added changes from Dominic Evans (oldmanuk (at) gmail (dot) com)
376    to support HPUX 11.
377    
378    ----------------------------------------------------------------------
379    Name: John Reppy
380    Date: 2005/07/06
381    Tag:
382    Description:
383    
384    Changes to the SML/NJ library.  See smlnj-lib/CHANGES for details.
385    
386    ----------------------------------------------------------------------
387    Name: Matthias Blume (blume (at) tti - c (dot) org)
388    Date: 2005/07/06 00:45:00 CDT
389    Tag: blume-20050706-slice-copy
390    Description:
391    
392    Fixed reversed logic for deciding whether to "copy up" or "copy down"
393    in *-array-slice.sml.
394    
395    ----------------------------------------------------------------------
396    Name: Allen Leung (leunga (at) reservoir (dot) com)
397    Date: 2005/05/31 17:00:00 EST
398    Tag: leunga-20050531-cygwin-fault-2
399    Description:
400    
401    A typo in the cygwin code fixed.
402    
403    ----------------------------------------------------------------------
404    Name: Allen Leung (leunga (at) reservoir (dot) com)
405    Date: 2005/05/31 16:47:00 EST
406    Tag: leunga-20050531-cygwin-fault
407    Description:
408    
409    Updated Cygwin's fault/signal handling to match the Windows version.
410    Updated the export list.
411    
412    ----------------------------------------------------------------------
413    Name: Matthias Blume (blume (at) tti - c (dot) org)
414    Date: 2005/05/18 13:35:00 CDT
415    Tag: Release_110_54
416    Description:
417    
418    New working version (110.54).  NEW BOOTFILES!
419    
420    ----------------------------------------------------------------------
421    Name: Matthias Blume (blume (at) tti - c (dot) org)
422    Date: 2005/05/18 11:58:00 CDT
423    Tag: blume-20050518-installer
424    Description:
425    
426    Added support scripts for Mac OS X PackageMaker and modified
427    config/install.sh so that it supports re-dumping a heap image
428    after customization.
429    
430    ----------------------------------------------------------------------
431    Name: Matthias Blume (blume (at) tti - c (dot) org)
432    Date: 2005/05/18 10:55:00 CDT
433    Tag: blume-20050518-realdiv-noovld
434    Description:
435    
436    Un-overloaded / to work around bug in overloading resolution code.
437    
438    ----------------------------------------------------------------------
439    Name: Matthias Blume (blume (at) tti - c (dot) org)
440    Date: 2005/05/16 23:50:00 CDT
441    Tag: blume-20050516-redump-heap
442    Description:
443    
444    Added mechanism for re-creating a heap file for the interactive system
445    after configuration variables have been changed.
446    
447       CM.redump_heap : string -> unit
448    
449    This is much like SMLofNJ.exportML, but starting from the resulting
450    heap does not return to the caller of CM.redump_heap but
451    restarts the interactive system from scratch.  The original call of
452    CM.redump_heap does not return but ends the interactive session.
453    Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
454    
455    Internally, redump_heap winds the dynamic execution context back to
456    the point where the original heap image was created and re-executes
457    the heap image generation code in the boot code.
458    
459    ----------------------------------------------------------------------
460    Name: Matthias Blume (blume (at) tti - c (dot) org)
461    Date: 2005/05/09 21:55:00 CDT
462    Tag: blume-20050509-word64
463    Description:
464    
465    Added a hack to the existing hack known as Word64 to make fromString
466    behave correctly.  I am still not sure whether Word64.scan will work
467    as specified with respect to the interaction of radix and prefix.
468    
469    ----------------------------------------------------------------------
470    Name: Allen Leung (leunga (at) reservoir (dot) com)
471    Date: 2005/05/04 11:50:00 EST
472    Tag: leunga-20050504-checkgc
473    Description:
474    
475    Added a gc protocol checking phase.  This phase is enabled with
476    the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
477    
478    ----------------------------------------------------------------------
479    Name: Matthias Blume (blume (at) tti - c (dot) org)
480    Date: 2005/05/04 10:45:00 CDT
481    Tag: blume-20050504-intinf
482    Description:
483    
484    Fixed a bug in the implementation of div and mod for IntInf.
485    Thanks to Neophytos Michael for reporting the problem.
486    
487    ----------------------------------------------------------------------
488    Name: Matthias Blume (blume (at) tti - c (dot) org)
489    Date: 2005/05/04 10:35:00 CDT
490    Tag: blume-20050504-join
491    Description:
492    
493    Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
494    
495    ----------------------------------------------------------------------
496    Name: Matthias Blume (blume (at) tti - c (dot) org)
497    Date: 2005/02/28 23:40:00 CST
498    Tag: blume-20050228-mVar
499    Description:
500    
501    Fixed serious bug (brown paper bag variety) in new implementation of
502    structure Atom in CML.  (I had accidentally used a mailbox instead of
503    an mvar, leaving the door open for races.)
504    
505    ----------------------------------------------------------------------
506    Name: Matthias Blume (blume (at) tti - c (dot) org)
507    Date: 2005/02/25 15:00:00 CST
508    Tag: Release_110_53
509    Description:
510    
511    New working version (110.53).  NEW BOOTFILES!
512    
513    ----------------------------------------------------------------------
514    Name: Matthias Blume (blume (at) tti - c (dot) org)
515    Date: 2005/02/25 14:50:00 CST
516    Tag: blume-20050225-susp
517    Description:
518    
519    Brought back SMLofNJ.Susp.  The underlying suspension type is the one
520    implemented in Core -- which means that it is the same as the one used
521    by the lazy extension.
522    
523    ----------------------------------------------------------------------
524    Name: Matthias Blume (blume (at) tti - c (dot) org)
525    Date: 2005/02/24 16:50:00 CST
526    Tag: blume-20050224-cml-atom
527    Description:
528    
529    Simpler and at the same time more general implementation of structure
530    Atom in CML.
531    
532    ----------------------------------------------------------------------
533    Name: Matthias Blume (blume (at) tti - c (dot) org)
534    Date: 2005/02/15 17:35:00 CST
535    Tag: blume-20050215-tools
536    Description:
537    
538    Created new "tools" directory under "src" and moved "TraceDebugProf"
539    there.
540    
541    ----------------------------------------------------------------------
542    Name: Matthias Blume (blume (at) tti - c (dot) org)
543    Date: 2005/02/10 17:55:00 CST
544    Tag: blume-20050210-longlong
545    Description:
546    
547    Implemented "long long" arguments and results for NLFFI.  (Only the
548    PPC/MacOS implementation is complete, the other backends still need to
549    be updated.)
550    
551    ----------------------------------------------------------------------
552    Name: Matthias Blume (blume (at) tti - c (dot) org)
553    Date: 2005/01/24 17:40:00 CST
554    Tag: blume-20050124-mlyacc
555    Description:
556    
557    Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
558    problem with certain "as" patterns which previously got rendered
559    using incorrect syntax.
560    
561    ----------------------------------------------------------------------
562    Name: Matthias Blume (blume (at) tti - c (dot) org)
563    Date: 2005/01/18 12:00:00 CST
564    Tag: blume-20050118-profile
565    Description:
566    
567    Made time profiling code (interrupt handler) in runtime system aware
568    of new array representation.
569    
570    ----------------------------------------------------------------------
571    Name: Matthias Blume (blume (at) tti - c (dot) org)
572    Date: 2005/01/14 18:00:00 CST
573    Tag: blume-20050114-heap2exec
574    Description:
575    
576    Implemented new (but still experimental) heap2exec facility.  This is
577    tested under Mac OS X and should work under Linux (will test shortly).
578    It will probably also work on the Sparc (will test some time later).
579    
580      - removed old "HACKED_STANDALONE" hack from runtime
581    
582    To be able to test this, uncomment the request for "heap2asm" in
583    config/targets prior to installation.  (Notice that this is different
584    from "heap2exec" mentioned below.  Not a typo.)
585    
586    To perform an actual test, run the command
587    
588       $ bin/heap2exec heapfile execfile
589    
590    (You can put heap2exec on your shell's path.)
591    
592    For example, run
593    
594      $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
595    
596    This will create a standalone executable called "mly" which you can
597    then invoke directly as a command.
598    
599    ----------------------------------------------------------------------
600    Name: Matthias Blume (blume (at) tti - c (dot) org)
601    Date: 2005/01/07 11:44:00 CST
602    Tag: blume-20050107-mlstring
603    Description:
604    
605    fixed off-by-one error in ML_STRING macro (globals.c)
606    
607    ----------------------------------------------------------------------
608    Name: Matthias Blume (blume (at) tti - c (dot) org)
609    Date: 2004/12/23 18:00:00 CST
610    Tag: blume-20041223-santa
611    Description:
612    
613    Made ml-build script "smarter" (but only very little).
614    
615    ----------------------------------------------------------------------
616    Name: Matthias Blume (blume (at) tti - c (dot) org)
617    Date: 2004/12/21 15:05:00 CST
618    Tag: blume-20041221-longlong
619    Description:
620    
621    * Implemented access to signed and unsigned long long data in NLFFI.
622    (The parameter-passing part of the picture has not complete. But
623    data structure access seems to work.)
624    
625    * Fixed CM's incorrect assumption that the PPC is little-endian.
626      (On the Mac, it is big-endian.  And that's currently our only
627      PPC platform.)
628    
629    ----------------------------------------------------------------------
630    Name: Matthias Blume (blume (at) tti - c (dot) org)
631    Date: 2004/12/21 12:50:00 CST
632    Tag: blume-20041221-memory
633    Description:
634    
635    Some cleanup in the $c/memory.cm library: separated some concerns by
636    moving allocation code and memory access code each into their own
637    files.
638    
639    ----------------------------------------------------------------------
640    Name: Allen Leung (leunga (at) reservoir (dot) com)
641    Date: 2004/12/17 16:12:00 EST
642    Tag: leunga-20041217-cygwin-smlnj-home
643    Description:
644    
645        The Unix I/O library of SML/NJ on cygwin does not understand
646    Windows style pathname, so problems arise when SMLNJ_HOME is set
647    to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
648    to a POSIX pathname on cygwin.
649    
650    ----------------------------------------------------------------------
651    Name: Matthias Blume (blume (at) tti - c (dot) org)
652    Date: 2004/12/16 13:35:00 CST
653    Tag: Release_110_52
654    Description:
655    
656    Last-minute changes incorporated into 110.52.  Release tag moved.
657    
658    The changes:
659       - HashString.hashString' -> HashString.hashSubstring
660       - bug fix in UnivariateStats
661    
662    ----------------------------------------------------------------------
663    Name: Matthias Blume (blume (at) tti - c (dot) org)
664    Date: 2004/12/15 23:40:00 CST
665    Tag: blume-20041215-hashSubstring
666    Description:
667    
668    - HashString.hashString' -> HashString.hashSubstring
669    - corresponding changes in atom.sml
670    - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
671      added some comments
672    
673    ----------------------------------------------------------------------
674    Name: Matthias Blume (blume (at) tti - c (dot) org)
675    Date: 2004/12/15 15:30:00 CST
676    Tag: (Release_110_52)
677    Description:
678    
679    New working version (110.52).  NEW BOOTFILES!
680    
681    ----------------------------------------------------------------------
682    Name: Matthias Blume (blume (at) tti - c (dot) org)
683    Date: 2004/12/15 12:45:00 CST
684    Tag: blume-20041215-spaces
685    Description:
686    
687    More on the space problem (this time for Win32).
688    
689    ----------------------------------------------------------------------
690    Name: Matthias Blume (blume (at) tti - c (dot) org)
691    Date: 2004/12/14 17:30:00 CST
692    Tag: blume-20041214-spaces
693    Description:
694    
695    Hacked some of the scripts (in particular: the installer) to cope with
696    spaces in filenames a bit better.  But beware: the current "solution"
697    is likely still full of bugs and inherently incomplete.  (We need to
698    do away with those shell scripts for a comprehensive solution.)
699    
700    ----------------------------------------------------------------------
701    Name: Matthias Blume (blume (at) tti - c (dot) org)
702    Date: 2004/12/13 14:45:00 CST
703    Tag: blume-20041213-ml-makedepend
704    Description:
705    
706    Fixed bug in code for ml-makedepend.
707    
708    ----------------------------------------------------------------------
709    Name: Matthias Blume (blume (at) tti - c (dot) org)
710    Date: 2004/12/09 16:30:00 CST
711    Tag: blume-20041209-statistics
712    Description:
713    
714    Added two simple but potentially useful statistics modules to SML/NJ Library.
715    (See CHANGES file there.)
716    
717    ----------------------------------------------------------------------
718    Name: Matthias Blume (blume (at) tti - c (dot) org)
719    Date: 2004/12/01 16:50:00 CST
720    Tag: blume-20041201-atom
721    Description:
722    
723    smlnj-lib:
724    
725      Added function HashString.hashString' for substrings.
726      Hand-inlined CharVector.foldl into HashString (for speed).
727      Modified implementation of structure Atom to avoid extracting
728      strings from substrings unless necessary.
729      (Also see CHANGES file for smlnj-lib.)
730    
731    ----------------------------------------------------------------------
732    Name: Matthias Blume (blume (at) tti - c (dot) org)
733    Date: 2004/11/24 22:15:00 CST
734    Tag: blume-20041124-cml
735    Description:
736    
737    Made sure CML compiles when Position = Int64.
738    
739    ----------------------------------------------------------------------
740    Name: Matthias Blume (blume (at) tti - c (dot) org)
741    Date: 2004/11/24 14:30:00 CST
742    Tag: blume-20041124-position
743    Description:
744    
745    The compiler can now be compiled in a mode that makes structure Position
746    equal to Int64.  The default, however, is unchanged (Position = Int31)
747    for the time being.
748    
749    To enable 64-bit positions, use the following procedure:
750    
751    1. Start sml
752    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
753    3. Type
754          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
755    4. Run CMB.make() as usual.
756    
757    This is barely tested.  The only test so far was a little SML program
758    counting the number of characters in an 8-gigabyte file by
759    reading it character-by-character.  That test was successful.
760    
761    In support of 64-bit positions, a number of new functions have been
762    added to the runtime system.
763    
764    ----------------------------------------------------------------------
765    Name: Matthias Blume (blume (at) tti - c (dot) org)
766    Date: 2004/11/23 14:45:00 CST
767    Tag: blume-20041123-useFile
768    Description:
769    
770    Fixed a problem with unhelpful error messages related to problems with
771    .cm- or .sml files that appear as part of the sml command line.
772    
773    ----------------------------------------------------------------------
774    Name: Matthias Blume (blume (at) tti - c (dot) org)
775    Date: 2004/11/18 15:40:00 CST
776    Tag: Release_110_51
777    Description:
778    
779    New working version (110.51).  NEW BOOTFILES!
780    
781    ----------------------------------------------------------------------
782    Name: Matthias Blume (blume (at) tti - c (dot) org)
783    Date: 2004/11/18 15:35:00 CST
784    Tag: <none>
785    Description:
786    
787    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
788    Joemann).
789    
790    ----------------------------------------------------------------------
791    Name: Allen Leung (leunga (at) reservoir (dot) com)
792    Date: 2004/11/17 16:05:21 EST 2004
793    Tag: leunga-20041117-mlrisc-live-kill
794    Description:
795    
796       Added support for MLTree constructs LIVE and KILL
797    to all the architectures.
798    
799    ----------------------------------------------------------------------
800    Name: Matthias Blume (blume (at) tti - c (dot) org)
801    Date: 2004/11/13 00:20:00 CST
802    Tag: blume-20041113-versiontool
803    Description:
804    
805    - Stripped down the versiontool: It now only handles the version number.
806      The date string is generated at bootstrap time (during makeml).
807    
808    - In a previous commit, fixed a minor issue with how polyequal is being
809      translated.  In particular, the code now "looks through" abstractions.
810      This results in slightly fewer polyEqual warnings and hopefully slightly
811      more efficient code.  Important examples for where this matters are
812      the new int64 and word64 types.
813    
814    ----------------------------------------------------------------------
815    Name: Matthias Blume (blume (at) tti - c (dot) org)
816    Date: 2004/11/12 00:30:00 CST
817    Tag: blume-20041112-int64
818    Description:
819    
820    Structure Int64 fully hooked in.  (The implementation is not very
821    efficient, though.)
822    
823    ----------------------------------------------------------------------
824    Name: Matthias Blume (blume (at) tti - c (dot) org)
825    Date: 2004/11/11 17:30:00 CST
826    Tag: blume-20041111-more64
827    Description:
828    
829    All the pieces of Word64 are now there, with the exception of the
830    conversions from and to LargeWord. (Eventually these need to be identities,
831    but for the time being they don't even make sense because LargeWord is
832    32-bit wide.)
833    
834    Also started to add similar support for Int64, but major pieces of that
835    are still missing.
836    
837    ----------------------------------------------------------------------
838    Name: Matthias Blume (blume (at) tti - c (dot) org)
839    Date: 2004/11/11 00:15:00 CST
840    Tag: blume-20041111-word64
841    Description:
842    
843    Structure Word64 is now (almost) complete, word literals and patterns
844    seem to work.  There are a few odd pieces missing.  In particular,
845    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
846    at the moment.
847    
848    Making Word64 official would mean that LargeWord becomes Word64.  But
849    this requires extreme care because most word-word conversions have to
850    go through LargeWord, so making a mistake means loss of efficiency or
851    worse.  Eventually there will be a solution similar to (but actually
852    simpler than) what I did with IntInf.
