Home My Page Projects Code Snippets Project Openings SML/NJ
Summary Activity Forums Tracker Lists Tasks Docs Surveys News SCM Files

SCM Repository

[smlnj] Annotation of /sml/trunk/NOTES/HISTORY
ViewVC logotype

Annotation of /sml/trunk/NOTES/HISTORY

Parent Directory Parent Directory | Revision Log Revision Log


Revision 2304 - (view) (download)

1 : blume 2237 This is the HISTORY file for the SML/NJ repository.
2 : dbm 570
3 :     An entry should be made for _every_ commit to the repository.
4 :     The entries in this file will be used when creating the README
5 :     for new versions, so keep that in mind when writing the
6 :     description.
7 :    
8 :     The form of an entry should be:
9 :    
10 : mblume 1931
11 :     Name: Your Name (web page or e-mail)
12 : blume 2155 Date: yyyy/mm/dd hh:mm:ss
13 : mblume 1931 Tag: cvs-tag-for-this-change
14 : mblume 1480 Description:
15 :    
16 :     ----------------------------------------------------------------------
17 : blume 2304 Name: Matthias Blume (http://www.tti-c.org/blume/)
18 :     Date: 2007/02/02 15:20:00 CST
19 :     Tag: release-110.62
20 :     Description:
21 :    
22 :     New working release (110.62). NEW BOOTFILES!
23 :    
24 :     ----------------------------------------------------------------------
25 : adrassi 2296 Name: Aaron Turon (http://www.cs.uchicago.edu/~adrassi)
26 :     Date: 2007/01/31
27 :     Description:
28 :    
29 :     Brought ml-lpt manual up-to-date with the code. Many minor
30 :     improvements to the tools, and a few minor bugfixes. Prepared
31 :     ml-lpt-lib for integration with new UTF8 structure. Changed
32 :     the interface for creating streams in ml-ulex (we now support
33 :     stream creation from several kinds of sources). Overall, the
34 :     tools are now quite stable, and their interfaces are unlikely to
35 :     change in a way that would break compatibility. More work,
36 :     however, is needed in the documentation.
37 :    
38 :     ----------------------------------------------------------------------
39 : blume 2277 Name: Matthias Blume (http://www.tti-c.org/blume/)
40 : blume 2298 Date: 2007/02/01 10:05:00 CST
41 :     Description:
42 :    
43 :     Implemented library installer. Moved CM plugin code for
44 :     ml-burg, ml-lex, and ml-yacc out of CM source tree and into
45 :     their respective trees. Implemented CM plugin for ml-ulex
46 :     and ml-antlr. Used library installer for ml-burg, ml-ulex,
47 :     and ml-antlr.
48 :    
49 :     For ml-yacc and ml-lex we continue to have permanently "plugged-in"
50 :     CM tools. (It turns out to be too messy to do otherwise because
51 :     there is too much code that during installation relies on the presence
52 :     of these tools -- resulting in a tricky ordering problem.)
53 :    
54 :     ----------------------------------------------------------------------
55 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
56 : blume 2277 Date: 2007/01/30 11:05:00 CST
57 :     Description:
58 :    
59 :     Added SMLofNJ.shiftArgs which is like a shell's "shift" command.
60 :     Modified CM's startup code to use shiftArgs as it processes
61 :     command line arguments. This way, the init code in each .sml-file
62 :     or library that is mentioned at the top level will see only
63 :     those arguments that have not yet been processed at this point.
64 :     In other words, the init code can "seize control" and process the
65 :     remaining command line.
66 :    
67 :     ----------------------------------------------------------------------
68 : jhr 2269 Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
69 :     Date: 2007/01/28
70 :     Description:
71 :    
72 :     Added the UTF8 structure and signature from the Moby compiler to
73 :     the SML/NJ library (and the CML library). These modules will replace
74 :     the version in the ml-lpt-lib.
75 :    
76 :     ----------------------------------------------------------------------
77 : blume 2237 Name: Matthias Blume (http://www.tti-c.org/blume/)
78 : blume 2267 Date: 2007/01/26 15:10:00 CST
79 :     Description:
80 :    
81 :     Added entries to handle ml-lpt-lib.cm in installer.
82 :    
83 :     ----------------------------------------------------------------------
84 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
85 : blume 2259 Date: 2006/12/15 11:10:00 CST
86 : blume 2237 Tag: release-110.61
87 :     Description:
88 :    
89 :     New working release (110.61). NEW BOOTFILES!
90 :    
91 : blume 2259 This supersedes the pre-brown-paper-bug release (see below).
92 :    
93 : blume 2237 ----------------------------------------------------------------------
94 : blume 2259 Name: Matthias Blume (http://www.tti-c.org/blume/)
95 :     Date: 2006/12/15 11:00:00 CST
96 :     Description:
97 :    
98 :     Fixed brown-paper-bag bug with CM's pathname handling, which made
99 :     installation fail under Win32.
100 :    
101 :     ----------------------------------------------------------------------
102 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
103 :     Date: 2006/12/14 12:00:00 CST
104 :     Description:
105 :    
106 :     Obsolete!
107 :    
108 :     [New working release (110.61). NEW BOOTFILES!]
109 :    
110 :     ----------------------------------------------------------------------
111 : jhr 2234 Name: John Reppy (http://cs.uchicago.edu/~jhr)
112 :     Date: 2006/14/06 09:50:00 CST
113 :     Description:
114 :    
115 :     Fixed the code in runtime/c-libs/posix-tty/{tcgetattr,tcsetattr}.c to
116 :     get the c_cc termios data copied correctly. Also moved the allocation
117 :     of the string to avoid problems if it caused a GC.
118 :    
119 :     Thanks to Timothy Bourke for the bug report and fix.
120 :    
121 :     ----------------------------------------------------------------------
122 : blume 2232 Name: Matthias Blume (http://www.tti-c.org/blume)
123 :     Date: 2006/12/08 09:50:00 CST
124 :     Description:
125 :    
126 :     Added code to CM's "standard shell tool" implementation which
127 :     causes it to tolerate (with a warning) the situation where the shell
128 :     command fails (e.g., due to the shell command's non-existence)
129 :     as long as all target files exist.
130 :    
131 :     This makes it possible to, e.g., build ml-yacc from sources even if
132 :     svn checkout messed up the time stamps on files in such a way that
133 :     yacc.grm is younger than yacc.grm.sml or yacc.grm.sig. (Ml-yacc would
134 :     be needed to re-process yacc.grm, but obviously it might not yet be
135 :     available at that time.)
136 :    
137 :     ----------------------------------------------------------------------
138 : jhr 2228 Name: John Reppy (http://cs.uchicago.edu/~jhr)
139 :     Date: 2006/12/06 12:06:00 CST
140 :     Description:
141 :    
142 :     Fixed the types of recvVecFrom, recvVecFrom', recvVecFromNB, and
143 :     recvVecFromNB' in the SOCKET signature. This error is actually
144 :     in the SML Basis specification too.
145 :    
146 :     ----------------------------------------------------------------------
147 : mblume 1923 Name: Matthias Blume (http://www.tti-c.org/blume/)
148 : blume 2226 Date: 2006/12/05 22:40:00 CST
149 :     Description:
150 :    
151 :     CM now reports undefined anchors as errors and aborts execution
152 :     rather than silently pressing on using bogus values.
153 :    
154 :     ----------------------------------------------------------------------
155 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
156 : blume 2224 Date: 2006/11/29 10:00:00 CST
157 :     Description:
158 :    
159 :     Use Say.vsay for printing the "[autoloading]" message, so
160 :     #set CM.Control.verbose false (or -Ccm.verbose=false) can be used
161 :     to suppress them.
162 :    
163 :     ----------------------------------------------------------------------
164 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
165 : blume 2219 Date: 2006/11/10 11:00:00 CST
166 :     Description:
167 :    
168 :     Fixed bug in CM where "with:" specifications that affect compilation
169 :     (as opposed to parsing) were ignored.
170 :    
171 :     ----------------------------------------------------------------------
172 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
173 : blume 2193 Date: 2006/11/09 14:30:00 CST
174 :     Tag: release-110.60
175 :     Description:
176 :    
177 :     New working release (110.60). NEW BOOTFILES!
178 :    
179 :     This is the first subversion-hosted release. There are also changes
180 :     to the directory layout. Here is a table:
181 :    
182 :     (All pathnames are relative to the SML/NJ "root" directory.)
183 :    
184 :     config -> config
185 :    
186 :     src/cm -> base/cm
187 :     src/compiler -> base/compiler
188 :     src/runtime -> base/runtime
189 :     src/system -> base/system
190 :    
191 :     src/cm/pgraph -> pgraph
192 :    
193 :     src/READMES -> base/READMES
194 :     src/$notefile -> base/NOTES/$notefile
195 :     for <notefile> in BOOT CVSNOTES CYGWININSTALL HISTORY INSTALL
196 :     MACOSXINSTALL WININSTALL
197 :    
198 :     src/smlnj-lib -> smlnj-lib
199 :     src/MLRISC -> MLRISC
200 :     ckit -> ckit
201 :     src/cml -> cml
202 :     src/eXene -> eXene
203 :     src/heap2asm -> heap2asm
204 :     src/lexgen -> lexgen
205 :     src/ml-burg -> ml-burg
206 :     src/ml-lex -> ml-lex
207 :     src/ml-yacc -> ml-yacc
208 :     src/ml-nlffi-lib -> nlffi/lib
209 :     src/ml-nlffigen -> nlffi/gen
210 :     src/smlnj-c -> smlnj-c
211 :     src/tools/TraceDebugProf -> trace-debug-profile
212 :    
213 :     In addition, there is also a new
214 :    
215 :     ml-lpt
216 :    
217 :     directory containing two new program generator tools: ml-ulex and
218 :     ml-antlr (a lexer generator that handles unicode and an ANTRL-inspired
219 :     LL(k) parser generator). These tools are currently "beta-quality"
220 :    
221 :     The latest versios of the sources can now be obtained anonymously
222 :     via subversion. For this, it is useful to first check out
223 :    
224 :     svn://smlnj-gforge.cs.uchicago.edu/smlnj/admin
225 :    
226 :     and put the resulting directory on your shell's PATH. This provides
227 :     access to three shell scripts: checkout-all.sh, stat-all.sh,
228 :     and refresh-all.sh.
229 :    
230 :     To create a freshly checked-out copy of the sources, do
231 :    
232 :     checkout-all.sh [dir]
233 :    
234 :     where dir is the optional SML/NJ root directory (default is ".").
235 :    
236 :     This creates the above directory layout. Each subdirectory of
237 :     the root is under individual subversion control. The stat-all.sh
238 :     and refresh-al.sh scripts apply "svn stat" or "svn update" to
239 :     each of these subtrees.
240 :    
241 :     ----------------------------------------------------------------------
242 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
243 : blume 2155 Date: 2006/11/02 12:20:00 CST
244 :     Description:
245 :    
246 :     Reorganized directory layout.
247 :    
248 :     This is a temporary solution, more reorganization is to come.
249 :    
250 :     The basic idea is to have a number of toplevel trees, each corresponding
251 :     to a well-defined part of the overall system. Each part can be
252 :     maintained individually, even in separate source repositories,
253 :     although currently we still serve everything out of the main
254 :     smlnj-gforge tree.
255 :    
256 :     Installer and and scripts have been updated to reflect the new
257 :     layout. The installer (base/src/system/smlnj/installer) is now
258 :     "scriptable" to avoid burning too much knowledge about the
259 :     layout into SML source code. The main script used by the installer
260 :     is in config/actions.
261 :    
262 :     The main change is that many of the subdirectories of what used to
263 :     be known as "src" have moved to the toplevel. The "src" directory
264 :     itself has moved down into a subtree called "base". (We may eventually
265 :     get rid of the extra level of indirection represented by "src".)
266 :    
267 :     The layout is now as follows:
268 :    
269 :     toplevel tree name | default repository
270 :     | (using gf=svn://smlnj-gforge.cs.uchicago.edu/smlnj)
271 :     -------------------+----------------------------------------------------
272 :     config | $gf/config/trunk
273 :     |
274 :     base | $gf/sml/trunk
275 :     |
276 :     smlnj-lib | $gf/smlnj-lib/trunk
277 :     MLRISC | $gf/MLRISC/trunk
278 :     |
279 :     ml-yacc | $gf/ml-yacc/trunk
280 :     ml-lex | $gf/ml-lex/trunk
281 :     ml-burg | $gf/ml-burg/trunk
282 :     lexgen | $gf/lexgen/trunk
283 :     heap2asm | $gf/heap2asm/trunk
284 :     |
285 :     cml | $gf/cml/trunk
286 :     eXene | $gf/eXene/trunk
287 :     ckit | $gf/ckit/trunk
288 :     nlffi | $gf/nlffi/trunk
289 :     |
290 :     smlnj-c | $gf/smlnj-c/trunk
291 :    
292 :     In $gf/admin there are a few useful shell scripts for checking out
293 :     and maintaining the entire collection of trees:
294 :    
295 :     admin/checkout-all.sh [dir]
296 :     -- optionally creates dir and checks out all trees from their
297 :     default repositories; if dir is missing, checkout into .
298 :     admin/refresh-all.sh [dir]
299 :     -- looks at all trees (from the above list) in dir (default: .)
300 :     and runs "svn update" if the tree exists and is under
301 :     subversion control;
302 :     non-existing or non-subversion trees are skipped
303 :     admin/stat-all.sh [dir]
304 :     -- like refresh-all.sh, but runs "svn stat" instead
305 :    
306 :     ----------------------------------------------------------------------
307 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
308 :     Date: 2006/10/05 11:30:00 CDT
309 : mblume 2066 Tag: blume-20061005-merge-amd64
310 :     Description:
311 :    
312 :     Merged code for AMD64 backend (Mike Rainey's work).
313 :     Everything is hooked up but untested.
314 :    
315 :     ----------------------------------------------------------------------
316 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
317 : mblume 1931 Date: 2006/05/17 14:00:00 CST
318 :     Tag: Release_110_59
319 :     Description:
320 :    
321 : blume 2193 New working version (110.59). NEW BOOTFILES!
322 : mblume 1931
323 :     I am freezing 110.59. Changes other than the version-number increase:
324 :    
325 :     eXene:
326 :    
327 :     * committed changes to eXene from Alley Stoughton:
328 :     "fixed bugs in X authorization and resource handling, as well
329 :     as in the pile and viewport widgets"
330 :    
331 :     Runtime:
332 :     * fixed linking problem with NetBSD 3.x.
333 :    
334 :     Lexgen:
335 :     * lexgen tool handles non-ascii characters in 7-bit mode the same
336 :     way that ml-lex does
337 :    
338 :     * lexgen propagates exceptions the same way that ml-lex does
339 :    
340 :     CML:
341 :     * Fixed a bug in the SyncVar polling functions (iGetPoll,
342 :     mTakePoll, and mGetPoll) that could lead to livelock.
343 :    
344 :     ----------------------------------------------------------------------
345 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
346 : mblume 1926 Date: 2006/05/12 15:20:00 CST
347 :     Tag: blume-20060515-makedepend
348 :     Description:
349 :    
350 :     Implemented ml-makedepend (i.e., CM.sources) in a better (more robust)
351 :     way. This should hopefully fix the ml-makedepend problem permanently.
352 :    
353 :     ----------------------------------------------------------------------
354 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
355 : mblume 1925 Date: 2006/05/12 15:15:00 CST
356 :     Tag: blume-20060512-makedepend
357 :     Description:
358 :    
359 :     Fixed long-standing bug with ml-makedepend where it would output a
360 :     spurious dependency to a non-existing file. (This is a simple fix.
361 :     It might need further looking into.)
362 :    
363 :     ----------------------------------------------------------------------
364 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
365 : mblume 1923 Date: 2006/04/20 10:30:00 CDT
366 :     Tag: blume-20060420-heap2exec
367 :     Description:
368 :    
369 :     Committed patches received from Johannes 5 Joemann (joemann@befree.free.de)
370 :     that enable heap2exec under Linux and FreeBSD.
371 :    
372 :     ----------------------------------------------------------------------
373 : jhr 1921 Name: John Reppy
374 :     Date: 2006/04/14
375 : blume 2155 Tag:
376 : jhr 1921 Description:
377 :    
378 :     MLRISC changes:
379 :     1) renamed GAS_PSEUDO_OPS to AS_PSEUDO_OPS and put it in its own file.
380 :     2) added support for NOTB and XORB operators in pseudo-op expressions
381 :     3) added DarwinPseudoOp functor that supports Darwin's assembler syntax.
382 :     4) added support for 64-bit integer literals
383 :    
384 :     ----------------------------------------------------------------------
385 : jhr 1922 Name: Matthias Blume (http://www.tti-c.org/blume/)
386 : mblume 1907 Date: yyyy/mm/dd
387 :     Tag: Release_110_58
388 :     Description:
389 :    
390 :     New working version (110.58). NEW BOOTFILES!
391 :    
392 :     ----------------------------------------------------------------------
393 : mblume 1896 Name: Matthias Blume (http://www.tti-c.org/blume/)
394 : mblume 1905 Date: 2006/03/01 22:15:00 CST
395 :     Tag: blume-20060301-lexgen
396 :     Description:
397 :    
398 :     Incorporated several bugfixes to lexgen. Compiler now compiles to
399 :     fixpoint when using lexgen instead of ml-lex.
400 :    
401 :     ----------------------------------------------------------------------
402 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
403 : mblume 1902 Date: 2006/02/28 22:45:00 CST
404 :     Tag: blume-20060228-lexgen
405 :     Description:
406 :    
407 :     Removed ml-flex and added lexgen instead, using Aaron Turon's newly
408 :     provided tarball. The generated lexers still have problems.
409 :    
410 :     ----------------------------------------------------------------------
411 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
412 : mblume 1898 Date: 2006/02/26 23:30:00 CST
413 :     Tag: blume-20060226-ml-flex
414 :     Description:
415 :    
416 :     Removed ml-flex' dependency on regexp-lib.cm. Turned installation
417 :     of ml-flex on by default.
418 :    
419 :     ----------------------------------------------------------------------
420 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
421 : mblume 1896 Date: 2006/02/24 16:30:00 CST
422 :     Tag: blume-20060224-ml-flex
423 :     Description:
424 :    
425 :     Added ml-flex sources. Partially integrated, but should not be
426 :     turned on yet! (Read: leave it commented-out in config/targets!)
427 :    
428 :     ----------------------------------------------------------------------
429 : jhr 1893 Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
430 :     Date: 2006/02/23
431 :     Tag: jhr-20060223-x86-64
432 :     Description:
433 :     Changes to support building on x86-64 systems (using the 32-bit mode).
434 :     Also cleaned up signal handling on Linux. Support for pre-2.2 Linux
435 :     kernels dropped.
436 :    
437 :     ----------------------------------------------------------------------
438 : allenleung 1890 Name: Allen Leung (leunga (at) reservoir (dot) com)
439 :     Date: 2006/02/22 15:07:00 EST
440 :     Tag: leunga-20060222-mdl-gen
441 :     Description:
442 :    
443 :     CM has changed. Updated the script for rebuilding the MLRISC generated
444 :     files.
445 :    
446 :     ----------------------------------------------------------------------
447 : mblume 1886 Name: Matthias Blume (http://www.tti-c.org/blume/)
448 :     Date: 2006/02/14 17:30:00 CST
449 :     Tag: blume-20060214-intel-mac
450 :     Description:
451 :    
452 :     Hooked code for Darwin-specific Intel ABI into main compiler.
453 :     (This is a temporary solution which relies on the fact that the
454 :     compiler itself does not use NLFFI. Eventually we need to divorce
455 :     intel mac from generic x86 unix code and make separate sets of binaries.)
456 :    
457 :     ----------------------------------------------------------------------
458 : jhr 1882 Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
459 : jhr 1885 Date: 2006/02/13
460 :     Tag: jhr-20060213-intel-mac
461 :     Description:
462 :     Changed MLRISC x86 CCalls for partial support of MacOS X ABI.
463 :    
464 :     ----------------------------------------------------------------------
465 :     Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
466 : jhr 1882 Date: 2006/02/06
467 :     Tag: jhr-20060206-intel-mac
468 :     Description:
469 :     Changes to support MacOS X on Intel hardware. The C-calls support in
470 :     MLRISC must also be updated to support the MacOS X ABI.
471 :    
472 :     ----------------------------------------------------------------------
473 : mblume 1863 Name: Matthias Blume (http://www.tti-c.org/blume/)
474 : mblume 1879 Date: 2005/12/16 00:15:00 CST
475 :     Tag: blume-20051216-cm
476 :     Description:
477 :    
478 :     Improved error reporting and handling in CM.
479 :    
480 :     ----------------------------------------------------------------------
481 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
482 : mblume 1878 Date: 2005/11/21 14:00:00 CST
483 :     Tag: blume-20051121-html-lib
484 :     Description:
485 :    
486 :     Enabling $/html-lib.cm in config/preloads did not work. This is fixed
487 :     now. (Since the anchor mapping for html-lib.cm is not yet in effect
488 : blume 2155 at the time when config/preloads is processed, the library has to
489 : mblume 1878 be referred to by another name. In 110.57 this would be
490 :     $SMLNJ-LIB/HTML/html-lib.cm. I arranged for $smlnj/smlnj-lib/html-lib.cm
491 :     to be valid as well.)
492 :    
493 :     Thanks to Todd Wilson (Fresno) for alerting me to this issue.
494 :    
495 :     ----------------------------------------------------------------------
496 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
497 : mblume 1874 Date: 2005/11/19 13:45:00 CST
498 :     Tag: Release_110_57
499 :     Description:
500 :    
501 :     Moved release tag for 110.57.
502 :    
503 :     ----------------------------------------------------------------------
504 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
505 : mblume 1873 Date: 2005/11/19 23:35:00 CST
506 :     Tag: blume-20051119-fish
507 :     Description:
508 :    
509 :     Fixed a problem in config/install.sh where it tries to "fish" the
510 :     name of the CM metadata directory from the wrong place (because
511 :     the physical location of basis.cm has changed). Also, corrected the
512 :     path anchor for $/html-lib.cm. (Thanks to M. Fluet for pointing out
513 :     these problems.)
514 :    
515 :     ----------------------------------------------------------------------
516 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
517 : mblume 1867 Date: 2005/11/16 14:30:00 CST
518 : mblume 1874 Tag: blume-orig-110_57
519 : mblume 1867 Description:
520 :    
521 :     New working version (110.57). NEW BOOTFILES!
522 :    
523 : mblume 1874 UPDATE: The release tag has been moved to include a fix for a brown-paper-
524 :     bag glitch (see above).
525 :    
526 : mblume 1867 ----------------------------------------------------------------------
527 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
528 : mblume 1866 Date: 2005/11/16 10:25:00 CST
529 :     Tag: blume-22051116-btrace
530 :     Description:
531 :    
532 :     Fixed problem with bogus exception message when using back-trace
533 :     facility.
534 :    
535 :     ----------------------------------------------------------------------
536 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
537 : mblume 1865 Date: 2005/11/15 22:35:00 CST
538 :     Tag: blume-22051115-array2copy
539 :     Description:
540 :    
541 :     Added simple implementation of Array2.copy. (Warning: mostly untested.)
542 :    
543 :     ----------------------------------------------------------------------
544 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
545 : mblume 1864 Date: 2005/11/15 21:35:00 CST
546 :     Tag: blume-22051115-versiontool
547 :     Description:
548 :    
549 :     Reversed change to src/system/smlnj/internals/versiontool.cm. This
550 :     file gets loaded as a tool -- by the equivalent of CM.make during the
551 :     run of CMB.make. Thus, CMB's path configuration is meaningless for
552 :     it. Instead, it has the status of "user code", so it should use
553 :     $/basis.cm to refer to the Basis library. (At least that's true for
554 :     the purpose of bootstrapping the previous change. In the future it
555 :     might make sense to have versiontool.cm refer to
556 :     $smlnj/basis/basis.cm, i.e., the version of the Basis that the
557 :     compiler itself uses.)
558 :    
559 :     Also patched src/system/testml to have it activate those extra anchor
560 :     bindings in config/extrapathconfig.
561 :    
562 :     ----------------------------------------------------------------------
563 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
564 : mblume 1863 Date: 2005/11/15 16:35:00 CST
565 :     Tag: blume-22051115-libraries
566 :     Description:
567 :    
568 :     This change affects the way the following libraries are tied into
569 :     the system:
570 :    
571 :     $/basis.cm
572 :     $/smlnj-lib.cm
573 :     $/pp-lib.cm
574 :     $/controls-lib.cm
575 :     $/html-lib.cm
576 :     $/ml-yacc-lib.cm
577 :    
578 :     These libraries are now internally (as seen from the source code
579 :     of the implementation itself) known by the following names:
580 :    
581 :     $smlnj/basis/basis.cm
582 :     $smlnj/smlnj-lib/smlnj-lib.cm
583 :     $smlnj/smlnj-lib/pp-lib.cm
584 :     $smlnj/smlnj-lib/controls-lib.cm
585 :     $smlnj/smlnj-lib/html-lib.cm
586 :     $smlnj/ml-yacc/ml-yacc-lib.cm
587 :    
588 :     This makes it possible to work with code that requires different
589 :     versions of these libraries, and which refers to these libraries using
590 :     their "default" names (i.e., the first set of names above). In other
591 :     words, one can un-define or re-define those default names without
592 :     compromising the proper functioning of the compiler itself.
593 :    
594 :     A similar procedure had already been performed for several of the
595 :     MLRISC libraries that are linked into the compiler. I did some
596 :     cleanup on this code.