853    
854    ----------------------------------------------------------------------
855    Name: Matthias Blume (blume (at) tti - c (dot) org)
856    Date: 2004/11/10 18:12:00 CST
857    Tag: blume-20041110-64bit
858    Description:
859    
860    More 64-bit hacking (but still not even half-way there yet).
861    Also, some assorted improvements to the handling of 8-bit words.
862    
863    ----------------------------------------------------------------------
864    Name: Matthias Blume (blume (at) tti - c (dot) org)
865    Date: 2004/11/09 17:50:00 CST
866    Tag: <none>
867    Description:
868    
869    Started adding some infrastructure for supporting 64-bit int- and
870    word-types.  (Still in its very early stages.)
871    
872    ----------------------------------------------------------------------
873    Name: Matthias Blume (blume (at) tti - c (dot) org)
874    Date: 2004/10/28 10:45:00 CDT
875    Tag: Release_110_50
876    Description:
877    
878    New working version (110.50).  NEW BOOTFILES!
879    
880                 =====================
881    
882    Also:
883    
884      - Changend config/srcarchiveurl from a file just
885        containing the URL string into a file containing
886        shell script code.  The code has access to the $VERSION variable.
887      - Made corresponding changes to config/install.sh and config/unpack.
888      - Default contents of config/srcarchiveurl uses $VERSION and
889        normally does not have to be edited to reflect a version change.
890    
891      (As a result, a version change can be done by just editing
892       config/version, the rest is now automatic.)
893    
894    ----------------------------------------------------------------------
895    Name: Matthias Blume (blume (at) tti - c (dot) org)
896    Date: 2004/10/27 17:50:00 CDT
897    Tag: blume-20041027-btrace-msg
898    Description:
899    
900    BackTrace.monitor now also reports the source of the exception that
901    triggered the trace.
902    
903    ----------------------------------------------------------------------
904    Name: Matthias Blume (blume (at) tti - c (dot) org)
905    Date: 2004/10/27 17:20:00 CDT
906    Tag: blume-20041027-x86-c-calls
907    Description:
908    
909    This is the HISTORY entry for two earlier commits, both concerning
910    the x86 c-calls code in MLRISC:
911    
912      - added a missing LOAD in the code that deals with struct arguments
913      - made sure the caller does not add the wrong number of bytes to the
914        stack pointer after a call of a function returning a struct
915        (the callee already pops the implicit argument which points to
916         the space reserved for the result)
917    
918    ----------------------------------------------------------------------
919    Name: Allen Leung (leunga (at) reservoir (dot) com)
920    Date: 2004/10/24 14:00:00 EST
921    Tag: leunga-20041024-x86-gas-fucomip
922    Description:
923    
924       John discovered a bug in the syntax of fucomip.
925       The opcodes FU?COMIP? have been changed to
926    
927           fu?comip? %st(i), %st
928    
929    ----------------------------------------------------------------------
930    Name: Matthias Blume (blume (at) tti - c (dot) org)
931    Date: 2004/10/20 15:06:00 CDT
932    Tag: blume-20041020-standalone-backtrace
933    Description:
934    
935    Added a mechanism for getting back-trace information from standalone
936    programs.  Here is how it works:
937    
938       1. The part of the program from which you want to get backtrace
939          information (usually the whole program) should be wrapped with
940          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
941          main program could be modified from something like
942    
943             fun main (pgm, args) = ...
944    
945         to
946    
947            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
948    
949       2. To be able to access BackTrace.monitor, you have to add
950          library $smlnj-tdp/plugins.cm to the .cm file that contains your
951          main function.
952    
953       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
954          CM might have created in the past for your project).
955    
956       4. Build the system using this command line:
957    
958             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
959                      myprog.cm MyProg.main myprog
960    
961          instead of the usual
962    
963             ml-build myprog.cm MyProg.main myprog
964    
965    I changed a library name:
966    
967       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
968    
969    New libraries:
970    
971       $smlnj-tdp/back-trace.cm
972          -- when loaded causes the back-trace plugin to be installed
973       $smlnj-tdp/coverage.cm
974          -- when loaded causes the coverage plugin to be installed
975    
976    ----------------------------------------------------------------------
977    Name: Matthias Blume (blume (at) tti - c (dot) org)
978    Date: 2004/10/18 16:45:00 CDT
979    Tag: blume-20041018-groupowner
980    Description:
981    
982    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
983    
984    Eliminated group owner specs from .cm files throughout the source tree.
985    
986    ----------------------------------------------------------------------
987    Name: Matthias Blume (blume (at) tti - c (dot) org)
988    Date: 2004/10/15 15:45:00 CDT
989    Tag: blume-20041015-coverage
990    Description:
991    
992    * Test coverage tool added!
993    * Further reorganization of tracing-, debugging-, and profiling support:
994    
995        - moved original BTImp -- now called BackTrace -- into a separate
996          library called $/trace-debug-profile.cm
997        - eliminated all mentions of BTrace from SMLofNJ.Internals
998        - only the instrumentation mechanism is now left in the compiler proper
999        - BackTrace module is a plugin which is NOT plugged in by default
1000        - Coverage module is another such plugin
1001    
1002    To get the benefits of any of these plugin modules, the code in
1003    question must be compiled with tdp instrumentation turned on.  This
1004    can be done by setting SMLofNJ.Internals.TDP.mode to true.
1005    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
1006    
1007    Plugins are selected at link time.  (Pre-compiled instrumented code
1008    can be re-loaded with different plugins in effect.)  When an
1009    instrumented module is linked, whatever plugins are at that time
1010    enabled will come into effect for that module.
1011    
1012    To enable the back-trace plugin, load library $/trace-debug-profile.cm
1013    and invoke BackTrace.install() (e.g., from the interactive prompt).
1014    To enable the coverage plugin, load the same library and invoke
1015    Coverage.install().
1016    
1017    Back-traces are generated automatically on uncaught exceptions and
1018    when the code in question explicitly invokes BackTrace.trigger().
1019    
1020    Coverage (and execution frequency-) information must be queried
1021    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
1022    
1023    ----------------------------------------------------------------------
1024    Name: Matthias Blume (blume (at) tti - c (dot) org)
1025    Date: 2004/10/14 17:40:00 CDT
1026    Tag: blume-20041014-tdp-core
1027    Description:
1028    
1029    Snapshot of a significant overhaul of how the trace/debug/profile support
1030    is hooked into the system (specifically: Core and SMLofNJ.Internals).
1031    
1032    ----------------------------------------------------------------------
1033    Name: Matthias Blume (blume (at) tti - c (dot) org)
1034    Date: 2004/10/13 16:34:00 CDT
1035    Tag: blume-20041013-tdp
1036    Description:
1037    
1038    Some rationalization of names:
1039    
1040       structure BTrace -> structure TDPInstrument
1041       etc.
1042    
1043    This is is preparation of using the original back-trace
1044    instrumentation for other purposes.  "TDP" stands for
1045    Trace/Debug/Profile.
1046    
1047    The control flag controlling whether instrumentation is on or off is now
1048    registered under a different name, so instead of running sml as
1049    
1050     sml -Cinstrument.btrace-mode=true
1051    
1052    one has to say
1053    
1054     sml -Ctdp.instrument=true
1055    
1056    ----------------------------------------------------------------------
1057    Name: Matthias Blume (blume (at) tti - c (dot) org)
1058    Date: 2004/10/11 16:37:00 CDT
1059    Tag: blume-20041011-regions
1060    Description:
1061    
1062    Made some minor modifications to elabcore.sml to have source regions
1063    be propagated more tightly -- resulting in better (i.e., smaller)
1064    regions being reported in error- and debug messages.
1065    
1066    ----------------------------------------------------------------------
1067    Name: Matthias Blume (blume (at) tti - c (dot) org)
1068    Date: 2004/10/08 22:50:00 CDT
1069    Tag: blume-20041008-cmkw
1070    Description:
1071    
1072    Fixed handling of keywords in .cm files: After seeing "is" the lexer
1073    treats subsequent occurrences of "group", "library", "source", "is",
1074    "*", and "-" as ordinary identifiers rather than keywords.
1075    
1076    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
1077    is supposed to accept a tool argument called "source", but this did
1078    not work because of the clash with the keyword.
1079    
1080    ----------------------------------------------------------------------
1081    Name: Matthias Blume (blume (at) tti - c (dot) org)
1082    Date: 2004/10/07 16:00:00 CDT
1083    Tag: blume-20041007-cleanup
1084    Description:
1085    
1086    Assorted cleanup work:
1087    
1088      - got rid of intstrmap in favor of using the library's
1089        hash table implementation
1090      - threw out most of the pathnames stuff, as it was not used anyway
1091      - simplified tokentable implementation
1092      - fixed some minor spelling errors
1093    
1094    ----------------------------------------------------------------------
1095    Name: Matthias Blume (blume (at) tti - c (dot) org)
1096    Date: 2004/10/06 15:15:15 CDT
1097    Tag: blume-20041006-handler
1098    Description:
1099    
1100    Cleaned up the absyn to reflect the invariant that HANDLE always
1101    carries a FNexp as part of the type definition.  This eliminates some
1102    superfluous sanity checks at runtime down the road.
1103    
1104    Some minor cleanup of the btrace code.
1105    
1106    ----------------------------------------------------------------------
1107    Name: Matthias Blume (blume (at) tti - c (dot) org)
1108    Date: 2004/10/01 10:20:30 CDT
1109    Tag: blume-20041001-slave
1110    Description:
1111    
1112    Added hack to make slave mode work in the presence of the version
1113    tool.  (Still, since the master does two passes over the code for
1114    CMB.make, the release number gets bumped twice when slaves are
1115    attached.  I don't know if this is worth fixing...)
1116    
1117    ----------------------------------------------------------------------
1118    Name: Matthias Blume (blume (at) tti - c (dot) org)
1119    Date: 2004/09/30 10:55:00 CDT
1120    Tag: blume-20040930-version
1121    Description:
1122    
1123    * Moved the "version" magic into its own little library under
1124      src/system/smlnj/internal.  This avoids expensive reconstruction of
1125      a stable src/compiler/core.cm.
1126    
1127    * At the same time, structure CompilerVersion is now known as
1128      structure SMLNJVersion.
1129    
1130    * Arranged for the version tool to NOT kick in when rebuilding the system
1131      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
1132      Also, loading the versiontool does not work when rebuilding the system
1133      because CM is not properly initialized at that time.
1134    
1135    ----------------------------------------------------------------------
1136    Name: Matthias Blume (blume (at) tti - c (dot) org)
1137    Date: 2004/09/29 14:00:00 CDT
1138    Tag: blume-20040929-autoversion
1139    Description:
1140    
1141    Implemented some CM magic to have
1142        file src/compiler/TopLevel/main/version.sml
1143    generated automagically.
1144    The version is taken from two files: config/version and config/release.
1145    The first is expected to contain a two-part version number such as 110.49.
1146    The second should contain a single number, but it may be missing.
1147    
1148    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
1149    time the version tool is loaded (which is the first time you say CMB.make),
1150    then the tool will increment the value stored in config/release every
1151    time CMB.make is invoked.
1152    
1153    The binfile format is now insensitive to anything beyond the first
1154    two components of a version number, so bumping the release does not render
1155    binfiles incompatible.  Auto-bumping can be used to keep track of versions
1156    during development without invalidating existing binfiles.
1157    
1158    In any case, every CMB.make updates the date information in version.sml.
1159    (This is the date that is printed in the banner.)
1160    
1161    ----------------------------------------------------------------------
1162    Name: Matthias Blume (blume (at) tti - c (dot) org)
1163    Date: 2004/09/28 10:53:00 CDT
1164    Tag: blume-20040928-controls
1165    Description:
1166    
1167    Some cleanup of the controls code.
1168    
1169    ----------------------------------------------------------------------
1170    Name: Matthias Blume (blume (at) tti - c (dot) org)
1171    Date: 2004/09/27 22:08:00 CDT
1172    Tag: blume-20040927-controls
1173    Description:
1174    
1175    Added two pieces of functionality to the Controls interface:
1176    
1177    1.   val save'restore: 'a control -> unit -> unit
1178    
1179       grabs the current value of the control in stage 1 and restores it
1180       in stage 2.
1181    
1182    2.   val set' : 'a control * 'a -> unit -> unit
1183    
1184       stores the given value into the control in stage 2 (i.e., delayed)
1185       but does all error checking in stage 1.
1186       (This is for string controls that need to do parse their argument --
1187        something that might fail.  In some cases, notably in CM, one
1188        already knows the intended argument but wants to delay the actual
1189        assignment until a time when error recovery would be more difficult.)
1190    
1191    Changed the handling of controls in tool arguments to classes "sml" and
1192    "lazysml":
1193       - use Controls.save'restore as a more robust way of restoring the
1194         old value (in particular: without having to re-parse the string)
1195       - use controls to handle the "overload" keyword in the init group
1196         (I believe this change actually fixes a long-standing obscure bug.)
1197    
1198    ----------------------------------------------------------------------
1199    Name: Matthias Blume (blume (at) tti - c (dot) org)
1200    Date: 2004/09/27 17:00:00 CDT
1201    Tag: blume-20040927-lazysml
1202    Description:
1203    
1204    Added a new tool class called "lazysml" to CM's tool chest.  The only
1205    difference to "sml" is that compilation is done with Control.lazysml
1206    set to true.  A source of class "lazysml" is automatically recognized
1207    by a file name suffix of ".lml".
1208    
1209    In addition to the above feature, the original class "sml" now also
1210    supports a tool argument "lazy" which has the same effect.  As a
1211    result, the following three lines are equivalent:
1212    
1213        foo.sml : lazysml
1214        foo.sml : sml (lazy)
1215        foo.sml (lazy)
1216    
1217    The setting goes into effect both during parsing and during
1218    compilation.  The original setting is restored right after parsing and
1219    after compilation, respectively.
1220    
1221    In addition to all the above, there is also a general mechanism to set
1222    ANY of the "controls" that are available at the command line via
1223    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
1224    apply as well. (In fact, "lazy" is implemented as a special case of
1225    the general mechanism.)
1226    
1227    The .cm file syntax uses a new keyword tool argument called "with".
1228    There are several ways of indicating the desired settings:
1229    
1230        foo.sml (with:parser.quotations=true)
1231        foo.sml (with:(name:parser.quotations value:true))
1232        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
1233        foo.sml (with:(name1=value1 name2=value2 ...))
1234        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
1235    
1236        etc.
1237    
1238    Another possible abbreviation is to leave out the =v or value:v part
1239    if the name refers to a boolean control (in which case the value is
1240    taken to be true).  Thus, one could get lazy sml also by saying:
1241    
1242        foo.sml (with:parser.lazy-keyword=true)
1243        foo.sml (with:parser.lazy-keyword)
1244        foo.sml (with:(name:parser.lazy-keyword value:true))
1245        foo.sml (with:(name:parser.lazy-keyword))
1246    
1247    ----------------------------------------------------------------------
1248    Name: Matthias Blume (blume (at) tti - c (dot) org)
1249    Date: 2004/09/24 16:22:00 CDT
1250    Tag: blume-20040924-ppc-long-branch
1251    Description:
1252    
1253    Turned message about "emiting long form of branch" off by default.
1254    Added a control flag to turn it back on when desired.
1255    
1256    ----------------------------------------------------------------------
1257    Name: Matthias Blume (blume (at) tti - c (dot) org)
1258    Date: 2004/09/24 16:05:00 CDT
1259    Tag: blume-20040924-rounding
1260    Description:
1261    
1262    Applied patch for setting rounding modes under Mac OS X.  Thanks to
1263    Melissa O'Neill for providing the code!
1264    
1265    ----------------------------------------------------------------------
1266    Name: Matthias Blume (blume (at) tti - c (dot) org)
1267    Date: 2004/09/23 17:30:00 CDT
1268    Tag: blume-20040923-envvars
1269    Description:
1270    
1271    1. Changed definition of type ControlRegistry.registry_tree to
1272       include control_info (i.e., the name of the controlling environment
1273       variable).
1274    
1275    2. Added command-line flags -e and -E to print the names of environment
1276       variables that can be used to control internal settings.
1277       (This uses the new API mentioned in 1.)
1278    
1279    ----------------------------------------------------------------------
1280    Name: Matthias Blume (blume (at) tti - c (dot) org)
1281    Date: 2004/09/13 16:50:00 CDT
1282    Tag: Release_110_49
1283    Description:
1284    
1285    New working version (110.49).  NEW BOOTFILES!
1286    
1287    ----------------------------------------------------------------------
1288    Name: Matthias Blume
1289    Date: 2004/09/13 16:20:00 CDT
1290    Tag: blume-20040913-config-mlrisc
1291    Description:
1292    
1293    Put target "mlrisc" back into the default list.
1294    (There is no harm in having it, and some users have expressed their
1295    wish to have "mlrisc" included by default.)
1296    
1297    ----------------------------------------------------------------------
1298    Name: John Reppy
1299    Date: 2004/09/13
1300    Tag: jhr-20040913-signals
1301    Description:
1302    
1303    Fixed the signal masking code to properly nest mask/unmask operations
1304    on a per-signal basis.
1305    
1306    ----------------------------------------------------------------------
1307    Name: Matthias Blume (blume (at) tti - c (dot) org)
1308    Date: 2004/09/08 13:20:00 CDT
1309    Tag: blume-20040908-heap-magic
1310    Description:
1311    
1312    Bumped the heap macig number to 0x09082004 to account for the changed
1313    layout of the ML frame under MacOS X.
1314    
1315    ----------------------------------------------------------------------
1316    Name: Allen Leung (leunga (at) reservoir (dot) com)
1317    Date: 2004/09/03 11:26:00 EST
1318    Tag: leunga-20040903-cygwin-install
1319    Description:
1320    
1321       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
1322    The Cygwin runtime is turned on by setting the environment
1323    variable SMLNJ_CYGWIN_RUNTIME to 1.
1324    
1325    ----------------------------------------------------------------------
1326    Name: Matthias Blume (blume (at) tti - c (dot) org)
1327    Date: 2004/08/31 17:15:00 CDT
1328    Tag: blume-20040831-core
1329    Description:
1330    
1331    Added some exports to src/compiler/core.cm upon request by J. Joemann.
1332    
1333    ----------------------------------------------------------------------
1334    Name: Matthias Blume (blume (at) tti - c (dot) org)
1335    Date: 2004/08/30 17:55:00 CDT
1336    Tag: blume-20040830-installer
1337    Description:
1338    
1339    Upon request by Johannes Joemann:
1340    
1341    - improved ML code of installer to fall back to coping when renaming
1342      fails (i.e., when source and target are on different file systems);
1343      the code compiles but has yet to be tested in anger
1344    - removed mlrisc from list of default targets (config/targets)
1345    
1346    ----------------------------------------------------------------------
1347    Name: Matthias Blume (blume (at) tti - c (dot) org)
1348    Date: 2004/08/27 17:20:00 CDT
1349    Tag: blume-20040827-ptreql
1350    Description:
1351    
1352    Added ptreql primop to structure InlineT (upon request from Larry
1353    Paulson).
1354    
1355    ----------------------------------------------------------------------
1356    Name: Allen Leung (leunga (at) reservoir (dot) com)
1357    Date: 2004/08/15 21:21:00 EST
1358    Tag: leunga-110_48-udgraph
1359    Description:
1360    
1361    Another bug fix from Carl Hauser:
1362    
1363    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
1364    > 48c48
1365    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
1366    > e::L)
1367    > ---
1368    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
1369    > Without this, any deletion of an edge in an undirected graph does severe
1370    > violence to the graph.
1371    
1372    ----------------------------------------------------------------------
1373    Name: Allen Leung (leunga (at) reservoir (dot) com)
1374    Date: 2004/08/10 23:35:00 EST
1375    Tag: leunga-110_48-ppc
1376    Description:
1377    
1378       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
1379    
1380    ----------------------------------------------------------------------
1381    Name: Matthias Blume (blume (at) tti - c (dot) org)
1382    Date: 2004/08/10 12:00:00 CDT
1383    Tag: Release_110_48
1384    Description:
1385    
1386    New working version (110.48).  NEW BOOTFILES!
1387    
1388    ----------------------------------------------------------------------
1389    Name: Allen Leung (leunga (at) reservoir (dot) com)
1390    Date: 2004/08/09 12:21:00 EST
1391    Tag: leunga-110_47-dijsktra
1392    Description:
1393    
1394    Bug fix from Carl Hauser:
1395    
1396    single_source_shortest_paths in dijkstra.sml was observed to get wrong
1397    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1398    
1399    The problem is that following the expression A.update(dist,s,Num.zero)
1400    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1401    
1402    ----------------------------------------------------------------------
1403    Name: Matthias Blume (blume (at) tti - c (dot) org)
1404    Date: 2004/08/06 18:10:00 CDT
1405    Tag: blume-20040806-cmdline
1406    Description:
1407    
1408    Fiddled with handling of command-line options:
1409    
1410      * sml now quits after processing the command line
1411        if -H, -S, -h<n>, or -s<n> appears as the last
1412        command-line argument
1413      * a new option -q terminates the session when encountered on
1414        the command line; subsequent arguments will be ignored
1415      * bug fixes: short (erroneous) arguments are no longer ignored
1416        completely
1417    
1418    ----------------------------------------------------------------------
1419    Name: Allen Leung (leunga (at) reservoir (dot) com)
1420    Date: 2004/08/04 18:17:00 EST
1421    Tag: leunga-110_47-ppc-ibm-asm
1422    Description:
1423    
1424      - Added minimal IBM assembly syntax support for PowerPC.
1425    
1426      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1427        been altered in the runtime.  We need an automatic way to keep the file
1428        in sync.
1429    
1430    ----------------------------------------------------------------------
1431    Name: Matthias Blume (blume (at) tti - c (dot) org)
1432    Date: 2004/08/04 14:00:00 CDT
1433    Tag: Release_110_47
1434    Description:
1435    
1436    New working version (110.47).  NEW BOOTFILES!
1437    
1438    ----------------------------------------------------------------------
1439    Name: Matthias Blume (blume (at) tti - c (dot) org)
1440    Date: 2004/08/03 14:25:00 CDT
1441    Tag: blume-20040803-callingconv
1442    Description:
1443    
1444    Added low-level support for choosing C calling conventions by
1445    twiddling the type of rawccall.  (See
1446    src/compiler/Semant/types/cproto.sml for details.)
1447    
1448    ----------------------------------------------------------------------
1449    Name: Matthias Blume (blume (at) tti - c (dot) org)
1450    Date: 2004/08/02 15:55:00 CDT
1451    Tag: blume-20040802-backout
1452    Description:
1453    
1454    Backed out of change to win32-filesys.c.  The earlier patch to
1455    get_file_time caused CM to produce files with the wrong time stamp.
1456    
1457    ----------------------------------------------------------------------
1458    Name: Matthias Blume (blume (at) tti - c (dot) org)
1459    Date: 2004/08/02 14:45:00 CDT
1460    Tag: blume-20040802-nlffi-win32
1461    Description:
1462    
1463    Added NLFFI support for Win32, adapted from a patch provided by David
1464    Hansel.  This is currently completely untested.  Also, the issue
1465    concerning stdcall vs. ccall is still unresolved.
1466    
1467    ----------------------------------------------------------------------
1468    Name: Matthias Blume (blume (at) tti - c (dot) org)
1469    Date: 2004/07/30 17:55:00 CDT
1470    Tag: blume-20040730-various
1471    Description:
1472    
1473    Gearing up towards 110.47...
1474    
1475    - various minor bugfixes to ml-nlffigen
1476    - a beginning of a manual for nlffi
1477    
1478    - eliminated 'export name=value' in config/install.sh as this does
1479      not work with certain versions of /bin/sh
1480      (Thanks to David King at Motorola for catching this.)
1481    
1482    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1483      - added a test for tm==NULL to gmtime.c and localtime.c
1484      - applied patch for incorrect GetFileTime under win32
1485      - toSeconds -> toMilliseconds in Win32/win32-process.sml
1486    
1487    ----------------------------------------------------------------------
1488    Name: Matthias Blume (blume (at) tti - c (dot) org)
1489    Date: 2004/07/21 18:20:00 CDT
1490    Tag: blume-20040721-nlffigen
1491    Description:
1492    
1493    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1494      for a typedef to an incomplete type, but leave out the "typ" member.
1495      (This is just for consistency.)
1496    - Started to produce what is supposed to become better (i.e., comprehensive)
1497      documentation of what ml-nlffigen does and produces.
1498    
1499    ----------------------------------------------------------------------
1500    Name: Matthias Blume (blume (at) tti - c (dot) org)
1501    Date: 2004/07/14 16:25:00 CDT
1502    Tag: blume-20040714-union
1503    Description:
1504    
1505    Added C_UNION to c-calls/c-types.sml and updated the machinery
1506    (ml-nlffigen, cproto.sml) that conveys C function interface
1507    information to the code generator.
1508    
1509    However, the actual architecture-specific implementation of function
1510    arguments and results that are C unions is still not implemented.
1511    
1512    ----------------------------------------------------------------------
1513    Name: Allen Leung (leunga (at) reservoir (dot) com)
1514    Date: 2004/07/14 14:38:00 EST
1515    Tag: leunga-110_46_1-ppc-lwzu
1516    Description:
1517    
1518        Added these instructions to the PowerPC architecture:
1519        LBZU(X), LHZU(X), LWZU(X),
1520        STWU(X), STFDU, STFSU
1521    
1522        etc...
1523    
1524        Note: I haven't added their instruction encoding into the description.
1525    
1526    ----------------------------------------------------------------------
1527    Name: Allen Leung (leunga (at) reservoir (dot) com)
1528    Date: 2004/07/13 15:04:00 EST
1529    Tag: leunga-110_46_1-ppc-lwarx
1530    Description:
1531    
1532       Added the two instructions LWARX and STWCX to the PowerPC
1533    instruction set.
1534    
1535       A (untested) rewrite of loop-structure.sml.  The old version
1536    is completely broken.
1537    
1538    ----------------------------------------------------------------------
1539    Name: Matthias Blume (blume (at) tti - c (dot) org)
1540    Date: 2004/07/13 13:50:00 CDT
1541    Tag: blume-20040713-nlffi
1542    Description:
1543    
1544    - use paramAlloc to report c-calls with too many arguments
1545      (for PPC version where parameter area is pre-allocated)
1546    - added ccall_maxargspace to machspec (to implement the above)
1547    - made "make" commend in CM's "make" tool configurable
1548    - added option (default: on) for passing the name of the SML/NJ's "bin"
1549      directory to "make"; the call looks like this:
1550    
1551         make <options> SMLNJ_BINDIR=<dir> <target>
1552    
1553      This can be used by the Makefile to, e.g., pick the "right" version
1554      of ml-nlffigen.
1555    - minor code tweaks
1556    
1557    ----------------------------------------------------------------------
1558    Name: Matthias Blume (blume (at) tti - c (dot) org)
1559    Date: 2004/07/12 22:50:00 CDT
1560    Tag: blume-110_46_1-macosx-nlffi
1561    Description:
1562    
1563    NLFFI under Mac OS X now working (sort of).  This is largely untested,
1564    though.
1565    
1566    Note:  1. You have to make a new, clean build of the runtime system.
1567           2. There are new BOOTFILES, you have to use them!
1568              (Doing the bootstrap process yourself would be *very* painful!
1569               If you absolutely have to do it, build the system under
1570               a different architecture and then cross-compile.)
1571    
1572    Version bumped to 110.46.1 to account for runtime data format changes.
1573    
1574    ----------------------------------------------------------------------
1575    Name: Matthias Blume (blume (at) tti - c (dot) org)
1576    Date: 2004/06/18 14:30:00 CDT
1577    Tag: blume-20040618-unix
1578    Description:
1579    
1580    Changed the implementation of structure Unix so that the same stream
1581    is returned every time one of the {text,bin}{In,Out}streamOf functions
1582    is invoked on the same proc.  This is not what the spec currently
1583    says -- although IMO it arguably should.  (See discussion below.)
1584    
1585    ----------------------------------------------------------------------
1586    Name: Matthias Blume (blume (at) tti - c (dot) org)
1587    Date: 2004/06/17 18:15:00 CDT
1588    Tag: Release_110_46
1589    Description:
1590    
1591    New working version (110.46).  NEW BOOTFILES!
1592    
1593    ----------------------------------------------------------------------
1594    Name: Matthias Blume (blume (at) tti - c (dot) org)
1595    Date: 2004/06/17 17:20:00 CDT
1596    Tag: blume-20040617-timer-unix
1597    Description:
1598    
1599    Changed the interface of structures Timer and Unix to match the most
1600    recent Basis spec.
1601    
1602    In the case of Unix there still seems to be an open/weird issue:
1603    
1604       The {text,bin}{In,Out}streamOf functions are supposed to create
1605       fresh streams whenever they are called -- as opposed to have them
1606       return the same stream every time.  This design is supposed to
1607       prevent space leaks caused by proc values hanging on to streams.
1608    
1609       The reap function, on the other hand, is supposed to close the
1610       streams.  This cannot be done without having a handle on the
1611       stream in proc after all...
1612    
1613    I took the liberty to implement the following stopgap solution:
1614    
1615      The proc value hangs on to the most recently created stream(s).
1616      Reap closes those.  If either or both of the two streams hadn't
1617      been created at all yet, then reap will close the corresponding
1618      file descriptors directly.
1619    
1620    PS: I don't understand the original space leak argument anymore.  If
1621    a proc hangs on to the imperative stream, then I/O operations on those
1622    will advance the state of the cached stream and avoid the space leak.
1623    
1624    ----------------------------------------------------------------------
1625    Name: Matthias Blume (blume (at) tti - c (dot) org)
1626    Date: 2004/05/28 16:45:00 CDT
1627    Tag: blume-20040528-basis
1628    Description:
1629    
1630    Added signature PACK_REAL and exported functor PrimIO.
1631    
1632    ----------------------------------------------------------------------
1633    Name: Matthias Blume (blume (at) tti - c (dot) org)
1634    Date: 2004/05/25 16:00:00 CDT
1635    Tag: blume-20040525-group-owner
1636    Description:
1637    
1638    CM now ignores (but still accepts) the "owner" information in group
1639    descriptions.  The owner of a group is its next enclosing
1640    library. Each group must have a unique owner.  (There is a virtual
1641    "toplevel" library that owns groups which are not nested within a real
1642    library.)  Previously, each group had to explicitly declare its owner,
1643    and CM would check that such a declaration is correct.  The new scheme
1644    is to have CM check that for each group there is precisely one owning
1645    library.
1646    
1647    The advantage of the new scheme is that the programmer no longer needs
1648    to maintain the somewhat annoying owner information.  The downside is
1649    that CM cannot enforce the ownership rule across multiple runs of
1650    CM.make.  Fortunately, enclosing the same group in two different
1651    libraries A and B which are not part of the same program does not
1652    cause real problems.
1653    
1654    ----------------------------------------------------------------------
1655    Name: Matthias Blume (blume (at) tti - c (dot) org)
1656    Date: 2004/05/20 16:00:00 CDT
1657    Tag: blume-20040520-win32
1658    Description:
1659    
1660    Made the win32 version work again.  (Strangely, a misplaced comma had
1661    slipped into win32-process.c which prevented the runtime from being
1662    compiled correctly.)
1663    
1664    Also, included a minor addition to ml-build.bat analogous to what was
1665    done in blume-20040519-ml-build.
1666    
1667    ----------------------------------------------------------------------
1668    Name: Matthias Blume (blume (at) tti - c (dot) org)
1669    Date: 2004/05/19 22:10:00 CDT
1670    Tag: blume-20040519-ml-build
1671    Description:
1672    
1673    Arranged for ml-build to clean up after itself a little bit better.
1674    The script generates a temporary SML source file and compiles it using
1675    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1676    gets rid of those at the end, so they don't accumulate under .cm.
1677    
1678    This required a minor change to install.sh because the name of the
1679    metadata directory (default: .cm) is actually configurable at
1680    installation time.
1681    
1682    ----------------------------------------------------------------------
1683    Name: Matthias Blume (blume (at) tti - c (dot) org)
1684    Date: 2004/05/18 15:50:00 CDT
1685    Tag: blume-20040518-mkreader
1686    Description:
1687    
1688    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1689    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1690    
1691    ----------------------------------------------------------------------
1692    Name: Matthias Blume (blume (at) tti - c (dot) org)
1693    Date: 2004/05/11 14:35:00 CDT
1694    Tag: blume-20040511-win32sock
1695    Description:
1696    
1697    Added previously missing support for many socket-related functions
1698    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1699    for the voluminous patch!
1700    
1701    (I have not tested this patch under win32 yet.)
1702    
1703    Here is David's e-mail:
1704    
1705    Hi,
1706    
1707    Attached to this email you find a diff against sml/nj 110.45
1708    that will enable socket support under Windows.
1709    
1710    To apply the patch (using unix or cygwin)
1711    1) gunzip runtime.diff.gz
1712    2) "cd" into "src/runtime" in the source tree of a fresh
1713       110.45 installation.
1714    3) patch -p 1 < [your/path/to]runtime.diff
1715    
1716    The code compiles fine but has NOT yet been extensively tested.
1717    I only ran a few tests for basic socket client functionality
1718    (which worked fine).  Especially the functions that use ioctl
1719    are not tested at all and might not work (see below).
1720    
1721    I implemented this since we want to move to a newer version of sml/nj
1722    but need socket support in order to use it.  This is the first time I
1723    even had a look at the sml/nj source,  so please review my changes
1724    before making this part of the distribution!  Here are a few issues
1725    that I think might be better for someone to solve who is more
1726    familiar with the sml/nj source (and socket programming):
1727    
1728    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1729      exception since I could not figure out what the windows equivalent
1730      of these functions is
1731    
1732    - In sockets-osdep.h there are a some #include statements that are
1733      only used in a few files that include sockets-osdep.h
1734    
1735    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1736      does not process its return value since I don't know how to
1737      report an error upwards.
1738    
1739    - It would probably be good to have a call to WSACleanup() when
1740      the library is unloaded (if there is such a possibility).
1741      Otherwise I think Windows will take care of this automatically
1742      when the process finishes.
1743    
1744    - I used ioctlsocket() as a replacement for ioctl() but I have
1745      no idea if that is actually the proper replacement on Windows.
1746    
1747    - All these issues are marked in the code by "FIXME" comments.
1748    
1749    We use sml/nj extensively in our products and are quite happy
1750    with it.  I hope this contribution will help you.