597 :    
598 :     A new file in the config directory (named extrapathconfig) is
599 :     responsible for setting up path anchors that the compiler itself does
600 :     not need, but that are typically required by user code.
601 :    
602 :     ----------------------------------------------------------------------
603 : mblume 1862 Name: Matthias Blume
604 :     Date: 2005/11/07 17:05:00 CST
605 :     Tag: blume-20051107-slice
606 :     Description:
607 :    
608 :     Fixed erroneous out-of-bounds test in the "update" function of
609 :     various *ArraySlice modules. (Thanks to Vesa A. Norrman for pointing
610 :     out the problem.)
611 :    
612 :     Pushed some Basis changes through ML-Lex, CML, and eXene.
613 :    
614 :     ----------------------------------------------------------------------
615 : jhr 1858 Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
616 : jhr 1861 Date: 2005/11/07
617 :     Description:
618 :     Fixed a Basis incompatibility: The depreciated function Substring.all
619 :     was removed (use Substring.full instead).
620 :    
621 :     ----------------------------------------------------------------------
622 :     Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
623 : jhr 1858 Date: 2005/11/05
624 :     Description:
625 :     Tweaked interval set API in SML/NJ library; see the CHANGES file for details.
626 :    
627 :     ----------------------------------------------------------------------
628 : mblume 1856 Name: Matthias Blume (http://www.tti-c.org/blume/)
629 :     Date: 2005/11/02 16:25:00 CST
630 :     Tag: blume-20051102-hex
631 :     Description:
632 :    
633 :     * Runtime system bootstrap code now accepts hex digits in BOOTLIST
634 :     in either upper case or lower case format.
635 :     * Pushed changes to names of Pack<N>{Big,Little} structures through
636 :     CML and eXene.
637 :    
638 :     ----------------------------------------------------------------------
639 : jhr 1852 Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
640 : jhr 1855 Date: 2005/11/02
641 :     Description:
642 :     Fixed a Basis incompatibility: Pack<N>{Big,Little} structures should
643 :     be called PackWord<N>{Big,Little}.
644 :    
645 :     ----------------------------------------------------------------------
646 :     Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
647 : jhr 1852 Date: 2005/10/28
648 :     Description:
649 :     Fixed a minor Basis incompatibility: hex digits should be upper case.
650 :    
651 :     ----------------------------------------------------------------------
652 : mblume 1846 Name: Matthias Blume (blume (at) tti - c (dot) org)
653 :     Date: 2005/10/25 16:00:00 CDT
654 :     Tag: Release_110_56
655 :     Description:
656 :    
657 :     New working version (110.56). NEW BOOTFILES!
658 :    
659 :     ----------------------------------------------------------------------
660 : jhr 1838 Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
661 : jhr 1840 Date: 2005/10/25
662 :     Description:
663 :     Added interval sets to utility library (signatures INTERVAL_DOMAIN
664 :     and INTERVAL_SET, and functor IntervalSetFn).
665 :    
666 :     ----------------------------------------------------------------------
667 :     Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
668 : jhr 1838 Date: 2005/10/14
669 :     Description:
670 :     Add Zhong Shao's fix for datatype equality functions.
671 :    
672 :     ----------------------------------------------------------------------
673 : allenleung 1836 Name: Allen Leung (leunga (at) reservoir (dot) com)
674 : allenleung 1837 Date: 2005/10/14 12:30:00 EST
675 :     Tag: leunga-20051014-x86-spill
676 :     Description:
677 :    
678 :     Bug fix a bug found by Carl Hauser.
679 :     There was a typo in the reload code for FCMP in x86SpillInstr.sml.
680 :    
681 :     ----------------------------------------------------------------------
682 :     Name: Allen Leung (leunga (at) reservoir (dot) com)
683 : allenleung 1836 Date: 2005/10/14 11:15:00 EST
684 :     Tag: leunga-20051014-x86-asm
685 :     Description:
686 :    
687 :     Removed some debugging code in file x86Asm.sml.
688 :     The function emit_operand was printing out debugging output.
689 :    
690 :     ----------------------------------------------------------------------
691 : jhr 1813 Name: John Reppy
692 : jhr 1831 Date: 2005/07/27
693 :     Tag:
694 :     Description:
695 :    
696 : jhr 1832 Fixed ml-lex to recognize "\r" as representing carriage return.
697 :    
698 :     ----------------------------------------------------------------------
699 :     Name: John Reppy
700 :     Date: 2005/07/27
701 :     Tag:
702 :     Description:
703 :    
704 : jhr 1831 Fixed ml-yacc to work on files with non-native end-of-line
705 :     encodings (e.g., Windows text file on a Unix system).
706 :    
707 :     ----------------------------------------------------------------------
708 :     Name: John Reppy
709 : jhr 1824 Date: 2005/07/20
710 :     Tag:
711 :     Description:
712 :    
713 :     Added changes from Dominic Evans (oldmanuk (at) gmail (dot) com)
714 :     to support HPUX 11.
715 :    
716 :     ----------------------------------------------------------------------
717 :     Name: John Reppy
718 : jhr 1813 Date: 2005/07/06
719 :     Tag:
720 :     Description:
721 :    
722 :     Changes to the SML/NJ library. See smlnj-lib/CHANGES for details.
723 :    
724 :     ----------------------------------------------------------------------
725 : mblume 1808 Name: Matthias Blume (blume (at) tti - c (dot) org)
726 :     Date: 2005/07/06 00:45:00 CDT
727 :     Tag: blume-20050706-slice-copy
728 :     Description:
729 :    
730 :     Fixed reversed logic for deciding whether to "copy up" or "copy down"
731 :     in *-array-slice.sml.
732 :    
733 :     ----------------------------------------------------------------------
734 : allenleung 1806 Name: Allen Leung (leunga (at) reservoir (dot) com)
735 : allenleung 1807 Date: 2005/05/31 17:00:00 EST
736 :     Tag: leunga-20050531-cygwin-fault-2
737 :     Description:
738 :    
739 :     A typo in the cygwin code fixed.
740 :    
741 :     ----------------------------------------------------------------------
742 :     Name: Allen Leung (leunga (at) reservoir (dot) com)
743 : allenleung 1806 Date: 2005/05/31 16:47:00 EST
744 :     Tag: leunga-20050531-cygwin-fault
745 :     Description:
746 :    
747 : blume 2155 Updated Cygwin's fault/signal handling to match the Windows version.
748 : allenleung 1806 Updated the export list.
749 :    
750 :     ----------------------------------------------------------------------
751 : mblume 1791 Name: Matthias Blume (blume (at) tti - c (dot) org)
752 : mblume 1795 Date: 2005/05/18 13:35:00 CDT
753 :     Tag: Release_110_54
754 :     Description:
755 :    
756 :     New working version (110.54). NEW BOOTFILES!
757 :    
758 :     ----------------------------------------------------------------------
759 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
760 : mblume 1794 Date: 2005/05/18 11:58:00 CDT
761 :     Tag: blume-20050518-installer
762 :     Description:
763 :    
764 :     Added support scripts for Mac OS X PackageMaker and modified
765 :     config/install.sh so that it supports re-dumping a heap image
766 :     after customization.
767 :    
768 :     ----------------------------------------------------------------------
769 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
770 : mblume 1793 Date: 2005/05/18 10:55:00 CDT
771 :     Tag: blume-20050518-realdiv-noovld
772 :     Description:
773 :    
774 :     Un-overloaded / to work around bug in overloading resolution code.
775 :    
776 :     ----------------------------------------------------------------------
777 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
778 : mblume 1792 Date: 2005/05/16 23:50:00 CDT
779 :     Tag: blume-20050516-redump-heap
780 :     Description:
781 :    
782 :     Added mechanism for re-creating a heap file for the interactive system
783 :     after configuration variables have been changed.
784 :    
785 :     CM.redump_heap : string -> unit
786 :    
787 :     This is much like SMLofNJ.exportML, but starting from the resulting
788 : blume 2155 heap does not return to the caller of CM.redump_heap but
789 : mblume 1792 restarts the interactive system from scratch. The original call of
790 :     CM.redump_heap does not return but ends the interactive session.
791 :     Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
792 :    
793 :     Internally, redump_heap winds the dynamic execution context back to
794 :     the point where the original heap image was created and re-executes
795 :     the heap image generation code in the boot code.
796 :    
797 :     ----------------------------------------------------------------------
798 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
799 : mblume 1791 Date: 2005/05/09 21:55:00 CDT
800 :     Tag: blume-20050509-word64
801 :     Description:
802 :    
803 :     Added a hack to the existing hack known as Word64 to make fromString
804 :     behave correctly. I am still not sure whether Word64.scan will work
805 :     as specified with respect to the interaction of radix and prefix.
806 :    
807 :     ----------------------------------------------------------------------
808 : allenleung 1785 Name: Allen Leung (leunga (at) reservoir (dot) com)
809 :     Date: 2005/05/04 11:50:00 EST
810 :     Tag: leunga-20050504-checkgc
811 :     Description:
812 :    
813 : blume 2155 Added a gc protocol checking phase. This phase is enabled with
814 : allenleung 1785 the flag "check-gc". "debug-check-gc" turns on the verbose mode.
815 :    
816 :     ----------------------------------------------------------------------
817 : mblume 1741 Name: Matthias Blume (blume (at) tti - c (dot) org)
818 : mblume 1784 Date: 2005/05/04 10:45:00 CDT
819 :     Tag: blume-20050504-intinf
820 :     Description:
821 :    
822 :     Fixed a bug in the implementation of div and mod for IntInf.
823 :     Thanks to Neophytos Michael for reporting the problem.
824 :    
825 :     ----------------------------------------------------------------------
826 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
827 : mblume 1783 Date: 2005/05/04 10:35:00 CDT
828 :     Tag: blume-20050504-join
829 :     Description:
830 :    
831 :     Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
832 :    
833 :     ----------------------------------------------------------------------
834 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
835 : mblume 1770 Date: 2005/02/28 23:40:00 CST
836 :     Tag: blume-20050228-mVar
837 :     Description:
838 :    
839 :     Fixed serious bug (brown paper bag variety) in new implementation of
840 :     structure Atom in CML. (I had accidentally used a mailbox instead of
841 :     an mvar, leaving the door open for races.)
842 :    
843 :     ----------------------------------------------------------------------
844 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
845 : mblume 1761 Date: 2005/02/25 15:00:00 CST
846 :     Tag: Release_110_53
847 :     Description:
848 :    
849 :     New working version (110.53). NEW BOOTFILES!
850 :    
851 :     ----------------------------------------------------------------------
852 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
853 : mblume 1760 Date: 2005/02/25 14:50:00 CST
854 :     Tag: blume-20050225-susp
855 :     Description:
856 :    
857 :     Brought back SMLofNJ.Susp. The underlying suspension type is the one
858 :     implemented in Core -- which means that it is the same as the one used
859 :     by the lazy extension.
860 :    
861 :     ----------------------------------------------------------------------
862 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
863 : mblume 1759 Date: 2005/02/24 16:50:00 CST
864 :     Tag: blume-20050224-cml-atom
865 :     Description:
866 :    
867 :     Simpler and at the same time more general implementation of structure
868 :     Atom in CML.
869 :    
870 :     ----------------------------------------------------------------------
871 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
872 : mblume 1758 Date: 2005/02/15 17:35:00 CST
873 :     Tag: blume-20050215-tools
874 :     Description:
875 :    
876 :     Created new "tools" directory under "src" and moved "TraceDebugProf"
877 :     there.
878 :    
879 :     ----------------------------------------------------------------------
880 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
881 : mblume 1755 Date: 2005/02/10 17:55:00 CST
882 :     Tag: blume-20050210-longlong
883 :     Description:
884 :    
885 :     Implemented "long long" arguments and results for NLFFI. (Only the
886 :     PPC/MacOS implementation is complete, the other backends still need to
887 :     be updated.)
888 :    
889 :     ----------------------------------------------------------------------
890 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
891 : mblume 1751 Date: 2005/01/24 17:40:00 CST
892 :     Tag: blume-20050124-mlyacc
893 :     Description:
894 :    
895 :     Minor cleanup in ML-Yacc rule printing mechanism. This should fix a
896 :     problem with certain "as" patterns which previously got rendered
897 :     using incorrect syntax.
898 :    
899 :     ----------------------------------------------------------------------
900 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
901 : mblume 1750 Date: 2005/01/18 12:00:00 CST
902 :     Tag: blume-20050118-profile
903 :     Description:
904 :    
905 :     Made time profiling code (interrupt handler) in runtime system aware
906 :     of new array representation.
907 :    
908 :     ----------------------------------------------------------------------
909 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
910 : mblume 1747 Date: 2005/01/14 18:00:00 CST
911 :     Tag: blume-20050114-heap2exec
912 :     Description:
913 :    
914 :     Implemented new (but still experimental) heap2exec facility. This is
915 :     tested under Mac OS X and should work under Linux (will test shortly).