1751    
1752    Keep up the good work!
1753    
1754    David
1755    
1756    ----------------------------------------------------------------------
1757    Name: Matthias Blume (blume (at) tti - c (dot) org)
1758    Date: 2004/05/11 14:20:00 CDT
1759    Tag: blume-20040511-installml
1760    Description:
1761    
1762    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1763    the patch.)
1764    
1765    ----------------------------------------------------------------------
1766    Name: Matthias Blume (blume (at) tti - c (dot) org)
1767    Date: 2004/05/11 14:05:00 CDT
1768    Tag: blume-20040511-nlffi-netbsd
1769    Description:
1770    
1771    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1772    the patch.)
1773    
1774    ----------------------------------------------------------------------
1775    Name: Matthias Blume (blume (at) tti - c (dot) org)
1776    Date: 2004/05/11 12:05:00 CDT
1777    Tag: blume-20040511-exports
1778    Description:
1779    
1780    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1781    export lists in compiler-related .cm-files.
1782    
1783    ----------------------------------------------------------------------
1784    Name: Matthias Blume (blume (at) tti - c (dot) org)
1785    Date: 2004/05/11 11:35:00 CDT
1786    Tag: blume-20040511-allsource
1787    Description:
1788    
1789    The installer now honors the "src-smlnj" target again, although its meaning
1790    has changed from "all sources required for the compiler" to "all sources
1791    the installer knows about".  In other words, if you enable "src-smlnj"
1792    in the "targets" file, then the installer will pull in sources for
1793    everything.  (Notice that this refers to source code only.  Compiled
1794    code is still only installed for modules that were requested explicitly
1795    or which are required for other modules that were requested explicitly.)
1796    
1797    ----------------------------------------------------------------------
1798    Name: Matthias Blume (blume (at) tti - c (dot) org)
1799    Date: 2004/04/23 17:40:00 CDT
1800    Tag: blume-20040423-ieee-scan
1801    Description:
1802    
1803    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1804    in the exponent calculation we get INF or ZERO (depending on the mantissa
1805    and the sign of the exponent).
1806    
1807    ----------------------------------------------------------------------
1808    Name: Matthias Blume (blume (at) tti - c (dot) org)
1809    Date: 2004/04/23 10:40:00 CDT
1810    Tag: blume-20040423-ml-build
1811    Description:
1812    
1813    The ml-build script now terminates with a non-0 status when something
1814    goes wrong.
1815    
1816    ----------------------------------------------------------------------
1817    Name: Matthias Blume (blume (at) tti - c (dot) org)
1818    Date: 2004/04/22 16:35:00 CDT
1819    Tag: blume-20040422-Option
1820    Description:
1821    
1822    Made exception Option to be the same as exception Option.Option
1823    (as it should be).
1824    
1825    ----------------------------------------------------------------------
1826    Name: Allen Leung (leunga (at) reservoir (dot) com)
1827    Date: 2004/03/19 14:40:00 EST
1828    Tag: leunga-20040319-cygwin-nlffi
1829    Description:
1830    
1831    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1832    of SML/NJ.  The problem is that
1833    
1834       lib = dlopen(NULL, ...)
1835       f   = dlsym(lib, "malloc");
1836    
1837    does not work on Windows unless we explicitly export symbols
1838    such as 'malloc' during linking.  We fixed this by explicitly
1839    exporting the required symbols with the magic gcc incantation:
1840    
1841        -Wl,--export-all cygwin.def
1842    
1843    where cygwin.def is a file containing all the symbols that
1844    we wish to export.
1845    
1846    I suspect this is a Windows problem and we'll have to
1847    do the same (somehow with windows compilers) when
1848    we build the native win32 version with the system
1849    calls LoadLibrary/GetProcAddress.
1850    
1851    ----------------------------------------------------------------------
1852    Name: Matthias Blume (blume (at) tti - c (dot) org)
1853    Date: 2004/03/04 16:35:00 CST
1854    Tag: blume-20040304-intinf-fmt
1855    Description:
1856    
1857    Fixed problem with IntInf.fmt (sign would show up on the right instead
1858    of on the left for BIN, OCT, and HEX).
1859    
1860    ----------------------------------------------------------------------
1861    Name: Matthias Blume (blume (at) tti - c (dot) org)
1862    Date: 2004/03/04 11:25:00 CST
1863    Tag: blume-20040304-symlinks
1864    Description:
1865    
1866    Fixed problem with installer script (unix only) where bin/ml-yacc and
1867    friends pointed (via symlinks) to absolute locations instead of just
1868    .run-sml.  This was reported by Vesa A Norrman.
1869    
1870    ----------------------------------------------------------------------
1871    Name: Matthias Blume (blume (at) tti - c (dot) org)
1872    Date: 2004/02/13 14:50:00 CST
1873    Tag: Release_110_45
1874    Description:
1875    
1876    New working version (110.45). New bootfiles.
1877    
1878    ----------------------------------------------------------------------
1879    Name: Matthias Blume (blume (at) tti - c (dot) org)
1880    Date: 2004/01/26 15:15:15 CST
1881    Tag: blume-20040126-toplevel
1882    Description:
1883    
1884    Improved handling of exceptions at the interactive toplevel.
1885    
1886    ----------------------------------------------------------------------
1887    Name: Matthias Blume (blume (at) tti - c (dot) org)
1888    Date: 2004/01/26 11:25:00
1889    Tag: blume-20040126-app
1890    Description:
1891    
1892    Type of top-level "app" corrected.
1893    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1894    
1895    ----------------------------------------------------------------------
1896    Name: Matthias Blume (blume (at) tti - c (dot) org)
1897    Date: 2003/11/18 17:10 CST
1898    Tag: blume-20031118-basis-fiddle
1899    Description:
1900    
1901    - changed Timer interface to what might become the spec
1902    - POSIX_FLAGS -> BIT_FLAGS according to spec
1903    - some other minor discrepancies wrt. spec eliminated
1904    
1905    ----------------------------------------------------------------------
1906    Name: Matthias Blume (blume (at) tti - c (dot) org)
1907    Date: 2003/11/06 12:00:00 CST
1908    Tag: Release_110_44
1909    Description:
1910    
1911    New working version (110.44). New bootfiles.
1912    
1913    ----------------------------------------------------------------------
1914    Name: Matthias Blume (blume (at) tti - c (dot) org)
1915    Date: 2003/11/04 11:50:00 CST
1916    Tag: blume-20031104-move-libraries
1917    Description:
1918    
1919    Eliminated the "dont_move_libraries" directive in config/targets.
1920    (The mechanism was broken and could not be fixed easily.  Moreover,
1921    there does not seem to be any reason not to move all libraries into
1922    lib during installation.  I originally implemented this directive as a
1923    backward-compatibility feature when I first introduced the new CM.
1924    Now that things have been stable for a long time and going back to the
1925    old CM is not an option, there is no reason to keep it around.)
1926    
1927    ----------------------------------------------------------------------
1928    Name: Matthias Blume (blume (at) tti - c (dot) org)
1929    Date: 2003/11/03 16:00:00 CST
1930    Tag: blume-20031103-installdir
1931    Description:
1932    
1933    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1934    Richards for pointing out the problem and providing the solution.)
1935    
1936    ----------------------------------------------------------------------
1937    Name: Matthias Blume (blume (at) tti - c (dot) org)
1938    Date: 2003/10/01 17:05:00 CDT
1939    Tag: blume-20031001-lal-mlrisc
1940    Description:
1941    
1942    MLRISC bug fix from Lal.
1943    
1944    ----------------------------------------------------------------------
1945    Name: Matthias Blume (blume (at) tti - c (dot) org)
1946    Date: 2003/09/30 16:10:00 CDT
1947    Tag: blume-20030930-primio-bat
1948    Description:
1949    
1950    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1951    2. Improved .bat files (for Win32 port) to make things work under Win95.
1952       (thanks to Aaron S. Hawley for this one)
1953    
1954    ----------------------------------------------------------------------
1955    Name: Matthias Blume (blume (at) tti - c (dot) org)
1956    Date: 2003/09/26 16:05:00 CDT
1957    Tag: blume-20030926-wrappriv
1958    Description:
1959    
1960    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1961    
1962    ----------------------------------------------------------------------
1963    Name: Matthias Blume (blume (at) tti - c (dot) org)
1964    Date: 2003/09/26 15:00:00 CDT
1965    Tag: blume-20030926-110_43_3
1966    Description:
1967    
1968    - additional cleanup
1969    - version number bump, NEW BOOTFILES
1970    
1971    ----------------------------------------------------------------------
1972    Name: Matthias Blume (blume (at) tti - c (dot) org)
1973    Date: 2003/09/26 12:00:00 CDT
1974    Tag: blume-20030926-ppautoload
1975    Description:
1976    
1977    I modified the read-eval-print loop so that the autoloader gets
1978    invoked whenever the prettyprinter tries to look up a symbol that
1979    is not currently defined in the toplevel environment but which
1980    appears in CM's autoload registry.  As a result, we see far fewer of
1981    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1982    
1983    In addition to this I tried to clean up some pieces of the Basis
1984    implementation (e.g., Socket, Word8Array) in order to prevent other
1985    instances of these ?.Foo.Bar.xxx names from being printed.
1986    
1987    The mechanism that picks names for types still needs some work, though.
1988    (Right now it seems that if there is a type A.t which is defined to
1989    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1990    "?.B.u" although the perhaps more sensible solution would be to use
1991    "A.t" in this case.  In other words, the prettyprinter should follow
1992    a chain of DEFtycs not farther than there are corresponding toplevel
1993    names in the current environment.)
1994    
1995    ----------------------------------------------------------------------
1996    Name: Matthias Blume (blume (at) tti - c (dot) org)
1997    Date: 2003/09/24 16:31:00 CDT
1998    Tag: blume-20030924-installer
1999    Description:
2000    
2001    Another installer tweak:  All the ML code for the installer is now
2002    compiled during CMB.make and put into a little library called
2003    $smlnj/installer.cm.  The installation then simply invokes
2004    
2005       sml -m $smlnj/installer.cm
2006    
2007    and everything happens automagically.
2008    
2009    Win32:  ML code senses value of environment variable SMLNJ_HOME.
2010    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
2011            and BINDIR.
2012    
2013    The new scheme guarantees that the ML code responsible for the installation
2014    is in sync with the APIs of the main system.  Also, the installer is
2015    somewhat faster because the installer script is precompiled.
2016    
2017    ----------------------------------------------------------------------
2018    Name: Matthias Blume (blume (at) tti - c (dot) org)
2019    Date: 2003/09/24 15:35:00 CDT
2020    Tag: blume-20030924-synsock
2021    Description:
2022    
2023    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
2024    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
2025    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
2026    implementations of non-blocking operations from CML's Socket
2027    structure.
2028    
2029    ----------------------------------------------------------------------
2030    Name: Matthias Blume (blume (at) tti - c (dot) org)
2031    Date: 2003/09/24 15:10:05 CDT
2032    Tag: blume-20030924-sockets
2033    Description:
2034    
2035    1. Fixed SOCKET API and implementation to match Basis spec.
2036       This required changing the internal representation of sockets to one
2037       that remembers (for each socket file descriptor) whether it is currently
2038       blocking or non-blocking.  This state is maintained lazily (i.e., a system
2039       call is made only if the state actually needs to change).
2040    
2041    2. OS-specific details of sockets were moved into separate files, thus
2042       making it possible to unify the bulk of the socket implementations
2043       between Unix and Win32.
2044    
2045    3. CML's socket API changed accordingly.
2046       (Note that we need to remove non-blocking functions from this API
2047       since they are redundant in the case of CML!)
2048    
2049    4. CML's socket implementation now makes use of non-blocking functions
2050       provided by Basis, thus removing all OS-dependent code from this part
2051       of CML.
2052    
2053    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
2054    
2055    ----------------------------------------------------------------------
2056    Name: Matthias Blume (blume (at) tti - c (dot) org)
2057    Date: 2003/09/22 12:10:00 CDT
2058    Tag: blume-20030922-110_43_2
2059    Description:
2060    
2061    Made a new interim version and bootfiles for developer's bootstrapping
2062    convenience.
2063    
2064       110.43.2 -- NEW BOOTFILES
2065    
2066    ----------------------------------------------------------------------
2067    Name: Matthias Blume (blume (at) tti - c (dot) org)
2068    Date: 2003/09/19 15:55:00 CDT
2069    Tag: blume-20030919-cmdir
2070    Description:
2071    
2072    1. new-install.sh -> install.sh
2073    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
2074    3. tweaked installer so that another name instead of .cm can be chosen
2075       at install time (by setting the CM_DIR_ARC environment variable
2076       during installation); once installation is complete, the name is
2077       fixed
2078    
2079    ----------------------------------------------------------------------
2080    Name: Matthias Blume (blume (at) tti - c (dot) org)
2081    Date: 2003/09/18 16:00:00 CDT
2082    Tag: blume-20030918-110_43_1
2083    Description:
2084    
2085    Made a new interim version and bootfiles for developer's bootstrapping
2086    convenience.
2087    
2088       110.43.1 -- NEW BOOTFILES
2089    
2090    ----------------------------------------------------------------------
2091    Name: Matthias Blume (blume (at) tti - c (dot) org)
2092    Date: 2003/09/18 15:20:00 CDT
2093    Tag: blume-20030918-misc
2094    Description:
2095    
2096    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
2097       this seems to be controversial at the moment) and used those in
2098       Posix.ProcEnv.times.
2099    
2100    2. Added Time.{from,to}Nanoseconds to Time.
2101    
2102    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
2103       For example, fromLargeInt never needs to look at more than 3 "big
2104       digits" to get its 53 bits of precision.
2105    
2106    ----------------------------------------------------------------------
2107    Name: Matthias Blume (blume (at) tti - c (dot) org)
2108    Date: 2003/09/17 16:30:00 CDT
2109    Tag: blume-20030917-real32-slices
2110    Description:
2111    
2112    Added an entry to the primitive environment
2113    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
2114    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
2115    
2116    Removed some of the "magic" constants in real64.sml and replaced them
2117    with code that generates these values from their corresponding
2118    integer counterparts.
2119    
2120    Made all(?) the slice-related changes to the Basis and made everything
2121    compile again...
2122    
2123    ----------------------------------------------------------------------
2124    Name: Matthias Blume (blume (at) tti - c (dot) org)
2125    Date: 2003/09/15 17:45:00 CDT
2126    Tag: blume-20030915-rbase
2127    Description:
2128    
2129    Fixed bug in Real.fromLargeInt.
2130    
2131    ----------------------------------------------------------------------
2132    Name: Matthias Blume (blume (at) tti - c (dot) org)
2133    Date: 2003/09/13 18:11:00 CDT
2134    Tag: blume-20030913-libinstall
2135    Description:
2136    
2137    Minor bugfix in config/libinstall (set anchor with path to
2138    standalone tool after installing it, otherwise libraries that
2139    need ml-lex or ml-yacc won't compile the first time the installer
2140    runs).
2141    
2142    ----------------------------------------------------------------------
2143    Name: Matthias Blume (blume (at) tti - c (dot) org)
2144    Date: 2003/09/12 11:45:00 CDT
2145    Tag: blume-20030912-various
2146    Description:
2147    
2148     - fixed bug in Real.toLargeInt
2149     - fixed bug in Posix.ProcEnv.times
2150     - changed inputLine functions to return an option
2151     - minor installer improvements / bugfixes
2152     - changed default @SMLalloc parameter for x86/celeron to 64k
2153    
2154    ----------------------------------------------------------------------
2155    Name: Matthias Blume (blume (at) tti - c (dot) org)
2156    Date: 2003/09/09 22:00:00 CDT
2157    Tag: Release_110_43
2158    Description:
2159    
2160        New working release 110.43.  New bootfiles.
2161    
2162    ----------------------------------------------------------------------
2163    Name: Matthias Blume (blume (at) tti - c (dot) org)
2164    Date: 2003/09/09 19:20:00 CDT
2165    Tag: blume-20030909-installer
2166    Description:
2167    
2168    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
2169    Modified config/install.bat to take advantage of it.  Also modified
2170    config/install.sh (and called it config/new-install.sh) to take advantage
2171    of it on Unix systems.  (The SML code is (supposed to be) platform-
2172    independent.)
2173    
2174    The installer can now install everything under Win32
2175    as well as under *nix as long as it compiles.
2176    
2177    Other changes:
2178    
2179      - made CML compile again under Win32
2180      - made eXene compile under Win32 (by providing a fake structure UnixSock
2181        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
2182      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
2183        same as type int;  under Win32 it isn't
2184      - fixed some slice-related problems in the win32-specific parts of CML
2185      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
2186        (under Win32, the volume name is case-insensitive, and the
2187        OS.Path code compares volume names for equality)
2188    
2189    ----------------------------------------------------------------------
2190    Name: Matthias Blume (blume (at) tti - c (dot) org)
2191    Date: 2003/09/08 11:55:00 CDT
2192    Tag: blume-20030908-fullpath
2193    Description:
2194    
2195    Made Win32 version of OS.FileSys.fullPath return current directory
2196    when given an empty string.  This is what the spec says, and incidentally,
2197    CM depends on it.  (CM otherwise goes into an infinite loop in certain
2198    cases when presented with the name of a non-existing .cm file.)
2199    
2200    ----------------------------------------------------------------------
2201    Name: Matthias Blume (blume (at) tti - c (dot) org)
2202    Date: 2003/09/04 16:30:00 CDT
2203    Tag: blume-20030905-slices-etc
2204    Description:
2205    
2206      1. Changed interface to vectors and arrays in Basis to match
2207         (draft) Basis spec.
2208      2. Added signatures and implementations of slices according to
2209         Basis spec.
2210      3. Edited source code throughout the system to make it compile again
2211         under 1. and 2.  (In some cases code had to be added to have it
2212         match the new signatures.)
2213      4. MLRISC should be backward-compatible: the copies of the originals
2214         of files that needed to change under 3. were retained, the .cm files
2215         check the compiler version number and use old versions when
2216         appropriate.
2217      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
2218         dirstream -> string option (in accordance with Basis spec).
2219      6. When generating code that counts lines, ml-lex used function
2220         CharVector.foldli, taking advantage of its old interface.
2221         This has been replaced with the corresponding code from
2222         CharVectorSlice.  (html-lex must be re-lexed!)
2223      7. BitArray in smlnj-lib/Util has been extended/modified to match the
2224         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
2225      8. Removed temporary additions (fromInternal, toInternal) from the
2226         (now obsolete) IntInf in smlnj-lib/Util.
2227      9. Cleaned up structure Byte.
2228     10. Added localOffset, scan, and fromString to Date (according to spec).
2229         Cleaned/corrected implementation of Date.
2230         (Still need to check for correctness; implement better canonicalizeDate.)
2231     11. Added "scan" to signature IEEE_REAL.
2232     12. Some improvements to IntInf [in particular: efficiency-hack for
2233         mod and rem when second operand is 2 (for parity checks).]
2234     13. Changed representation of type Time.time, using a single IntInf.int
2235         value counting microseconds.  This considerably simplified the
2236         implementation of structure Time.  We now support negative time
2237         values; scan and fromString handle signs.
2238     14. Functor PrimIO now takes two additional arguments (VectorSlice and
2239         ArraySlice).
2240    
2241    ----------------------------------------------------------------------
2242    Name: Matthias Blume (blume (at) tti - c (dot) org)
2243    Date: 2003/08/28 17:00:00 CDT
2244    Tag: blume-20030828-intinf
2245    Description:
2246    
2247    This is a major update which comes with a version number bump
2248    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
2249    and an implementation of IntInf in the Basis.
2250    
2251    There are a fairly large number of related changes and updates throughout
2252    the system:
2253    
2254    Basis:
2255            - Implemented IntInf.
2256            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
2257            - Added some missing Real64 operations, most notably Real.toLargeInt.
2258            - Added FixedInt as a synonym for Int32.
2259    
2260    compiler:
2261            * Added support for a built-in intinf type.
2262              - literals
2263              - pattern matching
2264              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
2265              - overloading on literals and operations
2266    
2267            This required adding a primitive type intinf, some additional
2268            primops, and implementations for several non-trivial intinf
2269            operations in Core.  (The intinf type is completely abstract
2270            to the compiler; all operations get delegated back to the Core.)
2271    
2272            * Intinf equality is handled by polyequal.  However, the compiler
2273            does not print its usual warning in this case (since polyequal
2274            is the right thing to do there).
2275    
2276            * Improved the organization of structure InlineT.
2277    
2278            * A word about conversion primops:
2279            If conversions involving intinf do not cancel out during
2280            CPS contract, then the compiler must insert calls to Core functions.
2281            Since all core access must be resolved already during the FLINT
2282            translate phase, it would be too late a the time of CPS contract
2283            to add new Core calls.  For this reason, conversion primops
2284            for intinf carry two arguments:  1. the numeric argument that
2285            they are supposed to convert, and 2. the Core function that
2286            can help with this conversion if necessary.  If CPS contract
2287            eliminates a primop, then the associated Core function becomes
2288            dead and goes away.  Intinf conversion primops that do not get
2289            eliminated by CPS contract get rewritten into calls of their
2290            core functions by a separate, new phase.
2291    
2292    interactive system:
2293            - Control.Print.intinfDepth controls max length of intinf constants
2294            being printed. (Analogous to Control.Print.stringDepth.)
2295            - Cleanup in printutil and pputil: got rid of unused stuff and
2296            duplicates; replaced some of the code with code that makes better
2297            use of library functionality.
2298    
2299    CM:
2300            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
2301            are no longer silent.
2302    
2303    CKIT:
2304            Fixed mismatched uses of Int32 and LargeInt.  I always decided
2305            in favor of LargeInt -- which is now the same as IntInf.
2306            CKIT-knowledgable people should check whether this is what's
2307            intended and otherwise change things back to using Int32 or
2308            FixedInt.
2309    
2310    Throughout the code:
2311            Started using IntInf.int literals and built-in operations
2312            (e.g., comparison with 0) where this seems appropriate.
2313    
2314    
2315    ----------------------------------------------------------------------
2316    Name: Dave MacQueen (dbm@cs.uchicago.edu)
2317    Date: 2003/08/13 11:36:00 CDT
2318    Tag: dbm-20030813-mcz-merge1
2319    Description:
2320    
2321    Merging changes from the mcz-branch development branch into trunk.
2322    These changes involve replacement of the emulated old prettyprinter
2323    interface with direct use of the SML/NJ Lib PP library, and fixing
2324    of a couple of bugs (895, 1186) relating to error messages.  A new
2325    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
2326    has been added.
2327    
2328    ----------------------------------------------------------------------
2329    Name: Matthias Blume (blume (at) tti - c (dot) org)
2330    Date: 2003/08/11 15:45:00 CDT
2331    Tag: blume-20030811-windows
2332    Description:
2333    
2334    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
2335    
2336        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
2337    
2338    This patch restores SML/NJ's ability to run under win32.  There are a
2339    number of changes, including fixes for several bugs that had gone
2340    unnoticed until now:
2341    
2342      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
2343        trying to run the win32 version from within cygwin.)
2344    
2345      - There are a number of simple .bat scripts that substitute for their
2346        corresponding Unix shell-scripts.  (See below.)
2347    
2348      - The internals of ml-build have been modified slightly.  The main
2349        difference is that instead of calling ".link-sml" (or link-sml.bat)
2350        using OS.Process.system, the ML process delegates this task back
2351        to the script.  Otherwise problems arise in mixed environments such
2352        as Cygwin where scripts look and work like Unix scripts, but
2353        where OS.Process.system cannot run them.
2354    
2355      - In CM, the srcpath pickler used native pathname syntax -- which
2356        is incorrect in the case of cross-compilation.  The new pickle format
2357        is independent of platform-specific naming conventions.
2358    
2359      - Path configuration files (such as lib/pathconfig) can now choose
2360        between native and standard syntax.  Placing a line of the form
2361    
2362            standard!
2363    
2364        into the file causes all subsequent paths to be interpreted using
2365        CM standard pathname syntax (= Unix conventions); a line
2366    
2367            native!
2368    
2369        switches back to native style.  This was needed so that
2370        path config files can be written portably, see src/system/pathconfig.
2371    
2372      - Runtime system:
2373    
2374          - win32-filesys.c:  get_file_time and set_file_time now
2375            access modification time, not creation time.
2376    
2377          - I/O code made aware of new array representation.
2378    
2379          - Bug fixes in X86.prim.masm.
2380    
2381       - src/system/makeml made aware of win32. (For use under cygwin
2382            and other Unix-environments for windows.)
2383    
2384       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
2385         which caused BinIO.inputAll to fail consistently.
2386    
2387    .bat scripts:
2388    
2389       Windows .bat scripts assume that SMLNJ_HOME is defined.
2390    
2391       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
2392         applications (sml, ml-yacc, ml-lex).
2393       - ml-build.bat: analogous to ml-build.
2394       - config\install.bat:  Analogous to config/install.sh.  This requires
2395         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2396         (nmake etc. must be on the path, and vcvars32 must have been run.)
2397         Moreover, sources for ml-lex and ml-yacc need to exist under src,
2398         and the bootfile hierarchy must have been unpacked under
2399         sml.boot.x86-win32.
2400         The script is very primitive and does a poor job at error checking.
2401         It only installs the base system, ml-lex, and ml-yacc.  No other
2402         libraries are being installed (i.e., you get only those that
2403         are part of the compiler.)
2404       - link-sml.bat:  analogous to .link-sml, but not currently used
2405    
2406    Unrelated bug fixes:
2407    
2408      - ml-nlffigen now exports structures ST_* corresponding to incomplete
2409        types.
2410      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
2411        otherwise.)
2412    
2413    ----------------------------------------------------------------------
2414    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2415    Date: 2003/06/17
2416    Tag: macqueen-20030617-bug895
2417    Description:
2418    
2419    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2420    Tag will be used for new development branch (mcz-branch) for
2421    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2422    
2423    ----------------------------------------------------------------------
2424    Name: Matthias Blume (blume (at) tti - c (dot) org)
2425    Date: 2003/05/27 16:55:00 CDT
2426    Tag: blume-20030527-polyeq
2427    Description:
2428    
2429    Tried to eliminated most cases of polymorphic equality.
2430    
2431    ----------------------------------------------------------------------
2432    Name: Matthias Blume (blume (at) tti - c (dot) org)
2433    Date: 2003/05/21 17:45:00 CDT
2434    Tag: blume-20030517-complete
2435    Description:
2436    
2437    Two changes:
2438    
2439      1. Added a flag for controlling whether non-exhaustive bindings will
2440         be treated as errors (default is false).
2441      2. Cleaned up the *entire* source tree so that CMB.make goes through
2442         without a single non-exhaustive match- or bind warning.
2443    
2444    ----------------------------------------------------------------------
2445    Name: Matthias Blume (blume (at) tti - c (dot) org)
2446    Date: 2003/05/17 10:20:00 CDT
2447    Tag: blume-20030517-absyn
2448    Description:
2449    
2450    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2451    
2452       This mainly affects the quality of error messages.  However, some
2453       of the code is now more straightforward than before. (Treatment of
2454       the above four constructs in translate.sml is much simpler than
2455       the "macro-expansion" that was going on before.  Plus, the mach-
2456       compiler no longer gets invoked just to be able to compile an
2457       if-expression.)
2458    
2459    2. The ErrorMsg.Error exception is now caught and absorbed by the
2460       interactive loop.
2461    
2462    ----------------------------------------------------------------------
2463    Name: Allen Leung
2464    Date: 2003/05/16 13:05:00 CDT
2465    Tag: leunga-20030516-cygwin-runtime
2466    Description:
2467    
2468        Ported the runtime system to cygwin, which uses the unix
2469    x86-unix bin files.  Missing/buggy features:
2470    
2471       o getnetbyname, getnetbyaddr: these functions seem to be missing in
2472         the Cygwin library.
2473       o Ctrl-C handling may be flaky.
2474       o Windows system calls and Windows I/O are not supported.
2475    
2476       A new set of binfiles is located at:
2477    
2478         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2479    
2480       This is only needed for bootstrapping the cygwin version of smlnj.
2481       Other x86 versions can use the existing binfiles.
2482    
2483    ----------------------------------------------------------------------
2484    Name: Matthias Blume
2485    Date: 2003/04/08 15:42:00 CDT
2486    Tag: blume-20030408-listpair
2487    Description:
2488    
2489    1. Added a target 'mlrisc' to installer.
2490    
2491    2. Added missing elements to structure ListPair.
2492    
2493    ----------------------------------------------------------------------
2494    Name: Allen Leung
2495    Date: 2003/01/07 10:40:00 EST
2496    Tag: leunga-20030107-int-rem
2497    Description:
2498    
2499    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2500    The arguments to the SUBL instruction were swapped.
2501    
2502    ----------------------------------------------------------------------
2503    Name: Matthias Blume
2504    Date: 2002/12/12 16:25:00 EST
2505    Tag: blume-20021212-risc-ra
2506    Description:
2507    
2508    Fixed a serious bug in the rewrite code for FP spilling/reloading that
2509    sent the RA into an infinite loop when floating point registers get
2510    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2511    110.37 and 110.38.)
2512    There was another set of potential problems related to the handling of
2513    MLRISC annotations (but those did not yet cause real problems, apparently).
2514    
2515    ----------------------------------------------------------------------
2516    Name: Matthias Blume
2517    Date: 2002/12/06 22:40:00 EST
2518    Tag: blume-20021206-cm-fileid
2519    Description:
2520    
2521    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2522    This fixes the problem of CM getting confused by files that suddenly
2523    change their identity (e.g., by getting unlinked and recreated by some
2524    text editor such as vi).  There might be a better/cheaper/cleaner way
2525    of doing this, but for now this will have to do.
2526    
2527    ----------------------------------------------------------------------
2528    Name: Matthias Blume
2529    Date: 2002/10/28 09:50:00 EST
2530    Tag: blume-20021028-typecheck
2531    Description:
2532    
2533    Exported structure Typecheck from $smlnj/viscomp/core.cm.
2534    
2535    ----------------------------------------------------------------------
2536    Name: Matthias Blume
2537    Date: 2002/10/17 09:10:00 EDT
2538    Tag: Release_110_42
2539    Description:
2540    
2541    In good old tradition, there has been a slight hiccup so that we have
2542    to patch 110.42 after the fact.   The old release tag has been replaced
2543    (see below).
2544    
2545    The change solves a problem with two competing approaches the
2546    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2547    each other's way.
2548    
2549    This change only affects the runtime system code and the installer script.
2550    (No new bootfiles.)
2551    
2552    ----------------------------------------------------------------------
2553    Name: Matthias Blume
2554    Date: 2002/10/16 12:00:00 EDT
2555    Tag: Release_110_42_removed
2556    Description:
2557    
2558    New working release.  New bootfiles.
2559    
2560    ----------------------------------------------------------------------
2561    Name: Matthias Blume
2562    Date: 2002/10/10 13:10:00 EDT
2563    Tag: blume-20021010-ppc-divs
2564    Description:
2565    
2566    The mltree operator DIVS must be implemented with an overflow check on
2567    the PPC because the hardware indicates divide-by-zero using "overflow" as
2568    well.
2569    
2570    ----------------------------------------------------------------------
2571    Name: Matthias Blume
2572    Date: 2002/07/23 11:45:00
2573    Tag: blume-20020723-smlnj-home
2574    Description:
2575    
2576    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2577    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2578    CM_PATHCONFIG is also set), the path configuration file is assumed
2579    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2580    the entire tree to some other place and everything will "just work".
2581    
2582    (Companion commands such as ml-build and ml-makedepend also sense this
2583    variable.)
2584    
2585    ----------------------------------------------------------------------
2586    Name: Matthias Blume
2587    Date: 2002/07/12 21:19:00 EDT
2588    Tag: blume-20020712-liveness
2589    Description:
2590    
2591    Exported two useful "step" functions from liveness module (MLRISC).
2592    
2593    ----------------------------------------------------------------------
2594    Name: Matthias Blume
2595    Date: 2002/07/05 16:00 EDT
2596    Tag: Release_110_41
2597    Description:
2598    
2599    New working release.  New bootfiles.
2600    
2601    ----------------------------------------------------------------------
2602    Name: Matthias Blume
2603    Date: 2002/07/05 10:25:00 EDT
2604    Tag: blume-20020705-btimp
2605    Description:
2606    
2607    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2608    other clients can set up backtracing support.
2609    
2610    ----------------------------------------------------------------------
2611    Name: Matthias Blume
2612    Date: 2002/06/25 17:23:00 EDT
2613    Tag: blume-20020625-fpmax
2614    Description:
2615    
2616    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2617    case.  (The sense of the isNaN test was reversed -- which made min and
2618    max always return their first argument.)
2619    
2620    ----------------------------------------------------------------------
2621    Name: Matthias Blume
2622    Date: 2002/06/11
2623    Tag: blume-20020611-unixpath
2624    Description:
2625    
2626    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2627    
2628    ----------------------------------------------------------------------
2629    Name: Matthias Blume
2630    Date: 2002/06/10 16:35:00 EDT
2631    Tag: blume-20020610-ieeereal
2632    Description:
2633    
2634    I back-ported my implementation of IEEEReal.fromString from the
2635    idlbasis-devel branch so that we can test it.
2636    
2637    Another small change is that ppDec tries to give more information
2638    than just "<sig>" in the case of functors.  However, this code is
2639    broken in some mysterious way if the functor's body's signature
2640    has not been declared by ascription but gets inferred from the
2641    implementation.  This needs fixing...
2642    
2643    ----------------------------------------------------------------------
2644    Name: Matthias Blume
2645    Date: 2002/05/31
2646    Tag: blume-20020531-btrace-mode
2647    Description:
2648    
2649    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2650    the wayside when I switched over to using Controls everywhere.)
2651    
2652    ----------------------------------------------------------------------
2653    Name: Lal George
2654    Date: 2002/05/23 12:21:40 EDT
2655    Tag: george-20020523-visual-labels
2656    Description:
2657    
2658    Labels are now displayed in the graphical output to make
2659    the fall-through and target blocks obvious.
2660    
2661    ----------------------------------------------------------------------
2662    Name: Matthias Blume
2663    Date: 2002/05/22 11:03:00 EDT
2664    Tag: blume-20020522-shrink
2665    Description:
2666    
2667    John tweaked yesterday's fix for 1131 to handle an out-of-memory
2668    situation that comes up when allocating huge arrays.
2669    
2670    ----------------------------------------------------------------------
2671    Name: Matthias Blume
2672    Date: 2002/05/21 16:00:00 EDT
2673    Tag: Release_110_40
2674    Description:
2675    
2676    New working release (110.40).  New bootfiles.
2677    
2678    [Also: John Reppy fixed GC bug 1131.]