916 :     It will probably also work on the Sparc (will test some time later).
917 :    
918 :     - removed old "HACKED_STANDALONE" hack from runtime
919 :    
920 :     To be able to test this, uncomment the request for "heap2asm" in
921 :     config/targets prior to installation. (Notice that this is different
922 :     from "heap2exec" mentioned below. Not a typo.)
923 :    
924 :     To perform an actual test, run the command
925 :    
926 :     $ bin/heap2exec heapfile execfile
927 :    
928 :     (You can put heap2exec on your shell's path.)
929 :    
930 :     For example, run
931 :    
932 :     $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
933 :    
934 :     This will create a standalone executable called "mly" which you can
935 :     then invoke directly as a command.
936 :    
937 :     ----------------------------------------------------------------------
938 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
939 : mblume 1746 Date: 2005/01/07 11:44:00 CST
940 :     Tag: blume-20050107-mlstring
941 :     Description:
942 :    
943 :     fixed off-by-one error in ML_STRING macro (globals.c)
944 :    
945 :     ----------------------------------------------------------------------
946 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
947 : mblume 1744 Date: 2004/12/23 18:00:00 CST
948 :     Tag: blume-20041223-santa
949 :     Description:
950 :    
951 :     Made ml-build script "smarter" (but only very little).
952 :    
953 :     ----------------------------------------------------------------------
954 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
955 : mblume 1742 Date: 2004/12/21 15:05:00 CST
956 :     Tag: blume-20041221-longlong
957 :     Description:
958 :    
959 :     * Implemented access to signed and unsigned long long data in NLFFI.
960 :     (The parameter-passing part of the picture has not complete. But
961 :     data structure access seems to work.)
962 :    
963 :     * Fixed CM's incorrect assumption that the PPC is little-endian.
964 :     (On the Mac, it is big-endian. And that's currently our only
965 :     PPC platform.)
966 :    
967 :     ----------------------------------------------------------------------
968 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
969 : mblume 1741 Date: 2004/12/21 12:50:00 CST
970 :     Tag: blume-20041221-memory
971 :     Description:
972 :    
973 :     Some cleanup in the $c/memory.cm library: separated some concerns by
974 :     moving allocation code and memory access code each into their own
975 :     files.
976 :    
977 :     ----------------------------------------------------------------------
978 : allenleung 1740 Name: Allen Leung (leunga (at) reservoir (dot) com)
979 :     Date: 2004/12/17 16:12:00 EST
980 :     Tag: leunga-20041217-cygwin-smlnj-home
981 :     Description:
982 :    
983 : blume 2155 The Unix I/O library of SML/NJ on cygwin does not understand
984 :     Windows style pathname, so problems arise when SMLNJ_HOME is set
985 :     to a Windows style pathname. _run-sml now converts SMLNJ_HOME
986 :     to a POSIX pathname on cygwin.
987 : allenleung 1740
988 :     ----------------------------------------------------------------------
989 : mblume 1737 Name: Matthias Blume (blume (at) tti - c (dot) org)
990 : mblume 1736 Date: 2004/12/16 13:35:00 CST
991 :     Tag: Release_110_52
992 :     Description:
993 :    
994 :     Last-minute changes incorporated into 110.52. Release tag moved.
995 :    
996 :     The changes:
997 :     - HashString.hashString' -> HashString.hashSubstring
998 :     - bug fix in UnivariateStats
999 :    
1000 :     ----------------------------------------------------------------------
1001 : mblume 1702 Name: Matthias Blume (blume (at) tti - c (dot) org)
1002 : mblume 1733 Date: 2004/12/15 23:40:00 CST
1003 :     Tag: blume-20041215-hashSubstring
1004 :     Description:
1005 :    
1006 :     - HashString.hashString' -> HashString.hashSubstring
1007 :     - corresponding changes in atom.sml
1008 :     - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
1009 :     added some comments
1010 :    
1011 :     ----------------------------------------------------------------------
1012 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1013 : mblume 1728 Date: 2004/12/15 15:30:00 CST
1014 : mblume 1736 Tag: (Release_110_52)
1015 : mblume 1728 Description:
1016 :    
1017 :     New working version (110.52). NEW BOOTFILES!
1018 :    
1019 :     ----------------------------------------------------------------------
1020 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1021 : mblume 1727 Date: 2004/12/15 12:45:00 CST
1022 :     Tag: blume-20041215-spaces
1023 :     Description:
1024 :    
1025 :     More on the space problem (this time for Win32).
1026 :    
1027 :     ----------------------------------------------------------------------
1028 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1029 : mblume 1725 Date: 2004/12/14 17:30:00 CST
1030 :     Tag: blume-20041214-spaces
1031 :     Description:
1032 :    
1033 :     Hacked some of the scripts (in particular: the installer) to cope with
1034 :     spaces in filenames a bit better. But beware: the current "solution"
1035 :     is likely still full of bugs and inherently incomplete. (We need to
1036 :     do away with those shell scripts for a comprehensive solution.)
1037 :    
1038 :     ----------------------------------------------------------------------
1039 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1040 : mblume 1724 Date: 2004/12/13 14:45:00 CST
1041 :     Tag: blume-20041213-ml-makedepend
1042 :     Description:
1043 :    
1044 :     Fixed bug in code for ml-makedepend.
1045 :    
1046 :     ----------------------------------------------------------------------
1047 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1048 : mblume 1720 Date: 2004/12/09 16:30:00 CST
1049 :     Tag: blume-20041209-statistics
1050 :     Description:
1051 :    
1052 :     Added two simple but potentially useful statistics modules to SML/NJ Library.
1053 :     (See CHANGES file there.)
1054 :    
1055 :     ----------------------------------------------------------------------
1056 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1057 : mblume 1718 Date: 2004/12/01 16:50:00 CST
1058 :     Tag: blume-20041201-atom
1059 :     Description:
1060 :    
1061 :     smlnj-lib:
1062 :    
1063 :     Added function HashString.hashString' for substrings.
1064 :     Hand-inlined CharVector.foldl into HashString (for speed).
1065 :     Modified implementation of structure Atom to avoid extracting
1066 :     strings from substrings unless necessary.
1067 :     (Also see CHANGES file for smlnj-lib.)
1068 :    
1069 :     ----------------------------------------------------------------------
1070 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1071 : mblume 1717 Date: 2004/11/24 22:15:00 CST
1072 :     Tag: blume-20041124-cml
1073 :     Description:
1074 :    
1075 :     Made sure CML compiles when Position = Int64.
1076 :    
1077 :     ----------------------------------------------------------------------
1078 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1079 : mblume 1715 Date: 2004/11/24 14:30:00 CST
1080 :     Tag: blume-20041124-position
1081 :     Description:
1082 :    
1083 :     The compiler can now be compiled in a mode that makes structure Position
1084 :     equal to Int64. The default, however, is unchanged (Position = Int31)
1085 :     for the time being.
1086 :    
1087 :     To enable 64-bit positions, use the following procedure:
1088 :    
1089 :     1. Start sml
1090 :     2. Autoload $smlnj/cmb.cm (if not already autoloaded)
1091 :     3. Type
1092 :     #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
1093 :     4. Run CMB.make() as usual.
1094 :    
1095 :     This is barely tested. The only test so far was a little SML program
1096 :     counting the number of characters in an 8-gigabyte file by
1097 :     reading it character-by-character. That test was successful.
1098 :    
1099 :     In support of 64-bit positions, a number of new functions have been
1100 :     added to the runtime system.
1101 :    
1102 :     ----------------------------------------------------------------------
1103 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1104 : mblume 1716 Date: 2004/11/23 14:45:00 CST
1105 :     Tag: blume-20041123-useFile
1106 : mblume 1712 Description:
1107 :    
1108 :     Fixed a problem with unhelpful error messages related to problems with
1109 :     .cm- or .sml files that appear as part of the sml command line.
1110 :    
1111 :     ----------------------------------------------------------------------
1112 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1113 : mblume 1703 Date: 2004/11/18 15:40:00 CST
1114 :     Tag: Release_110_51
1115 :     Description:
1116 :    
1117 :     New working version (110.51). NEW BOOTFILES!
1118 :    
1119 :     ----------------------------------------------------------------------
1120 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1121 : mblume 1702 Date: 2004/11/18 15:35:00 CST
1122 :     Tag: <none>
1123 :     Description:
1124 :    
1125 :     Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
1126 :     Joemann).
1127 :    
1128 :     ----------------------------------------------------------------------
1129 : allenleung 1701 Name: Allen Leung (leunga (at) reservoir (dot) com)
1130 :     Date: 2004/11/17 16:05:21 EST 2004
1131 :     Tag: leunga-20041117-mlrisc-live-kill
1132 :     Description:
1133 :    
1134 : blume 2155 Added support for MLTree constructs LIVE and KILL
1135 : allenleung 1701 to all the architectures.
1136 :    
1137 :     ----------------------------------------------------------------------
1138 : mblume 1687 Name: Matthias Blume (blume (at) tti - c (dot) org)
1139 : mblume 1689 Date: 2004/11/13 00:20:00 CST
1140 :     Tag: blume-20041113-versiontool
1141 :     Description:
1142 :    
1143 :     - Stripped down the versiontool: It now only handles the version number.
1144 :     The date string is generated at bootstrap time (during makeml).
1145 :    
1146 :     - In a previous commit, fixed a minor issue with how polyequal is being
1147 :     translated. In particular, the code now "looks through" abstractions.
1148 :     This results in slightly fewer polyEqual warnings and hopefully slightly
1149 :     more efficient code. Important examples for where this matters are
1150 :     the new int64 and word64 types.
1151 :    
1152 :     ----------------------------------------------------------------------
1153 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1154 : mblume 1687 Date: 2004/11/12 00:30:00 CST
1155 :     Tag: blume-20041112-int64
1156 :     Description:
1157 :    
1158 :     Structure Int64 fully hooked in. (The implementation is not very
1159 :     efficient, though.)
1160 :    
1161 :     ----------------------------------------------------------------------
1162 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1163 : mblume 1686 Date: 2004/11/11 17:30:00 CST
1164 :     Tag: blume-20041111-more64
1165 :     Description:
1166 :    
1167 :     All the pieces of Word64 are now there, with the exception of the
1168 :     conversions from and to LargeWord. (Eventually these need to be identities,
1169 :     but for the time being they don't even make sense because LargeWord is
1170 :     32-bit wide.)
1171 :    
1172 :     Also started to add similar support for Int64, but major pieces of that
1173 :     are still missing.
1174 :    
1175 :     ----------------------------------------------------------------------
1176 : mblume 1668 Name: Matthias Blume (blume (at) tti - c (dot) org)
1177 : mblume 1684 Date: 2004/11/11 00:15:00 CST
1178 :     Tag: blume-20041111-word64
1179 :     Description:
1180 :    
1181 :     Structure Word64 is now (almost) complete, word literals and patterns
1182 :     seem to work. There are a few odd pieces missing. In particular,
1183 :     I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
1184 :     at the moment.
1185 :    
1186 :     Making Word64 official would mean that LargeWord becomes Word64. But
1187 :     this requires extreme care because most word-word conversions have to
1188 :     go through LargeWord, so making a mistake means loss of efficiency or
1189 :     worse. Eventually there will be a solution similar to (but actually
1190 :     simpler than) what I did with IntInf.
1191 :    
1192 :     ----------------------------------------------------------------------
1193 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1194 : mblume 1683 Date: 2004/11/10 18:12:00 CST
1195 :     Tag: blume-20041110-64bit
1196 :     Description:
1197 :    
1198 :     More 64-bit hacking (but still not even half-way there yet).
1199 :     Also, some assorted improvements to the handling of 8-bit words.
1200 :    
1201 :     ----------------------------------------------------------------------
1202 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1203 : mblume 1682 Date: 2004/11/09 17:50:00 CST
1204 : mblume 1683 Tag: <none>
1205 : mblume 1682 Description:
1206 :    
1207 :     Started adding some infrastructure for supporting 64-bit int- and
1208 :     word-types. (Still in its very early stages.)
1209 :    
1210 :     ----------------------------------------------------------------------
1211 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1212 : mblume 1672 Date: 2004/10/28 10:45:00 CDT
1213 :     Tag: Release_110_50
1214 :     Description:
1215 :    
1216 :     New working version (110.50). NEW BOOTFILES!
1217 :    
1218 :     =====================
1219 :    
1220 :     Also:
1221 :    
1222 :     - Changend config/srcarchiveurl from a file just
1223 :     containing the URL string into a file containing
1224 :     shell script code. The code has access to the $VERSION variable.
1225 :     - Made corresponding changes to config/install.sh and config/unpack.
1226 :     - Default contents of config/srcarchiveurl uses $VERSION and
1227 :     normally does not have to be edited to reflect a version change.
1228 :    
1229 :     (As a result, a version change can be done by just editing
1230 :     config/version, the rest is now automatic.)
1231 :    
1232 :     ----------------------------------------------------------------------
1233 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1234 : mblume 1669 Date: 2004/10/27 17:50:00 CDT
1235 :     Tag: blume-20041027-btrace-msg
1236 :     Description:
1237 :    
1238 :     BackTrace.monitor now also reports the source of the exception that
1239 :     triggered the trace.
1240 :    
1241 :     ----------------------------------------------------------------------
1242 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1243 : mblume 1668 Date: 2004/10/27 17:20:00 CDT
1244 :     Tag: blume-20041027-x86-c-calls
1245 :     Description:
1246 :    
1247 :     This is the HISTORY entry for two earlier commits, both concerning
1248 :     the x86 c-calls code in MLRISC:
1249 :    
1250 :     - added a missing LOAD in the code that deals with struct arguments
1251 :     - made sure the caller does not add the wrong number of bytes to the
1252 :     stack pointer after a call of a function returning a struct
1253 :     (the callee already pops the implicit argument which points to
1254 :     the space reserved for the result)
1255 :    
1256 :     ----------------------------------------------------------------------
1257 : allenleung 1662 Name: Allen Leung (leunga (at) reservoir (dot) com)
1258 :     Date: 2004/10/24 14:00:00 EST
1259 :     Tag: leunga-20041024-x86-gas-fucomip
1260 :     Description:
1261 :    
1262 :     John discovered a bug in the syntax of fucomip.
1263 :     The opcodes FU?COMIP? have been changed to
1264 :    
1265 :     fu?comip? %st(i), %st
1266 :    
1267 :     ----------------------------------------------------------------------
1268 : mblume 1622 Name: Matthias Blume (blume (at) tti - c (dot) org)
1269 : mblume 1656 Date: 2004/10/20 15:06:00 CDT
1270 :     Tag: blume-20041020-standalone-backtrace
1271 :     Description:
1272 :    
1273 :     Added a mechanism for getting back-trace information from standalone
1274 :     programs. Here is how it works:
1275 :    
1276 :     1. The part of the program from which you want to get backtrace
1277 :     information (usually the whole program) should be wrapped with
1278 :     BackTrace.monitor. This is a (unit->'a)->'a function, and your
1279 :     main program could be modified from something like
1280 :    
1281 :     fun main (pgm, args) = ...
1282 :    
1283 :     to
1284 :    
1285 :     fun main (pgm, args) = BackTrace.monitor (fn () => ...)
1286 :    
1287 :     2. To be able to access BackTrace.monitor, you have to add
1288 :     library $smlnj-tdp/plugins.cm to the .cm file that contains your
1289 :     main function.
1290 :    
1291 :     3. Remove all compiled code (i.e., all the .cm/ subdirectories that
1292 :     CM might have created in the past for your project).
1293 :    
1294 :     4. Build the system using this command line:
1295 :    
1296 :     ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
1297 :     myprog.cm MyProg.main myprog
1298 :    
1299 :     instead of the usual
1300 :    
1301 :     ml-build myprog.cm MyProg.main myprog
1302 :    
1303 :     I changed a library name:
1304 :    
1305 :     $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
1306 :    
1307 :     New libraries:
1308 :    
1309 :     $smlnj-tdp/back-trace.cm
1310 :     -- when loaded causes the back-trace plugin to be installed
1311 :     $smlnj-tdp/coverage.cm
1312 :     -- when loaded causes the coverage plugin to be installed
1313 :    
1314 :     ----------------------------------------------------------------------
1315 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1316 : mblume 1655 Date: 2004/10/18 16:45:00 CDT
1317 :     Tag: blume-20041018-groupowner
1318 :     Description:
1319 :    
1320 :     Added an "obsolete" warning for the "group owner" syntax to CM's parser.
1321 :    
1322 :     Eliminated group owner specs from .cm files throughout the source tree.
1323 :    
1324 :     ----------------------------------------------------------------------
1325 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1326 : mblume 1653 Date: 2004/10/15 15:45:00 CDT
1327 :     Tag: blume-20041015-coverage
1328 :     Description:
1329 :    
1330 :     * Test coverage tool added!
1331 :     * Further reorganization of tracing-, debugging-, and profiling support:
1332 :    
1333 :     - moved original BTImp -- now called BackTrace -- into a separate
1334 :     library called $/trace-debug-profile.cm
1335 :     - eliminated all mentions of BTrace from SMLofNJ.Internals
1336 :     - only the instrumentation mechanism is now left in the compiler proper
1337 :     - BackTrace module is a plugin which is NOT plugged in by default
1338 :     - Coverage module is another such plugin
1339 :    
1340 :     To get the benefits of any of these plugin modules, the code in
1341 :     question must be compiled with tdp instrumentation turned on. This
1342 :     can be done by setting SMLofNJ.Internals.TDP.mode to true.
1343 :     (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
1344 :    
1345 :     Plugins are selected at link time. (Pre-compiled instrumented code
1346 :     can be re-loaded with different plugins in effect.) When an
1347 :     instrumented module is linked, whatever plugins are at that time
1348 :     enabled will come into effect for that module.
1349 :    
1350 :     To enable the back-trace plugin, load library $/trace-debug-profile.cm
1351 :     and invoke BackTrace.install() (e.g., from the interactive prompt).
1352 :     To enable the coverage plugin, load the same library and invoke
1353 :     Coverage.install().
1354 :    
1355 :     Back-traces are generated automatically on uncaught exceptions and
1356 :     when the code in question explicitly invokes BackTrace.trigger().
1357 :    
1358 :     Coverage (and execution frequency-) information must be queried
1359 :     explicitly by calling Coverage.not_covered and Coverage.hot_spots.
1360 :    
1361 :     ----------------------------------------------------------------------
1362 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1363 : mblume 1652 Date: 2004/10/14 17:40:00 CDT
1364 :     Tag: blume-20041014-tdp-core
1365 :     Description:
1366 :    
1367 :     Snapshot of a significant overhaul of how the trace/debug/profile support
1368 :     is hooked into the system (specifically: Core and SMLofNJ.Internals).
1369 :    
1370 :     ----------------------------------------------------------------------
1371 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1372 : mblume 1650 Date: 2004/10/13 16:34:00 CDT
1373 :     Tag: blume-20041013-tdp
1374 :     Description:
1375 :    
1376 :     Some rationalization of names:
1377 :    
1378 :     structure BTrace -> structure TDPInstrument
1379 :     etc.
1380 :    
1381 :     This is is preparation of using the original back-trace
1382 :     instrumentation for other purposes. "TDP" stands for
1383 :     Trace/Debug/Profile.
1384 :    
1385 :     The control flag controlling whether instrumentation is on or off is now
1386 :     registered under a different name, so instead of running sml as
1387 :    
1388 :     sml -Cinstrument.btrace-mode=true
1389 :    
1390 :     one has to say
1391 :    
1392 :     sml -Ctdp.instrument=true
1393 :    
1394 :     ----------------------------------------------------------------------
1395 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1396 : mblume 1645 Date: 2004/10/11 16:37:00 CDT
1397 :     Tag: blume-20041011-regions
1398 :     Description:
1399 :    
1400 :     Made some minor modifications to elabcore.sml to have source regions
1401 :     be propagated more tightly -- resulting in better (i.e., smaller)
1402 :     regions being reported in error- and debug messages.
1403 :    
1404 :     ----------------------------------------------------------------------
1405 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1406 : mblume 1644 Date: 2004/10/08 22:50:00 CDT
1407 :     Tag: blume-20041008-cmkw
1408 :     Description:
1409 :    
1410 :     Fixed handling of keywords in .cm files: After seeing "is" the lexer
1411 :     treats subsequent occurrences of "group", "library", "source", "is",
1412 :     "*", and "-" as ordinary identifiers rather than keywords.
1413 :    
1414 :     Most seriously, this fixes a problem with CM's "shell" tool. The tool
1415 :     is supposed to accept a tool argument called "source", but this did
1416 :     not work because of the clash with the keyword.
1417 :    
1418 :     ----------------------------------------------------------------------
1419 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1420 : mblume 1642 Date: 2004/10/07 16:00:00 CDT
1421 :     Tag: blume-20041007-cleanup
1422 :     Description:
1423 :    
1424 :     Assorted cleanup work:
1425 :    
1426 :     - got rid of intstrmap in favor of using the library's
1427 :     hash table implementation
1428 :     - threw out most of the pathnames stuff, as it was not used anyway
1429 :     - simplified tokentable implementation
1430 :     - fixed some minor spelling errors
1431 :    
1432 :     ----------------------------------------------------------------------
1433 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1434 : mblume 1641 Date: 2004/10/06 15:15:15 CDT
1435 :     Tag: blume-20041006-handler
1436 :     Description:
1437 :    
1438 :     Cleaned up the absyn to reflect the invariant that HANDLE always
1439 :     carries a FNexp as part of the type definition. This eliminates some
1440 :     superfluous sanity checks at runtime down the road.
1441 :    
1442 :     Some minor cleanup of the btrace code.
1443 :    
1444 :     ----------------------------------------------------------------------
1445 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1446 : mblume 1640 Date: 2004/10/01 10:20:30 CDT
1447 :     Tag: blume-20041001-slave
1448 :     Description:
1449 :    
1450 :     Added hack to make slave mode work in the presence of the version
1451 :     tool. (Still, since the master does two passes over the code for
1452 :     CMB.make, the release number gets bumped twice when slaves are
1453 :     attached. I don't know if this is worth fixing...)
1454 :    
1455 :     ----------------------------------------------------------------------
1456 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1457 : mblume 1639 Date: 2004/09/30 10:55:00 CDT
1458 :     Tag: blume-20040930-version
1459 :     Description:
1460 :    
1461 :     * Moved the "version" magic into its own little library under
1462 : mblume 1672 src/system/smlnj/internal. This avoids expensive reconstruction of
1463 :     a stable src/compiler/core.cm.
1464 : mblume 1639
1465 : mblume 1672 * At the same time, structure CompilerVersion is now known as
1466 :     structure SMLNJVersion.
1467 : mblume 1639
1468 :     * Arranged for the version tool to NOT kick in when rebuilding the system
1469 :     (makeml -rebuild, fixpt). Otherwise one would never reach a fixpoint.
1470 :     Also, loading the versiontool does not work when rebuilding the system
1471 :     because CM is not properly initialized at that time.
1472 :    
1473 :     ----------------------------------------------------------------------
1474 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1475 : mblume 1638 Date: 2004/09/29 14:00:00 CDT
1476 :     Tag: blume-20040929-autoversion
1477 :     Description:
1478 :    
1479 :     Implemented some CM magic to have
1480 :     file src/compiler/TopLevel/main/version.sml
1481 :     generated automagically.
1482 :     The version is taken from two files: config/version and config/release.
1483 :     The first is expected to contain a two-part version number such as 110.49.
1484 :     The second should contain a single number, but it may be missing.
1485 :    
1486 :     If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
1487 :     time the version tool is loaded (which is the first time you say CMB.make),
1488 :     then the tool will increment the value stored in config/release every
1489 :     time CMB.make is invoked.
1490 :    
1491 :     The binfile format is now insensitive to anything beyond the first
1492 :     two components of a version number, so bumping the release does not render
1493 :     binfiles incompatible. Auto-bumping can be used to keep track of versions
1494 :     during development without invalidating existing binfiles.
1495 :    
1496 :     In any case, every CMB.make updates the date information in version.sml.
1497 :     (This is the date that is printed in the banner.)
1498 :    
1499 :     ----------------------------------------------------------------------
1500 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1501 : mblume 1634 Date: 2004/09/28 10:53:00 CDT
1502 :     Tag: blume-20040928-controls
1503 :     Description:
1504 :    
1505 :     Some cleanup of the controls code.
1506 :    
1507 :     ----------------------------------------------------------------------
1508 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1509 : mblume 1633 Date: 2004/09/27 22:08:00 CDT
1510 :     Tag: blume-20040927-controls
1511 :     Description:
1512 :    
1513 :     Added two pieces of functionality to the Controls interface:
1514 :    
1515 :     1. val save'restore: 'a control -> unit -> unit
1516 :    
1517 :     grabs the current value of the control in stage 1 and restores it
1518 :     in stage 2.
1519 :    
1520 :     2. val set' : 'a control * 'a -> unit -> unit
1521 :    
1522 :     stores the given value into the control in stage 2 (i.e., delayed)
1523 :     but does all error checking in stage 1.
1524 :     (This is for string controls that need to do parse their argument --
1525 :     something that might fail. In some cases, notably in CM, one
1526 :     already knows the intended argument but wants to delay the actual
1527 :     assignment until a time when error recovery would be more difficult.)
1528 :    
1529 :     Changed the handling of controls in tool arguments to classes "sml" and
1530 :     "lazysml":
1531 :     - use Controls.save'restore as a more robust way of restoring the
1532 :     old value (in particular: without having to re-parse the string)
1533 :     - use controls to handle the "overload" keyword in the init group
1534 :     (I believe this change actually fixes a long-standing obscure bug.)