2679    
2680    ----------------------------------------------------------------------
2681    Name: Matthias Blume
2682    Date: 2002/05/21 12:35:00 EDT
2683    Tag: blume-20020521-cmdoc
2684    Description:
2685    
2686    CM documentation update.
2687    
2688    ----------------------------------------------------------------------
2689    Name: Matthias Blume
2690    Date: 2002/05/21 10:55:00 EDT
2691    Tag: blume-20020521-misc
2692    Description:
2693    
2694    - John tweaked runtime to be silent on heap export (except when
2695      GC messages are on).
2696    - I added a few more things (cross-compiling versions of CMB) to
2697      config/preloads (as suggestions).
2698    
2699    ----------------------------------------------------------------------
2700    Name: Matthias Blume
2701    Date: 2002/05/20 22:25:00 EDT
2702    Tag: blume-20020520-controls
2703    Description:
2704    
2705    - Added ControlUtil structure to control-lib.cm.
2706    - Use it throughout.
2707    - Used Controls facility to define MLRISC controls (as opposed to
2708      registering MLRISC control ref cells with Controls after the fact)
2709    - Fixed messed-up controls priorities.
2710    
2711    * Removed again all the stuff from config/preloads that one wouldn't
2712      be able to preload at the time the initial heap image is built.
2713      (Many libraries, e.g., CML, do not exist yet at this time.  The
2714       only libraries that can be preloaded via config/preloads are those
2715       that come bundled with the bootfiles.)
2716    
2717    ----------------------------------------------------------------------
2718    Name: Matthias Blume
2719    Date: 2002/05/20 10:59:00 EDT
2720    Tag: blume-20020520-preloads
2721    Description:
2722    
2723    Added a lot of commented-out suggestions for things to be included
2724    in config/preloads.
2725    
2726    ----------------------------------------------------------------------
2727    Name: Allen Leung
2728    Date: 2002/05/18 14:20:00 EDT
2729    Tag: leunga-20020518-mdl
2730    Description:
2731    
2732    o Made the mdl tool stuff compile and run again.
2733    o I've disabled all the stuff that depends on RTL specifications; they
2734      are all badly broken anyway.
2735    
2736    ----------------------------------------------------------------------
2737    Name: Matthias Blume
2738    Date: 2002/05/17 16:49:00 EDT
2739    Tag: blume-20020517-controls
2740    Description:
2741    
2742    0. John Reppy made several modifications to the SML/NJ library.
2743       In particular, there is a shiny new controls-lib.cm.
2744    
2745    1. Pushed new controls interface through compiler so that everything
2746       compiles again.
2747    
2748    2. Added FormatComb and FORMAT_COMB to the CML version of the
2749       SML/NJ library (so that CML compiles again).
2750    
2751    3. Modified init scripts because XXX_DEFAULT environment variables
2752       are no longer with us.  (Boot-time initialization is now done
2753       using the same environment variables that are also used for
2754       startup-time initialization of controls.)
2755    
2756    ----------------------------------------------------------------------
2757    Name: Lal George
2758    Date: 2002/05/15 09:20:10 EDT
2759    Tag: george-20020515-pseudo-op-decls
2760    Description:
2761    
2762            All pseudo-ops emitted before the first segment declaration
2763    such as TEXT, DATA, and BSS directives are assumed to be global
2764    declarations and are emitted first in the assembly file. This is
2765    useful in a number of situations where one has pseudo-ops that are not
2766    specific to any segment, and also works around the constraint that one
2767    cannot have client pseudo-ops in the TEXT segment.
2768    
2769            Because no segment is associated with these declarations it is
2770    an error to allocate any space or objects before the first segment
2771    directive and an exception will be raised. However, we cannot make
2772    this check for client pseudo-ops.
2773    
2774            These top level declarations are a field in the CFG graph_info.
2775    In theory you can continue to add to this field after the CFG has been
2776    built -- provided you know what you are doing;-)
2777    
2778    ----------------------------------------------------------------------
2779    Name: Matthias Blume
2780    Date: 2002/05/13 16:40:00 EDT
2781    Tag: blume-20020513-pp-etc
2782    Description:
2783    
2784    A few minor bugfixes:
2785    
2786      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2787        (Bogus printouts for redefined bindings still occur.  Compiler
2788        bug should no longer occur now.  We need to redo the prettyprinter
2789        from scratch.)
2790    
2791      - CM pathname printer now also adds escape sequences for ( and )
2792    
2793      - commend and docu fixes for ml-nlffi
2794    
2795    ----------------------------------------------------------------------
2796    Name: Matthias Blume
2797    Date: 2002/05/10 16:40:00 EDT
2798    Tag: blume-20020510-erg-textio
2799    Description:
2800    
2801    Applied the following bugfix provided by Emden Gansner:
2802    
2803        Output is corrupted when outputSubstr is used rather than output.
2804    
2805        The problem occurs when a substring
2806    
2807            ss = (s, dataStart, dataLen)
2808    
2809        where dataStart > 0, fills a stream buffer with avail bytes left.
2810        avail bytes of s, starting at index dataStart, are copied into the
2811        buffer, the buffer is flushed, and then the remaining dataLen-avail
2812        bytes of ss are copied into the beginning of the buffer. Instead of
2813        starting this copy at index dataStart+avail in s, the current code
2814        starts the copy at index avail.
2815    
2816        Fix:
2817        In text-io-fn.sml, change line 695 from
2818             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2819        to
2820             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2821    
2822    ----------------------------------------------------------------------
2823    Name: Matthias Blume
2824    Date: 2002/04/12 13:55:00 EDT
2825    Tag: blume-20020412-assyntax
2826    Description:
2827    
2828    1. Grabbed newer assyntax.h from the XFree86 project.
2829    2. Fiddled with how to compile X86.prim.asm without warnings.
2830    3. (Very) Minor cleanup in CM.
2831    
2832    ----------------------------------------------------------------------
2833    Name: Matthias Blume
2834    Date: 2002/04/01 (no joke!) 17:07:00 EST
2835    Tag: blume-20020401-x86div
2836    Description:
2837    
2838    Added full support for div/mod/rem/quot on the x86, using the machine
2839    instruction's two results (without clumsily recomputing the remainder)
2840    directly where appropriate.
2841    
2842    Some more extensive power-of-two support was added to the x86 instruction
2843    selector (avoiding expensive divs, mods, and muls where they can be
2844    replaced with cheaper shifts and masks).  However, this sort of thing
2845    ought to be done earlier, e.g., within the CPS optimizer so that
2846    all architectures benefit from it.
2847    
2848    The compiler compiles to a fixed point, but changes might be somewhat
2849    fragile nevertheless.  Please, report any strange things that you might
2850    see wrt. div/mod/quot/rem...
2851    
2852    ----------------------------------------------------------------------
2853    Name: Matthias Blume
2854    Date: 2002/03/29 17:22:00
2855    Tag: blume-20020329-div
2856    Description:
2857    
2858    Fixed my broken div/mod logic.  Unfortunately, this means that the
2859    inline code for div/mod now has one more comparison than before.
2860    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2861    The problem was with quotient = 0, because that alone does not tell
2862    us which way the rounding went.  One then has to look at whether
2863    remainder and divisor have the same sign...  :(
2864    
2865    Anyway, I replaced the bootfiles with fresh ones...
2866    
2867    ----------------------------------------------------------------------
2868    Name: Matthias Blume
2869    Date: 2002/03/29 14:10:00 EST
2870    Tag: blume-20020329-inlprims
2871    Description:
2872    
2873    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2874    
2875    Primops have changed. This means that the bin/boot-file formats have
2876    changed as well.
2877    
2878    To make sure that there is no confusion, I made a new version.
2879    
2880    
2881    CHANGES:
2882    
2883    * removed REMT from mltree (remainder should never overflow).
2884    
2885    * added primops to deal with divisions of all flavors to the frontend
2886    
2887    * handled these primops all the way through so they map to their respective
2888      MLRISC support
2889    
2890    * used these primops in the implementation of Int, Int32, Word, Word32
2891    
2892    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2893    
2894    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2895    
2896    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2897      floating point
2898    
2899    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2900      and Real (but Real.abs maps to a separate floating-point-only primop)
2901    
2902    
2903    TODO items:
2904    
2905    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2906      instructions because the machine instruction encoder cannot handle
2907      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2908      REM in the first place, and REM is supported by the x86 machine coder.)
2909    
2910    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2911      selection phase.  (The two can be streamlined because the hardware
2912      delivers both quotient and remainder at the same time anyway.)
2913    
2914    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2915      (Currently the behavior is inconsistent both across architectures and
2916      wrt. the draft Basis spec.)
2917    
2918    * Word8 should eventually be handled natively, too.
2919    
2920    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2921      as if there currently is no execution path that could trigger it in
2922      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2923      hold for things like multiplication and division.)
2924    
2925    ----------------------------------------------------------------------
2926    Name: Matthias Blume
2927    Date: 2002/03/27 16:27:00 EST
2928    Tag: blume-20020327-mlrisc-divisions
2929    Description:
2930    
2931    Added support for all four division operations (ML's div, mod, quot,
2932    and rem) to MLRISC.  In the course of doing so, I also rationalized
2933    the naming (no more annoying switch-around of DIV and QUOT), by
2934    parameterizing the operation by div_rounding_mode (which can be either
2935    DIV_TO_ZERO or DIV_TO_NEGINF).
2936    
2937    The generic MLTreeGen functor takes care of compiling all four
2938    operations down to only round-to-zero div.
2939    
2940    Missing pieces:
2941    
2942      * Doing something smarter than relying on MLTreeGen on architectures
2943        like, e.g., the x86 where hardware division delivers both quotient and
2944        remainder at the same time.  With this, the implementation of the
2945        round-to-neginf operations could be further streamlined.
2946    
2947      * Remove inlining support for div/mod/rem from the frontend and replace it
2948        with primops that get carried through to the backend.  Do this for all
2949        int and word types.
2950    
2951    ----------------------------------------------------------------------
2952    Name: Matthias Blume
2953    Date: 2002/03/25 17:25:00 EST
2954    Tag: blume-20020325-divmod
2955    Description:
2956    
2957    I improved (hopefully without breaking them) the implementation of Int.div,
2958    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2959    advantage of the following observations:
2960    
2961      Let  q = x quot y      r = x rem y
2962           d = x div  y      m = x mod y
2963    
2964    where "quot" is the round-to-zero version of integer division that
2965    hardware usually provides.  Then we have:
2966    
2967         r = x - q * y        where neither the * nor the - will overflow
2968         d = if q >= 0 orelse x = q * y then q else q - 1
2969                              where neither the * nor the - will overflow
2970         m = if q >= 0 orelse r = 0 then r else r + y
2971                              where the + will not overflow
2972    
2973    This results in substantial simplification of the generated code.
2974    The following table shows the number of CFG nodes and edges generated
2975    for
2976            fun f (x, y) = x OPER y
2977            (* with OPER \in div, mod, quot, rem *)
2978    
2979    
2980        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2981        --------------------------------------------------------
2982         div |         24 |         39 |         12 |         16
2983         mod |         41 |         71 |         12 |         16
2984        quot |          8 |         10 |          8 |         10
2985         rem |         10 |         14 |          8 |         10
2986    
2987    
2988    ----------------------------------------------------------------------
2989    Name: Matthias Blume
2990    Date: 2002/03/25 22:06:00 EST
2991    Tag: blume-20020325-cprotobug
2992    Description:
2993    
2994    Fixed a bug in cproto (c prototype decoder).
2995    
2996    ----------------------------------------------------------------------
2997    Name: Matthias Blume
2998    Date: 2002/03/25 16:00:00 EST
2999    Tag: blume-20020325-raw-primops
3000    Description:
3001    
3002    I did some cleanup to Allen's new primop code and
3003    replaced yesterday's bootfiles with new ones.
3004    (But they are stored in the same place.)
3005    
3006    ----------------------------------------------------------------------
3007    Name: Matthias Blume
3008    Date: 2002/03/24 22:40:00 EST
3009    Tag: blume-20020324-bootfiles
3010    Description:
3011    
3012    Made the bootfiles that Allen asked for.
3013    
3014    ----------------------------------------------------------------------
3015    Name: Allen Leung
3016    Date: 2002/03/23 15:50:00 EST
3017    Tag: leunga-20020323-flint-cps-rcc-primops
3018    Description:
3019    
3020      1. Changes to FLINT primops:
3021    
3022        (* make a call to a C-function;
3023         * The primop carries C function prototype information and specifies
3024         * which of its (ML-) arguments are floating point. C prototype
3025         * information is for use by the backend, ML information is for
3026         * use by the CPS converter. *)
3027      | RAW_CCALL of { c_proto: CTypes.c_proto,
3028                       ml_args: ccall_type list,
3029                       ml_res_opt: ccall_type option,
3030                       reentrant : bool
3031                     } option
3032       (* Allocate uninitialized storage on the heap.
3033        * The record is meant to hold short-lived C objects, i.e., they
3034        * are not ML pointers.  With the tag, the representation is
3035        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
3036        *)
3037      | RAW_RECORD of {tag:bool,sz:int}
3038      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
3039    
3040      2.  These CPS primops are now overloaded:
3041    
3042           rawload of {kind:numkind}
3043           rawstore of {kind:numkind}
3044    
3045          The one argument form is:
3046    
3047             rawload {kind} address
3048    
3049          The two argument form is:
3050    
3051             rawload {kind} [ml object, byte-offset]
3052    
3053      3. RAW_CCALL/RCC now takes two extra arguments:
3054    
3055         a. The first is whether the C call is reentrant, i.e., whether
3056            ML state should be saved and restored.
3057         b. The second argument is a string argument specifying the name of
3058            library and the C function.
3059    
3060         These things are currently not handled in the code generator, yet.
3061    
3062      4. In CProto,
3063    
3064         An encoding type of "bool" means "ml object" and is mapped into
3065         C prototype of PTR.  Note that "bool" is different than "string",
3066         even though "string" is also mapped into PTR, because "bool"
3067         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
3068    
3069      5. Pickler/unpicker
3070    
3071         Changed to handle RAW_RECORD and newest RAW_CCALL
3072    
3073      6. MLRiscGen,
3074    
3075         1. Changed to handle the new rawload/rawstore/rawrecord operators.
3076         2. Code for handling C Calls has been moved to a new module CPSCCalls,
3077            in the file CodeGen/cpscompile/cps-c-calls.sml
3078    
3079      7. Added the conditional move operator
3080    
3081             condmove of branch
3082    
3083         to cps.  Generation of this is still buggy so it is currently
3084         disabled.
3085    
3086    ----------------------------------------------------------------------
3087    Name: Lal George
3088    Date: 2002/03/22 14:18:25 EST
3089    Tag: george-20020322-cps-branch-prob
3090    Description:
3091    
3092    Implemented the Ball-Larus branch prediction-heuristics, and
3093    incorporated graphical viewers for control flow graphs.
3094    
3095    Ball-Larus Heuristics:
3096    ---------------------
3097    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
3098    
3099    By design it uses the Dempster-Shafer theory for combining
3100    probabilities.  For example, in the function:
3101    
3102        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
3103    
3104    the ball-larus heuristics predicts that the n=0 is unlikely
3105    (OH-heuristic), and the 'then' branch is unlikely because of the
3106    RH-heuristic -- giving the 'then' branch an even lower combined
3107    probability using the Dempster-Shafer theory.
3108    
3109    Finally, John Reppy's loop analysis in MLRISC, further lowers the
3110    probability of the 'then' branch because of the loop in the else
3111    branch.
3112    
3113    
3114    Graphical Viewing:
3115    ------------------
3116    I merely plugged in Allen's graphical viewers into the compiler. The
3117    additional code is not much. At the top level, saying:
3118    
3119            Control.MLRISC.getFlag "cfg-graphical-view" := true;
3120    
3121    will display the graphical view of the control flow graph just before
3122    back-patching.  daVinci must be in your path for this to work. If
3123    daVinci is not available, then the default viewer can be changed
3124    using:
3125    
3126            Control.MLRISC.getString "viewer"
3127    
3128    which can be set to "dot" or "vcg" for the corresponding viewers. Of
3129    course, these viewers must be in your path.
3130    
3131    The above will display the compilation unit at the level of clusters,
3132    many of which are small, boring, and un-interesting. Also setting:
3133    
3134            Control.MLRISC.getInt "cfg-graphical-view_size"
3135    
3136    will display clusters that are larger than the value set by the above.
3137    
3138    
3139    ----------------------------------------------------------------------
3140    Name: Matthias Blume
3141    Date: 2002/03/21 22:20:00 EST
3142    Tag: blume-20020321-kmp-bugfix
3143    Description:
3144    
3145    Changed the interface to the KMP routine in PreString and fixed
3146    a minor bug in one place where it was used.
3147    
3148    ----------------------------------------------------------------------
3149    Name: Allen Leung
3150    Date: 2002/03/21 20:30:00 EST
3151    Tag: leunga-20020321-cfg
3152    Description:
3153    
3154      Fixed a potential problem in cfg edge splitting.
3155    
3156    ----------------------------------------------------------------------
3157    Name: Allen Leung
3158    Date: 2002/03/21 17:15:00 EST
3159    Tag: leunga-20020321-x86-fp-cfg
3160    Description:
3161    
3162      1. Recoded the buggy parts of x86-fp.
3163    
3164         a. All the block reordering code has been removed.