1535 :    
1536 :     ----------------------------------------------------------------------
1537 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1538 : mblume 1632 Date: 2004/09/27 17:00:00 CDT
1539 :     Tag: blume-20040927-lazysml
1540 :     Description:
1541 :    
1542 :     Added a new tool class called "lazysml" to CM's tool chest. The only
1543 :     difference to "sml" is that compilation is done with Control.lazysml
1544 :     set to true. A source of class "lazysml" is automatically recognized
1545 :     by a file name suffix of ".lml".
1546 :    
1547 :     In addition to the above feature, the original class "sml" now also
1548 :     supports a tool argument "lazy" which has the same effect. As a
1549 :     result, the following three lines are equivalent:
1550 :    
1551 :     foo.sml : lazysml
1552 :     foo.sml : sml (lazy)
1553 :     foo.sml (lazy)
1554 :    
1555 :     The setting goes into effect both during parsing and during
1556 :     compilation. The original setting is restored right after parsing and
1557 :     after compilation, respectively.
1558 :    
1559 :     In addition to all the above, there is also a general mechanism to set
1560 :     ANY of the "controls" that are available at the command line via
1561 :     "-C..." on a per-sml-file basis. The same rules that apply for "lazy"
1562 :     apply as well. (In fact, "lazy" is implemented as a special case of
1563 :     the general mechanism.)
1564 :    
1565 :     The .cm file syntax uses a new keyword tool argument called "with".
1566 :     There are several ways of indicating the desired settings:
1567 :    
1568 :     foo.sml (with:parser.quotations=true)
1569 :     foo.sml (with:(name:parser.quotations value:true))
1570 :     foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
1571 :     foo.sml (with:(name1=value1 name2=value2 ...))
1572 :     foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
1573 :    
1574 :     etc.
1575 :    
1576 :     Another possible abbreviation is to leave out the =v or value:v part
1577 :     if the name refers to a boolean control (in which case the value is
1578 :     taken to be true). Thus, one could get lazy sml also by saying:
1579 :    
1580 :     foo.sml (with:parser.lazy-keyword=true)
1581 :     foo.sml (with:parser.lazy-keyword)
1582 :     foo.sml (with:(name:parser.lazy-keyword value:true))
1583 :     foo.sml (with:(name:parser.lazy-keyword))
1584 :    
1585 :     ----------------------------------------------------------------------
1586 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1587 : mblume 1631 Date: 2004/09/24 16:22:00 CDT
1588 :     Tag: blume-20040924-ppc-long-branch
1589 :     Description:
1590 :    
1591 :     Turned message about "emiting long form of branch" off by default.
1592 :     Added a control flag to turn it back on when desired.
1593 :    
1594 :     ----------------------------------------------------------------------
1595 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1596 : mblume 1630 Date: 2004/09/24 16:05:00 CDT
1597 :     Tag: blume-20040924-rounding
1598 :     Description:
1599 :    
1600 :     Applied patch for setting rounding modes under Mac OS X. Thanks to
1601 :     Melissa O'Neill for providing the code!
1602 :    
1603 :     ----------------------------------------------------------------------
1604 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1605 : mblume 1629 Date: 2004/09/23 17:30:00 CDT
1606 :     Tag: blume-20040923-envvars
1607 :     Description:
1608 :    
1609 :     1. Changed definition of type ControlRegistry.registry_tree to
1610 :     include control_info (i.e., the name of the controlling environment
1611 :     variable).
1612 :    
1613 :     2. Added command-line flags -e and -E to print the names of environment
1614 :     variables that can be used to control internal settings.
1615 :     (This uses the new API mentioned in 1.)
1616 :    
1617 :     ----------------------------------------------------------------------
1618 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1619 : mblume 1622 Date: 2004/09/13 16:50:00 CDT
1620 :     Tag: Release_110_49
1621 :     Description:
1622 :    
1623 :     New working version (110.49). NEW BOOTFILES!
1624 :    
1625 :     ----------------------------------------------------------------------
1626 : mblume 1621 Name: Matthias Blume
1627 :     Date: 2004/09/13 16:20:00 CDT
1628 :     Tag: blume-20040913-config-mlrisc
1629 :     Description:
1630 :    
1631 :     Put target "mlrisc" back into the default list.
1632 :     (There is no harm in having it, and some users have expressed their
1633 :     wish to have "mlrisc" included by default.)
1634 :    
1635 :     ----------------------------------------------------------------------
1636 : jhr 1620 Name: John Reppy
1637 :     Date: 2004/09/13
1638 :     Tag: jhr-20040913-signals
1639 :     Description:
1640 :    
1641 :     Fixed the signal masking code to properly nest mask/unmask operations
1642 :     on a per-signal basis.
1643 :    
1644 :     ----------------------------------------------------------------------
1645 : mblume 1619 Name: Matthias Blume (blume (at) tti - c (dot) org)
1646 :     Date: 2004/09/08 13:20:00 CDT
1647 :     Tag: blume-20040908-heap-magic
1648 :     Description:
1649 :    
1650 :     Bumped the heap macig number to 0x09082004 to account for the changed
1651 :     layout of the ML frame under MacOS X.
1652 :    
1653 :     ----------------------------------------------------------------------
1654 : allenleung 1618 Name: Allen Leung (leunga (at) reservoir (dot) com)
1655 :     Date: 2004/09/03 11:26:00 EST
1656 :     Tag: leunga-20040903-cygwin-install
1657 :     Description:
1658 :    
1659 :     Added a patch to _arch-n-opsys to enable the Cygwin runtime.
1660 :     The Cygwin runtime is turned on by setting the environment
1661 :     variable SMLNJ_CYGWIN_RUNTIME to 1.
1662 :    
1663 :     ----------------------------------------------------------------------
1664 : mblume 1615 Name: Matthias Blume (blume (at) tti - c (dot) org)
1665 : mblume 1617 Date: 2004/08/31 17:15:00 CDT
1666 :     Tag: blume-20040831-core
1667 :     Description:
1668 :    
1669 :     Added some exports to src/compiler/core.cm upon request by J. Joemann.
1670 :    
1671 :     ----------------------------------------------------------------------
1672 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1673 : mblume 1616 Date: 2004/08/30 17:55:00 CDT
1674 :     Tag: blume-20040830-installer
1675 :     Description:
1676 :    
1677 :     Upon request by Johannes Joemann:
1678 :    
1679 :     - improved ML code of installer to fall back to coping when renaming
1680 :     fails (i.e., when source and target are on different file systems);
1681 :     the code compiles but has yet to be tested in anger
1682 :     - removed mlrisc from list of default targets (config/targets)
1683 :    
1684 :     ----------------------------------------------------------------------
1685 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1686 : mblume 1615 Date: 2004/08/27 17:20:00 CDT
1687 :     Tag: blume-20040827-ptreql
1688 :     Description:
1689 :    
1690 :     Added ptreql primop to structure InlineT (upon request from Larry
1691 :     Paulson).
1692 :    
1693 :     ----------------------------------------------------------------------
1694 : allenleung 1613 Name: Allen Leung (leunga (at) reservoir (dot) com)
1695 : allenleung 1614 Date: 2004/08/15 21:21:00 EST
1696 :     Tag: leunga-110_48-udgraph
1697 :     Description:
1698 :    
1699 :     Another bug fix from Carl Hauser:
1700 :    
1701 :     diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
1702 :     > 48c48
1703 :     > < | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
1704 :     > e::L)
1705 :     > ---
1706 :     > > | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
1707 :     > Without this, any deletion of an edge in an undirected graph does severe
1708 :     > violence to the graph.
1709 :    
1710 :     ----------------------------------------------------------------------
1711 :     Name: Allen Leung (leunga (at) reservoir (dot) com)
1712 : allenleung 1613 Date: 2004/08/10 23:35:00 EST
1713 :     Tag: leunga-110_48-ppc
1714 :     Description:
1715 :    
1716 :     The IBM/MacOS syntax switch on PPC was incorrectly swapped.
1717 :    
1718 :     ----------------------------------------------------------------------
1719 : mblume 1602 Name: Matthias Blume (blume (at) tti - c (dot) org)
1720 : mblume 1615 Date: 2004/08/10 12:00:00 CDT
1721 : mblume 1602 Tag: Release_110_48
1722 :     Description:
1723 :    
1724 :     New working version (110.48). NEW BOOTFILES!
1725 :    
1726 :     ----------------------------------------------------------------------
1727 : allenleung 1599 Name: Allen Leung (leunga (at) reservoir (dot) com)
1728 :     Date: 2004/08/09 12:21:00 EST
1729 :     Tag: leunga-110_47-dijsktra
1730 :     Description:
1731 :    
1732 :     Bug fix from Carl Hauser:
1733 :    
1734 : blume 2155 single_source_shortest_paths in dijkstra.sml was observed to get wrong
1735 : allenleung 1599 answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1736 :    
1737 : blume 2155 The problem is that following the expression A.update(dist,s,Num.zero)
1738 : allenleung 1599 it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1739 :    
1740 :     ----------------------------------------------------------------------
1741 : mblume 1595 Name: Matthias Blume (blume (at) tti - c (dot) org)
1742 :     Date: 2004/08/06 18:10:00 CDT
1743 :     Tag: blume-20040806-cmdline
1744 :     Description:
1745 :    
1746 :     Fiddled with handling of command-line options:
1747 :    
1748 :     * sml now quits after processing the command line
1749 :     if -H, -S, -h<n>, or -s<n> appears as the last
1750 :     command-line argument
1751 :     * a new option -q terminates the session when encountered on
1752 :     the command line; subsequent arguments will be ignored
1753 :     * bug fixes: short (erroneous) arguments are no longer ignored
1754 :     completely
1755 :    
1756 :     ----------------------------------------------------------------------
1757 : allenleung 1593 Name: Allen Leung (leunga (at) reservoir (dot) com)
1758 :     Date: 2004/08/04 18:17:00 EST
1759 :     Tag: leunga-110_47-ppc-ibm-asm
1760 :     Description:
1761 :    
1762 :     - Added minimal IBM assembly syntax support for PowerPC.
1763 :    
1764 : blume 2155 - Cygwin: manually changed the file cygwin.def. Some exported symbols have
1765 : allenleung 1593 been altered in the runtime. We need an automatic way to keep the file
1766 :     in sync.
1767 :    
1768 :     ----------------------------------------------------------------------
1769 : mblume 1548 Name: Matthias Blume (blume (at) tti - c (dot) org)
1770 : mblume 1581 Date: 2004/08/04 14:00:00 CDT
1771 :     Tag: Release_110_47
1772 :     Description:
1773 :    
1774 :     New working version (110.47). NEW BOOTFILES!
1775 :    
1776 :     ----------------------------------------------------------------------
1777 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1778 : mblume 1580 Date: 2004/08/03 14:25:00 CDT
1779 :     Tag: blume-20040803-callingconv
1780 :     Description:
1781 :    
1782 :     Added low-level support for choosing C calling conventions by
1783 :     twiddling the type of rawccall. (See
1784 :     src/compiler/Semant/types/cproto.sml for details.)
1785 :    
1786 :     ----------------------------------------------------------------------
1787 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1788 : mblume 1579 Date: 2004/08/02 15:55:00 CDT
1789 :     Tag: blume-20040802-backout
1790 :     Description:
1791 :    
1792 :     Backed out of change to win32-filesys.c. The earlier patch to
1793 :     get_file_time caused CM to produce files with the wrong time stamp.
1794 :    
1795 :     ----------------------------------------------------------------------
1796 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1797 : mblume 1573 Date: 2004/08/02 14:45:00 CDT
1798 :     Tag: blume-20040802-nlffi-win32
1799 :     Description:
1800 :    
1801 :     Added NLFFI support for Win32, adapted from a patch provided by David
1802 :     Hansel. This is currently completely untested. Also, the issue
1803 :     concerning stdcall vs. ccall is still unresolved.
1804 :    
1805 :     ----------------------------------------------------------------------
1806 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1807 : mblume 1572 Date: 2004/07/30 17:55:00 CDT
1808 :     Tag: blume-20040730-various
1809 :     Description:
1810 :    
1811 :     Gearing up towards 110.47...
1812 :    
1813 :     - various minor bugfixes to ml-nlffigen
1814 :     - a beginning of a manual for nlffi
1815 :    
1816 :     - eliminated 'export name=value' in config/install.sh as this does
1817 :     not work with certain versions of /bin/sh
1818 :     (Thanks to David King at Motorola for catching this.)
1819 :    
1820 :     - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1821 :     - added a test for tm==NULL to gmtime.c and localtime.c
1822 :     - applied patch for incorrect GetFileTime under win32
1823 :     - toSeconds -> toMilliseconds in Win32/win32-process.sml
1824 :    
1825 :     ----------------------------------------------------------------------
1826 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1827 : mblume 1557 Date: 2004/07/21 18:20:00 CDT
1828 :     Tag: blume-20040721-nlffigen
1829 :     Description:
1830 :    
1831 :     - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1832 :     for a typedef to an incomplete type, but leave out the "typ" member.
1833 :     (This is just for consistency.)
1834 :     - Started to produce what is supposed to become better (i.e., comprehensive)
1835 :     documentation of what ml-nlffigen does and produces.
1836 :    
1837 :     ----------------------------------------------------------------------
1838 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1839 : mblume 1548 Date: 2004/07/14 16:25:00 CDT
1840 :     Tag: blume-20040714-union
1841 :     Description:
1842 :    
1843 :     Added C_UNION to c-calls/c-types.sml and updated the machinery
1844 :     (ml-nlffigen, cproto.sml) that conveys C function interface
1845 :     information to the code generator.
1846 :    
1847 :     However, the actual architecture-specific implementation of function
1848 :     arguments and results that are C unions is still not implemented.
1849 :    
1850 :     ----------------------------------------------------------------------
1851 : allenleung 1539 Name: Allen Leung (leunga (at) reservoir (dot) com)
1852 : allenleung 1546 Date: 2004/07/14 14:38:00 EST
1853 :     Tag: leunga-110_46_1-ppc-lwzu
1854 :     Description:
1855 :    
1856 :     Added these instructions to the PowerPC architecture:
1857 :     LBZU(X), LHZU(X), LWZU(X),
1858 : blume 2155 STWU(X), STFDU, STFSU
1859 :    
1860 : allenleung 1546 etc...
1861 :    
1862 :     Note: I haven't added their instruction encoding into the description.
1863 :    
1864 :     ----------------------------------------------------------------------
1865 :     Name: Allen Leung (leunga (at) reservoir (dot) com)
1866 : allenleung 1539 Date: 2004/07/13 15:04:00 EST
1867 :     Tag: leunga-110_46_1-ppc-lwarx
1868 :     Description:
1869 :    
1870 :     Added the two instructions LWARX and STWCX to the PowerPC
1871 : blume 2155 instruction set.
1872 : allenleung 1539
1873 :     A (untested) rewrite of loop-structure.sml. The old version
1874 :     is completely broken.
1875 :    
1876 :     ----------------------------------------------------------------------
1877 : mblume 1479 Name: Matthias Blume (blume (at) tti - c (dot) org)
1878 : mblume 1537 Date: 2004/07/13 13:50:00 CDT
1879 :     Tag: blume-20040713-nlffi
1880 :     Description:
1881 :    
1882 :     - use paramAlloc to report c-calls with too many arguments
1883 :     (for PPC version where parameter area is pre-allocated)
1884 :     - added ccall_maxargspace to machspec (to implement the above)
1885 :     - made "make" commend in CM's "make" tool configurable
1886 :     - added option (default: on) for passing the name of the SML/NJ's "bin"
1887 :     directory to "make"; the call looks like this:
1888 :    
1889 :     make <options> SMLNJ_BINDIR=<dir> <target>
1890 :    
1891 :     This can be used by the Makefile to, e.g., pick the "right" version
1892 :     of ml-nlffigen.
1893 :     - minor code tweaks
1894 :    
1895 :     ----------------------------------------------------------------------
1896 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1897 : mblume 1532 Date: 2004/07/12 22:50:00 CDT
1898 :     Tag: blume-110_46_1-macosx-nlffi
1899 :     Description:
1900 :    
1901 :     NLFFI under Mac OS X now working (sort of). This is largely untested,
1902 :     though.
1903 :    
1904 :     Note: 1. You have to make a new, clean build of the runtime system.
1905 :     2. There are new BOOTFILES, you have to use them!
1906 :     (Doing the bootstrap process yourself would be *very* painful!
1907 :     If you absolutely have to do it, build the system under
1908 :     a different architecture and then cross-compile.)
1909 :    
1910 :     Version bumped to 110.46.1 to account for runtime data format changes.
1911 :    
1912 :     ----------------------------------------------------------------------
1913 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1914 : mblume 1515 Date: 2004/06/18 14:30:00 CDT
1915 :     Tag: blume-20040618-unix
1916 :     Description:
1917 :    
1918 :     Changed the implementation of structure Unix so that the same stream
1919 :     is returned every time one of the {text,bin}{In,Out}streamOf functions
1920 :     is invoked on the same proc. This is not what the spec currently
1921 :     says -- although IMO it arguably should. (See discussion below.)
1922 :    
1923 :     ----------------------------------------------------------------------
1924 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1925 : mblume 1507 Date: 2004/06/17 18:15:00 CDT
1926 :     Tag: Release_110_46
1927 :     Description:
1928 :    
1929 :     New working version (110.46). NEW BOOTFILES!
1930 :    
1931 :     ----------------------------------------------------------------------
1932 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1933 : mblume 1506 Date: 2004/06/17 17:20:00 CDT
1934 :     Tag: blume-20040617-timer-unix
1935 :     Description:
1936 :    
1937 :     Changed the interface of structures Timer and Unix to match the most
1938 :     recent Basis spec.
1939 :    
1940 :     In the case of Unix there still seems to be an open/weird issue:
1941 :    
1942 :     The {text,bin}{In,Out}streamOf functions are supposed to create
1943 :     fresh streams whenever they are called -- as opposed to have them
1944 :     return the same stream every time. This design is supposed to
1945 :     prevent space leaks caused by proc values hanging on to streams.
1946 :    
1947 :     The reap function, on the other hand, is supposed to close the
1948 :     streams. This cannot be done without having a handle on the
1949 :     stream in proc after all...
1950 :    
1951 :     I took the liberty to implement the following stopgap solution:
1952 :    
1953 :     The proc value hangs on to the most recently created stream(s).
1954 :     Reap closes those. If either or both of the two streams hadn't
1955 :     been created at all yet, then reap will close the corresponding
1956 :     file descriptors directly.
1957 :    
1958 :     PS: I don't understand the original space leak argument anymore. If
1959 :     a proc hangs on to the imperative stream, then I/O operations on those
1960 :     will advance the state of the cached stream and avoid the space leak.
1961 :    
1962 :     ----------------------------------------------------------------------
1963 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1964 : mblume 1500 Date: 2004/05/28 16:45:00 CDT
1965 :     Tag: blume-20040528-basis
1966 :     Description:
1967 :    
1968 :     Added signature PACK_REAL and exported functor PrimIO.
1969 :    
1970 :     ----------------------------------------------------------------------
1971 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1972 : mblume 1498 Date: 2004/05/25 16:00:00 CDT
1973 :     Tag: blume-20040525-group-owner
1974 :     Description:
1975 :    
1976 :     CM now ignores (but still accepts) the "owner" information in group
1977 :     descriptions. The owner of a group is its next enclosing
1978 :     library. Each group must have a unique owner. (There is a virtual
1979 : mblume 1505 "toplevel" library that owns groups which are not nested within a real
1980 : mblume 1498 library.) Previously, each group had to explicitly declare its owner,
1981 :     and CM would check that such a declaration is correct. The new scheme
1982 :     is to have CM check that for each group there is precisely one owning
1983 :     library.
1984 :    
1985 :     The advantage of the new scheme is that the programmer no longer needs
1986 :     to maintain the somewhat annoying owner information. The downside is
1987 :     that CM cannot enforce the ownership rule across multiple runs of
1988 :     CM.make. Fortunately, enclosing the same group in two different
1989 :     libraries A and B which are not part of the same program does not
1990 :     cause real problems.
1991 :    
1992 :     ----------------------------------------------------------------------
1993 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1994 : mblume 1496 Date: 2004/05/20 16:00:00 CDT
1995 :     Tag: blume-20040520-win32
1996 :     Description:
1997 :    
1998 :     Made the win32 version work again. (Strangely, a misplaced comma had
1999 :     slipped into win32-process.c which prevented the runtime from being
2000 :     compiled correctly.)
2001 :    
2002 :     Also, included a minor addition to ml-build.bat analogous to what was
2003 :     done in blume-20040519-ml-build.
2004 :    
2005 :     ----------------------------------------------------------------------
2006 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2007 : mblume 1494 Date: 2004/05/19 22:10:00 CDT
2008 :     Tag: blume-20040519-ml-build
2009 :     Description:
2010 :    
2011 :     Arranged for ml-build to clean up after itself a little bit better.
2012 :     The script generates a temporary SML source file and compiles it using
2013 :     CM, so CM generates metadata (GUID, SKEL, objectfile) for it. It now
2014 :     gets rid of those at the end, so they don't accumulate under .cm.
2015 :    
2016 :     This required a minor change to install.sh because the name of the
2017 :     metadata directory (default: .cm) is actually configurable at
2018 :     installation time.
2019 :    
2020 :     ----------------------------------------------------------------------
2021 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2022 : mblume 1491 Date: 2004/05/18 15:50:00 CDT
2023 :     Tag: blume-20040518-mkreader
2024 :     Description:
2025 :    
2026 :     Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
2027 :     implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
2028 :    
2029 :     ----------------------------------------------------------------------
2030 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2031 : mblume 1489 Date: 2004/05/11 14:35:00 CDT
2032 :     Tag: blume-20040511-win32sock
2033 :     Description:
2034 :    
2035 :     Added previously missing support for many socket-related functions
2036 :     under win32. Thanks to David Hansel <hansel@reactive-systems.com>
2037 :     for the voluminous patch!
2038 :    
2039 :     (I have not tested this patch under win32 yet.)
2040 :    
2041 : mblume 1490 Here is David's e-mail:
2042 :    
2043 :     Hi,
2044 :    
2045 :     Attached to this email you find a diff against sml/nj 110.45
2046 :     that will enable socket support under Windows.
2047 :    
2048 :     To apply the patch (using unix or cygwin)
2049 :     1) gunzip runtime.diff.gz
2050 :     2) "cd" into "src/runtime" in the source tree of a fresh
2051 :     110.45 installation.
2052 :     3) patch -p 1 < [your/path/to]runtime.diff
2053 :    
2054 : blume 2155 The code compiles fine but has NOT yet been extensively tested.
2055 :     I only ran a few tests for basic socket client functionality
2056 :     (which worked fine). Especially the functions that use ioctl
2057 : mblume 1490 are not tested at all and might not work (see below).
2058 :    
2059 :     I implemented this since we want to move to a newer version of sml/nj
2060 :     but need socket support in order to use it. This is the first time I
2061 :     even had a look at the sml/nj source, so please review my changes
2062 :     before making this part of the distribution! Here are a few issues
2063 : blume 2155 that I think might be better for someone to solve who is more
2064 : mblume 1490 familiar with the sml/nj source (and socket programming):
2065 :    
2066 :     - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
2067 :     exception since I could not figure out what the windows equivalent
2068 :     of these functions is
2069 :    
2070 :     - In sockets-osdep.h there are a some #include statements that are
2071 :     only used in a few files that include sockets-osdep.h
2072 :    
2073 :     - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
2074 :     does not process its return value since I don't know how to
2075 :     report an error upwards.
2076 :    
2077 :     - It would probably be good to have a call to WSACleanup() when
2078 : blume 2155 the library is unloaded (if there is such a possibility).
2079 : mblume 1490 Otherwise I think Windows will take care of this automatically
2080 :     when the process finishes.
2081 :    
2082 :     - I used ioctlsocket() as a replacement for ioctl() but I have
2083 :     no idea if that is actually the proper replacement on Windows.
2084 :    
2085 :     - All these issues are marked in the code by "FIXME" comments.
2086 :    
2087 : blume 2155 We use sml/nj extensively in our products and are quite happy
2088 :     with it. I hope this contribution will help you.
2089 : mblume 1490
2090 :     Keep up the good work!
2091 :    
2092 :     David
2093 :    
2094 : mblume 1489 ----------------------------------------------------------------------
2095 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2096 : mblume 1488 Date: 2004/05/11 14:20:00 CDT
2097 :     Tag: blume-20040511-installml
2098 :     Description:
2099 :    
2100 :     Fixed two bugs in installml script. (Thanks to Vesa A. Norrman for
2101 :     the patch.)
2102 :    
2103 :     ----------------------------------------------------------------------
2104 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2105 : mblume 1487 Date: 2004/05/11 14:05:00 CDT
2106 :     Tag: blume-20040511-nlffi-netbsd
2107 :     Description:
2108 :    
2109 :     Added support for nlffi under netbsd. (Thanks to Vesa A. Norrman for
2110 :     the patch.)
2111 :    
2112 :     ----------------------------------------------------------------------
2113 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2114 : mblume 1486 Date: 2004/05/11 12:05:00 CDT
2115 :     Tag: blume-20040511-exports
2116 :     Description:
2117 :    
2118 :     As per request by Adam Chlipala <adam@hcoop.net>, extended various
2119 :     export lists in compiler-related .cm-files.