3165            We now depend on the block placement phases to do this work.
3166    
3167         b. Critical edge splitting code has been simplified and moved into the
3168            CFG modules, as where they belong.
3169    
3170         Both of these were quite buggy and complex.  The code is now much, much
3171         simpler.
3172    
3173      2. X86 backend.
3174    
3175         a. Added instructions for 64-bit support.  Instruction selection for
3176            64-bit has not been committed, however, since that
3177            requires changes to MLTREE which haven't been approved by
3178            Lal and John.
3179    
3180         b. Added support for FUCOMI and FUCOMIP when generating code for
3181            PentiumPro and above.  We only generate these instructions in
3182            the fast-fp mode.
3183    
3184         c. Added cases for JP and JNP in X86FreqProps.
3185    
3186      3. CFG
3187    
3188         CFG now has a bunch of methods for edge splitting and merging.
3189    
3190      4. Machine description.
3191    
3192         John's simplification of MLTREE_BASIS.fcond broke a few machine
3193         description things:
3194    
3195         rtl-build.{sig,sml} and hppa.mdl fixed.
3196    
3197         NOTE: the machine description stuff in the repository is still broken.
3198               Again, I can't put my fixes in because that involves
3199               changes to MLTREE.
3200    
3201    ----------------------------------------------------------------------
3202    Name: Matthias Blume
3203    Date: 2002/03/20 15:55:00 EST
3204    Tag: blume-20020320-kmp
3205    Description:
3206    
3207    Implemented Knuth-Morris-Pratt string matching in PreString and used
3208    it for String.isSubstring, Substring.isSubstring, and
3209    Substring.position.
3210    
3211    (Might need some stress-testing.  Simple examples worked fine.)
3212    
3213    ----------------------------------------------------------------------
3214    Name: Matthias Blume
3215    Date: 2002/03/19 16:37:00 EST
3216    Tag: blume-20020319-witnesses
3217    Description:
3218    
3219    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
3220    
3221    This implements a generic mechanism for changing constness qualifiers
3222    anywhere within big C types without resorting to outright "casts".
3223    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
3224    modify the constness at the outermost level.)
3225    The implementation of "convert" is based on the idea of "witness"
3226    values -- values that are not used by the operation but whose types
3227    "testify" to their applicability.  On the implementation side, "convert"
3228    is simply a projection (returning its second curried argument).  With
3229    cross-module inlining, it should not result in any machine code being
3230    generated.
3231    
3232    ----------------------------------------------------------------------
3233    Name: Matthias Blume
3234    Date: 2002/03/15 16:40:00 EST
3235    Tag: blume-20020315-basis
3236    Description:
3237    
3238    Provided (preliminary?) implementations for
3239    
3240      {String,Substring}.{concatWith,isSuffix,isSubstring}
3241    
3242    and
3243    
3244      Substring.full
3245    
3246    Those are in the Basis spec but they were missing in SML/NJ.
3247    
3248    ----------------------------------------------------------------------
3249    Name: Matthias Blume
3250    Date: 2002/03/14 21:30:00 EST
3251    Tag: blume-20020314-controls
3252    Description:
3253    
3254    Controls:
3255    ---------
3256    
3257    1. Factored out the recently-added Controls : CONTROLS stuff and put
3258       it into its own library $/controls-lib.cm.  The source tree for
3259       this is under src/smlnj-lib/Controls.
3260    
3261    2. Changed the names of types and functions in this interface, so they
3262       make a bit more "sense":
3263    
3264          module -> registry
3265          'a registry -> 'a group
3266    
3267    3. The interface now deals in ref cells only.  The getter/setter interface
3268       is (mostly) gone.
3269    
3270    4. Added a function that lets one register an already-existing ref cell.
3271    
3272    5. Made the corresponding modifications to the rest of the code so that
3273       everything compiles again.
3274    
3275    6. Changed the implementation of Controls.MLRISC back to something closer
3276       to the original.  In particular, this module (and therefore MLRISC)
3277       does not depend on Controls.  There now is some link-time code in
3278       int-sys.sml that registers the MLRISC controls with the Controls
3279       module.
3280    
3281    CM:
3282    ---
3283    
3284      * One can now specify the lambda-split aggressiveness in init.cmi.
3285    
3286    ----------------------------------------------------------------------
3287    Name: Allen Leung
3288    Date: 2002/03/13 17:30:00 EST
3289    Tag: leunga-20020313-x86-fp-unary
3290    Description:
3291    
3292    Bug fix for:
3293    
3294    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
3295    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
3296    > - fun f(x,(y,z)) = Real.~ y;
3297    > [autoloading]
3298    > [autoloading done]
3299    >       fchsl   (%eax), 184(%esp)
3300    > Error: MLRisc bug: X86MCEmitter.emitInstr
3301    >
3302    > uncaught exception Error
3303    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
3304    
3305    The problem was that the code generator did not generate any fp registers
3306    in this case, and the ra didn't know that it needed to run the X86FP phase to
3307    translate the pseudo fp instruction.   This only happened with unary fp
3308    operators in certain situations.
3309    
3310    ----------------------------------------------------------------------
3311    Name: Matthias Blume
3312    Date: 2002/03/13 14:00:00 EST
3313    Tag: blume-20020313-overload-etc
3314    Description:
3315    
3316    1. Added _overload as a synonym for overload for backward compatibility.
3317       (Control.overloadKW must be true for either version to be accepted.)
3318    
3319    2. Fixed bug in install script that caused more things to be installed
3320       than what was requested in config/targets.
3321    
3322    3. Made CM aware of the (_)overload construct so that autoloading
3323       works.
3324    
3325    ----------------------------------------------------------------------
3326    Name: Matthias Blume
3327    Date: 2002/03/12 22:03:00 EST
3328    Tag: blume-20020312-url
3329    Description:
3330    
3331    Forgot to update BOOT and srcarchiveurl.
3332    
3333    ----------------------------------------------------------------------
3334    Name: Matthias Blume
3335    Date: 2002/03/12 17:30:00 EST
3336    Tag: blume-20020312-version110392
3337    Description:
3338    
3339    Yet another version number bump (because of small changes to the
3340    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
3341    
3342    Changes:
3343    
3344      The new pid generation scheme described a few weeks ago was overly
3345      complicated.  I implemented a new mechanism that is simpler and
3346      provides a bit more "stability":  Once CM has seen a compilation
3347      unit, it keeps its identity constant (as long as you do not delete
3348      those crucial CM/GUID/* files).  This means that when you change
3349      an interface, compile, then go back to the old interface, and
3350      compile again, you arrive at the original pid.
3351    
3352      There now also is a mechanism that instructs CM to use the plain
3353      environment hash as a module's pid (effectively making its GUID
3354      the empty string).  For this, "noguid" must be specified as an
3355      option to the .sml file in question within its .cm file.
3356      This is most useful for code that is being generated by tools such
3357      as ml-nlffigen (because during development programmers tend to
3358      erase the tool's entire output directory tree including CM's cached
3359      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
3360      revert to the old, broken behavior of SML/NJ, but in specific cases
3361      where there is no danger of interface confusion, its use is ok
3362      (I think).
3363    
3364      ml-nlffigen by default generates "noguid" annotations.  They can be
3365      turned off by specifying -guid in its command line.
3366    
3367    ----------------------------------------------------------------------
3368    Name: Lal George
3369    Date: 2002/03/12 12 14:42:36 EST
3370    Tag: george-20020312-frequency-computation
3371    Description:
3372    
3373    Integrated jump chaining and static block frequency into the
3374    compiler. More details and numbers later.
3375    
3376    ----------------------------------------------------------------------
3377    Name: Lal George
3378    Date: 2002/03/11 11 22:38:53 EST
3379    Tag: george-20020311-jump-chain-elim
3380    Description:
3381    
3382    Tested the jump chain elimination on all architectures (except the
3383    hppa).  This is on by default right now and is profitable for the
3384    alpha and x86, however, it may not be profitable for the sparc and ppc
3385    when compiling the compiler.
3386    
3387    The gc test will typically jump to a label at the end of the cluster,
3388    where there is another jump to an external cluster containing the actual
3389    code to invoke gc. This is to allow factoring of common gc invocation
3390    sequences. That is to say, we generate:
3391    
3392            f:
3393               testgc
3394               ja   L1      % jump if above to L1
3395    
3396            L1:
3397               jmp L2
3398    
3399    
3400    After jump chain elimination the 'ja L1' instructions is converted to
3401    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
3402    up being implemented in their long form (if L2 is far away) using:
3403    
3404            jbe     L3      % jump if below or equal to L3
3405            jmp     L2
3406         L3:
3407            ...
3408    
3409    
3410    For large compilation units L2  may be far away.
3411    
3412    
3413    ----------------------------------------------------------------------
3414    Name: Matthias Blume
3415    Date: 2002/03/11 13:30:00 EST
3416    Tag: blume-20020311-mltreeeval
3417    Description:
3418    
3419    A functor parameter was missing.
3420    
3421    ----------------------------------------------------------------------
3422    Name: Allen Leung
3423    Date: 2002/03/11 10:30:00 EST
3424    Tag: leunga-20020311-runtime-string0
3425    Description:
3426    
3427       The representation of the empty string now points to a
3428    legal null terminated C string instead of unit.  It is now possible
3429    to convert an ML string into C string with InlineT.CharVector.getData.
3430    This compiles into one single machine instruction.
3431    
3432    ----------------------------------------------------------------------
3433    Name: Allen Leung
3434    Date: 2002/03/10 23:55:00 EST
3435    Tag: leunga-20020310-x86-call
3436    Description:
3437    
3438       Added machine generation for CALL instruction (relative displacement mode)
3439    
3440    ----------------------------------------------------------------------
3441    Name: Matthias Blume
3442    Date: 2002/03/08 16:05:00
3443    Tag: blume-20020308-entrypoints
3444    Description:
3445    
3446    Version number bumped to 110.39.1.  NEW BOOTFILES!
3447    
3448    Entrypoints: non-zero offset into a code object where execution should begin.
3449    
3450    - Added the notion of an entrypoint to CodeObj.
3451    - Added reading/writing of entrypoint info to Binfile.
3452    - Made runtime system bootloader aware of entrypoints.
3453    - Use the address of the label of the first function given to mlriscGen
3454      as the entrypoint.  This address is currently always 0, but it will
3455      not be 0 once we turn on block placement.
3456    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3457      with entry points) from mlriscGen.
3458    
3459    ----------------------------------------------------------------------
3460    Name: Allen Leung
3461    Date: 2002/03/07 20:45:00 EST
3462    Tag: leunga-20020307-x86-cmov
3463    Description:
3464    
3465       Bug fixes for CMOVcc on x86.
3466    
3467       1. Added machine code generation for CMOVcc
3468       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3469       3. CMOVcc cannot have an immediate operand as argument.
3470    
3471    ----------------------------------------------------------------------
3472    Name: Matthias Blume
3473    Date: 2002/03/07 16:15:00 EST
3474    Tag: blume-20020307-controls
3475    Description:
3476    
3477    This is a very large but mostly boring patch which makes (almost)
3478    every tuneable compiler knob (i.e., pretty much everything under
3479    Control.* plus a few other things) configurable via both the command
3480    line and environment variables in the style CM did its configuration
3481    until now.
3482    
3483    Try starting sml with '-h' (or, if you are brave, '-H')
3484    
3485    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3486    implements the underlying generic mechanism.
3487    
3488    The interface to some of the existing such facilities has changed somewhat.
3489    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3490    (The getFoo interface is still there for backward-compatibility, but its
3491    use is deprecated.)
3492    
3493    The ml-build script passes -Cxxx=yyy command-line arguments through so
3494    that one can now twiddle the compiler settings when using this "batch"
3495    compiler.
3496    
3497    TODO items:
3498    
3499    We should go through and throw out all controls that are no longer
3500    connected to anything.  Moreover, we should go through and provide
3501    meaningful (and correct!) documentation strings for those controls
3502    that still are connected.
3503    
3504    Currently, multiple calls to Controls.new are accepted (only the first
3505    has any effect).  Eventually we should make sure that every control
3506    is being made (via Controls.new) exactly once.  Future access can then
3507    be done using Controls.acc.
3508    
3509    Finally, it would probably be a good idea to use the getter-setter
3510    interface to controls rather than ref cells.  For the time being, both
3511    styles are provided by the Controls module, but getter-setter pairs are
3512    better if thread-safety is of any concern because they can be wrapped.
3513    
3514    *****************************************
3515    
3516    One bug fix: The function blockPlacement in three of the MLRISC
3517    backpatch files used to be hard-wired to one of two possibilities at
3518    link time (according to the value of the placementFlag).  But (I
3519    think) it should rather sense the flag every time.
3520    
3521    *****************************************
3522    
3523    Other assorted changes (by other people who did not supply a HISTORY entry):
3524    
3525    1. the cross-module inliner now works much better (Monnier)
3526    2. representation of weights, frequencies, and probabilities in MLRISC
3527       changed in preparation of using those for weighted block placement
3528       (Reppy, George)
3529    
3530    ----------------------------------------------------------------------
3531    Name: Lal George
3532    Date: 2002/03/07 14:44:24 EST 2002
3533    Tag: george-20020307-weighted-block-placement
3534    
3535    Tested the weighted block placement optimization on all architectures
3536    (except the hppa) using AMPL to generate the block and edge frequencies.
3537    Changes were required in the machine properties to correctly
3538    categorize trap instructions. There is an MLRISC flag
3539    "weighted-block-placement" that can be used to enable weighted block
3540    placement, but this will be ineffective without block/edge
3541    frequencies (coming soon).
3542    
3543    
3544    ----------------------------------------------------------------------
3545    Name: Lal George
3546    Date: 2002/03/05 17:24:48 EST
3547    Tag: george-20020305-linkage-cluster
3548    
3549    In order to support the block placement optimization, a new cluster
3550    is generated as the very first cluster (called the linkage cluster).
3551    It contains a single jump to the 'real' entry point for the compilation
3552    unit. Block placement has no effect on the linkage cluster itself, but
3553    all the other clusters  have full freedom in the manner in which they
3554    reorder blocks or functions.
3555    
3556    On the x86 the typical linkage code that is generated is:
3557       ----------------------
3558            .align 2
3559       L0:
3560            addl    $L1-L0, 72(%esp)
3561            jmp     L1
3562    
3563    
3564            .align  2
3565       L1:
3566       ----------------------
3567    
3568    72(%esp) is the memory location for the stdlink register. This
3569    must contain the address of the CPS function being called. In the
3570    above example, it contains the address of  L0; before
3571    calling L1 (the real entry point for the compilation unit), it
3572    must contain the address for L1, and hence
3573    
3574            addl $L1-L0, 72(%esp)
3575    
3576    I have tested this on all architectures except the hppa.The increase
3577    in code size is of course negligible
3578    
3579    ----------------------------------------------------------------------
3580    Name: Allen Leung
3581    Date: 2002/03/03 13:20:00 EST
3582    Tag: leunga-20020303-mlrisc-tools
3583    
3584      Added #[ ... ] expressions to mlrisc tools
3585    
3586    ----------------------------------------------------------------------
3587    Name: Matthias Blume
3588    Date: 2002/02/27 12:29:00 EST
3589    Tag: blume-20020227-cdebug
3590    Description:
3591    
3592    - made types in structure C and C_Debug to be equal
3593    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3594    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3595    
3596    ----------------------------------------------------------------------
3597    Name: Matthias Blume
3598    Date: 2002/02/26 12:00:00 EST
3599    Tag: blume-20020226-ffi
3600    Description:
3601    
3602    1. Fixed a minor bug in CM's "noweb" tool:
3603       If numbering is turned off, then truly don't number (i.e., do not
3604       supply the -L option to noweb).  The previous behavior was to supply
3605       -L'' -- which caused noweb to use the "default" line numbering scheme.
3606       Thanks to Chris Richards for pointing this out (and supplying the fix).
3607    
3608    2. Once again, I reworked some aspects of the FFI:
3609    
3610       A. The incomplete/complete type business:
3611    
3612       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3613         gone!
3614       - ML types representing an incomplete type are now *equal* to
3615         ML types representing their corresponding complete types (just like
3616         in C).  This is still safe because ml-nlffigen will not generate
3617         RTTI for incomplete types, nor will it generate functions that
3618         require access to such RTTI.   But when ML code generated from both
3619         incomplete and complete versions of the C type meet, the ML types
3620         are trivially interoperable.
3621    
3622         NOTE:  These changes restore the full generality of the translation
3623         (which was previously lost when I eliminated functorization)!
3624    
3625       B. Enum types:
3626    
3627       - Structure C now has a type constructor "enum" that is similar to
3628         how the "su" constructor works.  However, "enum" is not a phantom
3629         type because each "T enum" has values (and is isomorphic to
3630         MLRep.Signed.int).
3631       - There are generic access operations for enum objects (using
3632         MLRep.Signed.int).
3633       - ml-nlffigen will generate a structure E_foo for each "enum foo".
3634         * The structure contains the definition of type "mlrep" (the ML-side
3635         representation type of the enum).  Normally, mlrep is the same
3636         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3637         then mlrep will be defined as a datatype -- thus facilitating
3638         pattern matching on mlrep values.
3639         ("-ec" will be suppressed if there are duplicate values in an
3640          enumeration.)
3641         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3642         will be generated for each C enum constant xxx.
3643         * Conversion functions m2i and i2m convert between mlrep and
3644         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
3645         * Coversion functions c and ml convert between mlrep and "tag enum".
3646         * Access functions (get/set) fetch and store mlrep values.
3647       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3648         enumerations are merged into one single enumeration represented by
3649         structure E_'.
3650    
3651    ----------------------------------------------------------------------
3652    Name: Allen Leung
3653    Date: 2002/02/25 04:45:00 EST
3654    Tag: leunga-20020225-cps-spill
3655    
3656    This is a new implementation of the CPS spill phase.
3657    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3658    In case of problems, replace it with the old file spill.sml
3659    
3660    The current compiler runs into some serious performance problems when
3661    constructing a large record.  This can happen when we try to compile a
3662    structure with many items.  Even a very simple structure like the following
3663    makes the compiler slow down.
3664    
3665        structure Foo = struct
3666           val x_1 = 0w1 : Word32.int
3667           val x_2 = 0w2 : Word32.int
3668           val x_3 = 0w3 : Word32.int
3669           ...
3670           val x_N = 0wN : Word32.int
3671        end
3672    
3673    The following table shows the compile time, from N=1000 to N=4000,
3674    with the old compiler:
3675    
3676    N
3677    1000   CPS 100 spill                           0.04u  0.00s  0.00g
3678           MLRISC ra                               0.06u  0.00s  0.05g
3679              (spills = 0 reloads = 0)
3680           TOTAL                                   0.63u  0.07s  0.21g
3681    
3682    1100   CPS 100 spill                           8.25u  0.32s  0.64g
3683           MLRISC ra                               5.68u  0.59s  3.93g
3684              (spills = 0 reloads = 0)
3685           TOTAL                                   14.71u  0.99s  4.81g
3686    
3687    1500   CPS 100 spill                           58.55u  2.34s  1.74g
3688           MLRISC ra                               5.54u  0.65s  3.91g
3689              (spills = 543 reloads = 1082)
3690           TOTAL                                   65.40u  3.13s  6.00g
3691    
3692    2000   CPS 100 spill                           126.69u  4.84s  3.08g
3693           MLRISC ra                               0.80u  0.10s  0.55g
3694              (spills = 42 reloads = 84)
3695           TOTAL                                   129.42u  5.10s  4.13g
3696    
3697    3000   CPS 100 spill                           675.59u  19.03s  11.64g
3698           MLRISC ra                               2.69u  0.27s  1.38g
3699              (spills = 62 reloads = 124)
3700           TOTAL                                   682.48u  19.61s  13.99g
3701    
3702    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
3703           MLRISC ra                               4.96u  0.27s  2.72g
3704              (spills = 85 reloads = 170)
3705           TOTAL                                   2375.26u  57.21s  48.00g
3706    
3707    As you can see the old cps spill module suffers from some serious
3708    performance problem.  But since I cannot decipher the old code fully,
3709    instead of patching the problems up, I'm reimplementing it
3710    with a different algorithm.  The new code is more modular,
3711    smaller when compiled, and substantially faster
3712    (O(n log n) time and O(n) space).  Timing of the new spill module:
3713    
3714    4000  CPS 100 spill                           0.02u  0.00s  0.00g
3715          MLRISC ra                               0.25u  0.02s  0.15g
3716             (spills=1 reloads=3)
3717          TOTAL                                   7.74u  0.34s  1.62g
3718    
3719    Implementation details:
3720    
3721    As far as I can tell, the purpose of the CPS spill module is to make sure the
3722    number of live variables at any program point (the bandwidth)
3723    does not exceed a certain limit, which is determined by the
3724    size of the spill area.
3725    
3726    When the bandwidth is too large, we decrease the register pressure by
3727    packing live variables into spill records.  How we achieve this is
3728    completely different than what we did in the old code.
3729    
3730    First, there is something about the MLRiscGen code generator
3731    that we should be aware of:
3732    
3733    o MLRiscGen performs code motion!
3734    
3735       In particular, it will move floating point computations and
3736       address computations involving only the heap pointer to
3737       their use sites (if there is only a single use).
3738       What this means is that if we have a CPS record construction
3739       statement
3740    
3741           RECORD(k,vl,w,e)
3742    
3743       we should never count the new record address w as live if w
3744       has only one use (which is often the case).
3745    
3746       We should do something similar to floating point, but the transformation
3747       there is much more complex, so I won't deal with that.
3748    
3749    Secondly, there are now two new cps primops at our disposal:
3750    
3751     1. rawrecord of record_kind option
3752        This pure operator allocates some uninitialized storage from the heap.
3753        There are two forms:
3754    
3755         rawrecord NONE [INT n]  allocates a tagless record of length n
3756         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
3757                                     and initializes the tag.
3758    
3759     2. rawupdate of cty
3760          rawupdate cty (v,i,x)
3761          Assigns to x to the ith component of record v.
3762          The storelist is not updated.
3763    
3764    We use these new primops for both spilling and increment record construction.
3765    
3766     1. Spilling.
3767    
3768        This is implemented with a linear scan algorithm (but generalized
3769        to trees).  The algorithm will create a single spill record at the
3770        beginning of the cps function and use rawupdate to spill to it,
3771        and SELECT or SELp to reload from it.  So both spills and reloads
3772        are fine-grain operations.  In contrast, in the old algorithm
3773        "spills" have to be bundled together in records.
3774    
3775        Ideally, we should sink the spill record construction to where
3776        it is needed.  We can even split the spill record into multiple ones
3777        at the places where they are needed.  But CPS is not a good
3778        representation for global code motion, so I'll keep it simple and
3779        am not attempting this.
3780    
3781     2. Incremental record construction (aka record splitting).
3782    
3783        Long records with many component values which are simulatenously live
3784        (recall that single use record addresses are not considered to
3785         be live) are constructed with rawrecord and rawupdate.
3786        We allocate space on the heap with rawrecord first, then gradually
3787        fill it in with rawupdate.  This is the technique suggested to me
3788        by Matthias.
3789    
3790        Some restrictions on when this is applicable:
3791        1. It is not a VECTOR record.  The code generator currently does not handle
3792           this case. VECTOR record uses double indirection like arrays.
3793        2. All the record component values are defined in the same "basic block"
3794           as the record constructor.  This is to prevent speculative
3795           record construction.
3796    
3797    ----------------------------------------------------------------------
3798    Name: Allen Leung
3799    Date: 2002/02/22 01:02:00 EST
3800    Tag: leunga-20020222-mlrisc-tools
3801    
3802    Minor bug fixes in the parser and rewriter
3803    
3804    ----------------------------------------------------------------------
3805    Name: Allen Leung
3806    Date: 2002/02/21 20:20:00 EST
3807    Tag: leunga-20020221-peephole
3808    
3809    Regenerated the peephole files.  Some contained typos in the specification
3810    and some didn't compile because of pretty printing bugs in the old version
3811    of 'nowhere'.
3812    
3813    ----------------------------------------------------------------------
3814    Name: Allen Leung
3815    Date: 2002/02/19 20:20:00 EST
3816    Tag: leunga-20020219-mlrisc-tools
3817    Description:
3818    
3819       Minor bug fixes to the mlrisc-tools library:
3820    
3821       1.  Fixed up parsing colon suffixed keywords
3822       2.  Added the ability to shut the error messages up
3823       3.  Reimplemented the pretty printer and fixed up/improved
3824           the pretty printing of handle and -> types.
3825       4.  Fixed up generation of literal symbols in the nowhere tool.
3826       5.  Added some SML keywords to to sml.sty
3827    
3828    ----------------------------------------------------------------------
3829    Name: Matthias Blume
3830    Date: 2002/02/19 16:20:00 EST
3831    Tag: blume-20020219-cmffi
3832    Description:
3833    
3834    A wild mix of changes, some minor, some major:
3835    
3836    * All C FFI-related libraries are now anchored under $c:
3837        $/c.cm      --> $c/c.cm
3838        $/c-int.cm  --> $c/internals/c-int.cm
3839        $/memory.cm --> $c/memory/memory.cm
3840    
3841    * "make" tool (in CM) now treats its argument pathname slightly
3842      differently:
3843        1. If the native expansion is an absolute name, then before invoking
3844           the "make" command on it, CM will apply OS.Path.mkRelative
3845           (with relativeTo = OS.FileSys.getDir()) to it.
3846        2. The argument will be passed through to subsequent phases of CM
3847           processing without "going native".  In particular, if the argument
3848           was an anchored path, then "make" will not lose track of that anchor.
3849    
3850    * Compiler backends now "know" their respective C calling conventions
3851      instead of having to be told about it by ml-nlffigen.  This relieves
3852      ml-nlffigen from one of its burdens.
3853    
3854    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
3855    
3856    * Export C_DEBUG and C_Debug from $c/c.cm.
3857    
3858    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
3859      subtyping relationship between incomplete pointers and their complete
3860      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
3861      with the convention of instantiating 'o with ('t, 'c) obj whenever
3862      the pointer target type is complete.  In the incomplete case, 'o
3863      will be instantiated with some "'c iobj" -- a type obtained by
3864      using one of the functors PointerToIncompleteType or PointerToCompleteType.
3865    
3866      Operations that work on both incomplete and complete pointer types are
3867      typed as taking an 'o ptr while operations that require the target to
3868      be known are typed as taking some ('t, 'c) obj ptr.
3869    
3870      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
3871      where void is an eqtype without any values.  This makes it possible
3872      to work on voidptr values using functions meant to operate on light
3873      incomplete pointers.
3874    
3875    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
3876      been vastly simplified.
3877    
3878    ----------------------------------------------------------------------
3879    Name: Matthias Blume
3880    Date: 2002/02/19 10:48:00 EST
3881    Tag: blume-20020219-pqfix
3882    Description:
3883    
3884    Applied Chris Okasaki's bug fix for priority queues.
3885    
3886    ----------------------------------------------------------------------
3887    Name: Matthias Blume
3888    Date: 2002/02/15 17:05:00
3889    Tag: Release_110_39
3890    Description:
3891    
3892    Last-minute retagging is becoming a tradition... :-(
3893    
3894    This is the working release 110.39.
3895    
3896    ----------------------------------------------------------------------
3897    Name: Matthias Blume
3898    Date: 2002/02/15 16:00:00 EST
3899    Tag: Release_110_39-orig
3900    Description:
3901    
3902    Working release 110.39.  New bootfiles.
3903    
3904    (Update: There was a small bug in the installer so it wouldn't work
3905    with all shells.  So I retagged. -Matthias)
3906    
3907    ----------------------------------------------------------------------
3908    Name: Matthias Blume
3909    Date: 2002/02/15 14:17:00 EST
3910    Tag: blume-20020215-showbindings
3911    Description:
3912    
3913    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
3914    the latter can be useful for exploring what bindings are available at
3915    the interactive prompt.  (The first function returns only the list
3916    of symbols that are really bound, the second prints those but also the
3917    ones that CM's autoloading mechanism knows about.)
3918    
3919    ----------------------------------------------------------------------
3920    Name: Matthias Blume
3921    Date: 2002/02/15 12:08:00 EST
3922    Tag: blume-20020215-iptrs
3923    Description:
3924    
3925    Two improvements to ml-nlffigen:
3926    
3927      1. Write files only if they do not exist or if their current contents
3928         do not coincide with what's being written.  (That is, avoid messing
3929         with the time stamps unless absolutely necessary.)
3930    
3931      2. Implement a "repository" mechanism for generated files related
3932         to "incomplete pointer types".   See the README file for details.
3933    
3934    ----------------------------------------------------------------------
3935    Name: Matthias Blume
3936    Date: 2002/02/14 11:50:00 EST
3937    Tag: blume-20020214-quote
3938    Description:
3939    
3940    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
3941    because of the new and improved tag generation scheme.  (Thanks to Allen
3942    Leung for pointing it out.)
3943    
3944    ----------------------------------------------------------------------
3945    Name: Lal George
3946    Date: 2002/02/14 09:55:27 EST 2002
3947    Tag: george-20020214-isabelle-bug
3948    Description:
3949    
3950    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
3951    of Isabelle on the x86.
3952    
3953    From Allen:
3954    -----------
3955     I've found the problem:
3956    
3957         in ra-core.sml, I use the counter "blocked" to keep track of the
3958         true number of elements in the freeze queue.  When the counter goes
3959         to zero, I skip examining the queue.  But I've messed up the
3960         bookkeeping in combine():
3961    
3962             else ();
3963             case !ucol of
3964               PSEUDO => (if !cntv > 0 then
3965                     (if !cntu > 0 then blocked := !blocked - 1 else ();
3966                                        ^^^^^^^^^^^^^^^^^^^^^^^
3967                      moveu := mergeMoveList(!movev, !moveu)
3968                     )
3969                  else ();
3970    
3971         combine() is called to coalesce two nodes u and v.
3972         I think I was thinking that if the move counts of u and v are both
3973         greater than zero then after they are coalesced then one node is
3974         removed from the freeze queue.  Apparently I was thinking that
3975         both u and v are of low degree, but that's clearly not necessarily true.
3976    
3977    
3978    02/12/2002:
3979        Here's the patch.  HOL now compiles.
3980    
3981        I don't know how this impact on performance (compile
3982        time or runtime).  This bug caused the RA (especially on the x86)
3983        to go thru the potential spill phase when there are still nodes on the
3984        freeze queue.
3985    
3986    
3987    
3988    
3989    ----------------------------------------------------------------------
3990    Name: Matthias Blume
3991    Date: 2002/02/13 22:40:00 EST
3992    Tag: blume-20020213-fptr-rtti
3993    Description:
3994    
3995    Fixed a bug in ml-nlffigen that was introduced with one of the previous
3996    updates.
3997    
3998    ----------------------------------------------------------------------
3999    Name: Matthias Blume
4000    Date: 2002/02/13 16:41:00 EST
4001    Tag: blume-20020213-cmlpq
4002    Description:
4003    
4004    Added new priority queue export symbols (which have just been added to
4005    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
4006    not compile and the installer would choke.)
4007    
4008    ----------------------------------------------------------------------
4009    Name: Matthias Blume
4010    Date: 2002/02/13 16:15:00 EST
4011    Tag: blume-20020213-various
4012    Description:
4013    
4014    1. More tweaks to ml-nlffigen:
4015    
4016       - better internal datastructures (resulting in slight speedup)
4017       - "-match" option requires exact match
4018       - "localized" gensym counters (untagged structs/unions nested within
4019         other structs/unions or within typedefs get a fresh counter; their
4020         tag will be prefixed by a concatenation of their parents' tags)
4021       - bug fixes (related to calculation of transitive closure of types
4022         to be included in the output)
4023    
4024    2. Minor Basis updates:
4025    
4026       - added implementations for List.collate and Option.app
4027    
4028    ----------------------------------------------------------------------
4029    Name: Matthias Blume
4030    Date: 2002/02/11 15:55:00 EST
4031    Tag: blume-20020211-gensym
4032    Description:
4033    
4034    Added a "-gensym" option to command line of ml-nlffigen.  This can be
4035    used to specify a "stem" -- a string that is inserted in all "gensym'd"
4036    names (ML structure names that correspond to unnamed C structs, unions,
4037    and enums), so that separate runs of ml-nlffigen do not clash.
4038    
4039    ----------------------------------------------------------------------
4040    Name: Matthias Blume
4041    Date: 2002/02/11 12:05:00 EST
4042    Tag: blume-20020211-gensml
4043    Description:
4044    
4045    A quick fix for a problem with GenSML (in the pgraph-util library):
4046    Make generation of toplevel "local" optional.  (Strictly speaking,
4047    signature definitions within "local" are not legal SML.)
4048    
4049    Other than that: updates to INSTALL and cm/TODO.
4050    
4051    ----------------------------------------------------------------------
4052    Name: Matthias Blume
4053    Date: 2002/02/08 15:00:00 EST
4054    Tag: blume-20020208-uniquepid
4055    Description:
4056    
4057    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
4058    
4059    1. The installer (config/install.sh) has gotten smarter:
4060    
4061         - Configuration options are a bit easier to specify now
4062           (in config/targets).
4063         - Bug in recognizing .tar.bz2 files fixed.
4064         - Installer automatically resolves dependencies between
4065           configuration options (e.g., if you ask for eXene, you will
4066           also get cml -- regardless whether you asked for it or not).
4067         - Installer can run in "quieter mode" by setting the environment
4068           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
4069           "completely silent", though.
4070         - Build HashCons library as part of smlnj-lib.
4071    
4072    2. A new scheme for assigning persistent identifiers to compilation
4073       units (and, by extension, to types etc.) has been put into place.
4074       This fixes a long-standing bug where types and even dynamic values
4075       can get internally confused, thereby compromising type safety
4076       (abstraction) and dynamic correctness.  See
4077    
4078         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
4079    
4080       for an example of how things could go wrong until now.
4081    
4082       The downside of the new scheme is that pids are not quite as
4083       persistent as they used to be: CM will generate a fresh pid
4084       for every compilation unit that it thinks it sees for the first
4085       time.  That means that if you compile starting from a clean, fresh
4086       source tree at two different times, you end up with different
4087       binaries.
4088    
4089       Cutoff recompilation, however, has not been compromised because
4090       CM keeps pid information in special caches between runs.
4091    
4092    ----------------------------------------------------------------------
4093    Name: Lal George
4094    Date: 2002/02/07 15:34:13 EST 2002
4095    Tag: <none>
4096    Description:
4097    
4098    Compilers that generate assembly code may produce  global labels
4099