2120 :    
2121 :     ----------------------------------------------------------------------
2122 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2123 : mblume 1485 Date: 2004/05/11 11:35:00 CDT
2124 :     Tag: blume-20040511-allsource
2125 :     Description:
2126 :    
2127 :     The installer now honors the "src-smlnj" target again, although its meaning
2128 :     has changed from "all sources required for the compiler" to "all sources
2129 :     the installer knows about". In other words, if you enable "src-smlnj"
2130 :     in the "targets" file, then the installer will pull in sources for
2131 :     everything. (Notice that this refers to source code only. Compiled
2132 :     code is still only installed for modules that were requested explicitly
2133 :     or which are required for other modules that were requested explicitly.)
2134 :    
2135 :     ----------------------------------------------------------------------
2136 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2137 : mblume 1482 Date: 2004/04/23 17:40:00 CDT
2138 :     Tag: blume-20040423-ieee-scan
2139 :     Description:
2140 :    
2141 :     Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
2142 :     in the exponent calculation we get INF or ZERO (depending on the mantissa
2143 :     and the sign of the exponent).
2144 :    
2145 :     ----------------------------------------------------------------------
2146 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2147 : mblume 1480 Date: 2004/04/23 10:40:00 CDT
2148 :     Tag: blume-20040423-ml-build
2149 :     Description:
2150 :    
2151 :     The ml-build script now terminates with a non-0 status when something
2152 :     goes wrong.
2153 :    
2154 :     ----------------------------------------------------------------------
2155 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2156 : mblume 1479 Date: 2004/04/22 16:35:00 CDT
2157 :     Tag: blume-20040422-Option
2158 :     Description:
2159 :    
2160 :     Made exception Option to be the same as exception Option.Option
2161 :     (as it should be).
2162 :    
2163 :     ----------------------------------------------------------------------
2164 : allenleung 1468 Name: Allen Leung (leunga (at) reservoir (dot) com)
2165 :     Date: 2004/03/19 14:40:00 EST
2166 :     Tag: leunga-20040319-cygwin-nlffi
2167 :     Description:
2168 :    
2169 :     Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
2170 :     of SML/NJ. The problem is that
2171 :    
2172 :     lib = dlopen(NULL, ...)
2173 :     f = dlsym(lib, "malloc");
2174 :    
2175 : blume 2155 does not work on Windows unless we explicitly export symbols
2176 :     such as 'malloc' during linking. We fixed this by explicitly
2177 : allenleung 1468 exporting the required symbols with the magic gcc incantation:
2178 :    
2179 :     -Wl,--export-all cygwin.def
2180 :    
2181 :     where cygwin.def is a file containing all the symbols that
2182 : blume 2155 we wish to export.
2183 : allenleung 1468
2184 : blume 2155 I suspect this is a Windows problem and we'll have to
2185 :     do the same (somehow with windows compilers) when
2186 : allenleung 1468 we build the native win32 version with the system
2187 : blume 2155 calls LoadLibrary/GetProcAddress.
2188 : allenleung 1468
2189 :     ----------------------------------------------------------------------
2190 : mblume 1384 Name: Matthias Blume (blume (at) tti - c (dot) org)
2191 : mblume 1465 Date: 2004/03/04 16:35:00 CST
2192 :     Tag: blume-20040304-intinf-fmt
2193 :     Description:
2194 :    
2195 :     Fixed problem with IntInf.fmt (sign would show up on the right instead
2196 :     of on the left for BIN, OCT, and HEX).
2197 :    
2198 :     ----------------------------------------------------------------------
2199 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2200 : mblume 1464 Date: 2004/03/04 11:25:00 CST
2201 :     Tag: blume-20040304-symlinks
2202 :     Description:
2203 :    
2204 :     Fixed problem with installer script (unix only) where bin/ml-yacc and
2205 :     friends pointed (via symlinks) to absolute locations instead of just
2206 :     .run-sml. This was reported by Vesa A Norrman.
2207 :    
2208 :     ----------------------------------------------------------------------
2209 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2210 : mblume 1455 Date: 2004/02/13 14:50:00 CST
2211 :     Tag: Release_110_45
2212 :     Description:
2213 :    
2214 :     New working version (110.45). New bootfiles.
2215 :    
2216 :     ----------------------------------------------------------------------
2217 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2218 : mblume 1448 Date: 2004/01/26 15:15:15 CST
2219 :     Tag: blume-20040126-toplevel
2220 :     Description:
2221 :    
2222 :     Improved handling of exceptions at the interactive toplevel.
2223 :    
2224 :     ----------------------------------------------------------------------
2225 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2226 :     Date: 2004/01/26 11:25:00
2227 : mblume 1446 Tag: blume-20040126-app
2228 :     Description:
2229 :    
2230 :     Type of top-level "app" corrected.
2231 : mblume 1447 Added code for setting vp_limitPtrMask to Win32-specific runtime.
2232 : mblume 1446
2233 :     ----------------------------------------------------------------------
2234 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2235 : mblume 1427 Date: 2003/11/18 17:10 CST
2236 :     Tag: blume-20031118-basis-fiddle
2237 :     Description:
2238 :    
2239 :     - changed Timer interface to what might become the spec
2240 :     - POSIX_FLAGS -> BIT_FLAGS according to spec
2241 :     - some other minor discrepancies wrt. spec eliminated
2242 :    
2243 :     ----------------------------------------------------------------------
2244 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2245 : mblume 1414 Date: 2003/11/06 12:00:00 CST
2246 :     Tag: Release_110_44
2247 :     Description:
2248 :    
2249 :     New working version (110.44). New bootfiles.
2250 :    
2251 :     ----------------------------------------------------------------------
2252 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2253 : mblume 1411 Date: 2003/11/04 11:50:00 CST
2254 :     Tag: blume-20031104-move-libraries
2255 :     Description:
2256 :    
2257 :     Eliminated the "dont_move_libraries" directive in config/targets.
2258 :     (The mechanism was broken and could not be fixed easily. Moreover,
2259 :     there does not seem to be any reason not to move all libraries into
2260 :     lib during installation. I originally implemented this directive as a
2261 :     backward-compatibility feature when I first introduced the new CM.
2262 :     Now that things have been stable for a long time and going back to the
2263 :     old CM is not an option, there is no reason to keep it around.)
2264 :    
2265 :     ----------------------------------------------------------------------
2266 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2267 : mblume 1409 Date: 2003/11/03 16:00:00 CST
2268 :     Tag: blume-20031103-installdir
2269 :     Description:
2270 :    
2271 :     Made installer honor INSTALLDIR variable again. (Thanks to Chris
2272 :     Richards for pointing out the problem and providing the solution.)
2273 :    
2274 :     ----------------------------------------------------------------------
2275 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2276 : mblume 1400 Date: 2003/10/01 17:05:00 CDT
2277 :     Tag: blume-20031001-lal-mlrisc
2278 :     Description:
2279 :    
2280 :     MLRISC bug fix from Lal.
2281 :    
2282 :     ----------------------------------------------------------------------
2283 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2284 : mblume 1399 Date: 2003/09/30 16:10:00 CDT
2285 :     Tag: blume-20030930-primio-bat
2286 : mblume 1398 Description:
2287 :    
2288 :     1. Added openVector, nullRd, and nullWr to PRIM_IO.
2289 :     2. Improved .bat files (for Win32 port) to make things work under Win95.
2290 :     (thanks to Aaron S. Hawley for this one)
2291 :    
2292 :     ----------------------------------------------------------------------
2293 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2294 : mblume 1395 Date: 2003/09/26 16:05:00 CDT
2295 :     Tag: blume-20030926-wrappriv
2296 :     Description:
2297 :    
2298 :     Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
2299 :    
2300 :     ----------------------------------------------------------------------
2301 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2302 : mblume 1394 Date: 2003/09/26 15:00:00 CDT
2303 :     Tag: blume-20030926-110_43_3
2304 :     Description:
2305 :    
2306 :     - additional cleanup
2307 :     - version number bump, NEW BOOTFILES
2308 :    
2309 :     ----------------------------------------------------------------------
2310 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2311 : mblume 1393 Date: 2003/09/26 12:00:00 CDT
2312 :     Tag: blume-20030926-ppautoload
2313 :     Description:
2314 :    
2315 :     I modified the read-eval-print loop so that the autoloader gets
2316 :     invoked whenever the prettyprinter tries to look up a symbol that
2317 :     is not currently defined in the toplevel environment but which
2318 :     appears in CM's autoload registry. As a result, we see far fewer of
2319 :     those ?.Foo.Bar.xxx names in the prettyprinter's output.
2320 :    
2321 :     In addition to this I tried to clean up some pieces of the Basis
2322 :     implementation (e.g., Socket, Word8Array) in order to prevent other
2323 :     instances of these ?.Foo.Bar.xxx names from being printed.
2324 :    
2325 :     The mechanism that picks names for types still needs some work, though.
2326 :     (Right now it seems that if there is a type A.t which is defined to
2327 : mblume 1394 be B.u, but B is unavailable at toplevel, then A.t gets printed as
2328 :     "?.B.u" although the perhaps more sensible solution would be to use
2329 : mblume 1393 "A.t" in this case. In other words, the prettyprinter should follow
2330 :     a chain of DEFtycs not farther than there are corresponding toplevel
2331 :     names in the current environment.)
2332 :    
2333 :     ----------------------------------------------------------------------
2334 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2335 : mblume 1391 Date: 2003/09/24 16:31:00 CDT
2336 :     Tag: blume-20030924-installer
2337 :     Description:
2338 :    
2339 :     Another installer tweak: All the ML code for the installer is now
2340 :     compiled during CMB.make and put into a little library called
2341 :     $smlnj/installer.cm. The installation then simply invokes
2342 :    
2343 :     sml -m $smlnj/installer.cm
2344 :    
2345 :     and everything happens automagically.
2346 :    
2347 :     Win32: ML code senses value of environment variable SMLNJ_HOME.
2348 :     Unix: ML code senses values of environment variables ROOT, CONFIGDIR,
2349 :     and BINDIR.
2350 :    
2351 :     The new scheme guarantees that the ML code responsible for the installation
2352 :     is in sync with the APIs of the main system. Also, the installer is
2353 :     somewhat faster because the installer script is precompiled.
2354 :    
2355 :     ----------------------------------------------------------------------
2356 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2357 : mblume 1390 Date: 2003/09/24 15:35:00 CDT
2358 :     Tag: blume-20030924-synsock
2359 :     Description:
2360 :    
2361 :     Added a signature SYNCHRONOUS_SOCKET to basis.cm. This is like SOCKET
2362 :     but excludes all non-blocking operations. Defined SOCKET (in Basis)
2363 :     and CML_SOCKET in terms of SYNCHRONOUS_SOCKET. Removed superfluous
2364 :     implementations of non-blocking operations from CML's Socket
2365 :     structure.
2366 :    
2367 :     ----------------------------------------------------------------------
2368 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2369 : mblume 1389 Date: 2003/09/24 15:10:05 CDT
2370 :     Tag: blume-20030924-sockets
2371 :     Description:
2372 :    
2373 :     1. Fixed SOCKET API and implementation to match Basis spec.
2374 :     This required changing the internal representation of sockets to one
2375 :     that remembers (for each socket file descriptor) whether it is currently
2376 :     blocking or non-blocking. This state is maintained lazily (i.e., a system
2377 :     call is made only if the state actually needs to change).
2378 :    
2379 :     2. OS-specific details of sockets were moved into separate files, thus
2380 :     making it possible to unify the bulk of the socket implementations
2381 :     between Unix and Win32.
2382 :    
2383 :     3. CML's socket API changed accordingly.
2384 :     (Note that we need to remove non-blocking functions from this API
2385 :     since they are redundant in the case of CML!)
2386 :    
2387 :     4. CML's socket implementation now makes use of non-blocking functions
2388 :     provided by Basis, thus removing all OS-dependent code from this part
2389 :     of CML.
2390 :    
2391 :     5. Changed Real64.precision from 52 to 53. Minor cleanup in Real64 code.
2392 :    
2393 :     ----------------------------------------------------------------------
2394 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2395 : mblume 1387 Date: 2003/09/22 12:10:00 CDT
2396 : mblume 1388 Tag: blume-20030922-110_43_2
2397 : mblume 1387 Description:
2398 :    
2399 :     Made a new interim version and bootfiles for developer's bootstrapping
2400 :     convenience.
2401 :    
2402 :     110.43.2 -- NEW BOOTFILES
2403 :    
2404 :     ----------------------------------------------------------------------
2405 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2406 : mblume 1385 Date: 2003/09/19 15:55:00 CDT
2407 :     Tag: blume-20030919-cmdir
2408 :     Description:
2409 :    
2410 :     1. new-install.sh -> install.sh
2411 :     2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
2412 :     3. tweaked installer so that another name instead of .cm can be chosen
2413 :     at install time (by setting the CM_DIR_ARC environment variable
2414 :     during installation); once installation is complete, the name is
2415 :     fixed
2416 :    
2417 :     ----------------------------------------------------------------------
2418 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2419 : mblume 1384 Date: 2003/09/18 16:00:00 CDT
2420 :     Tag: blume-20030918-110_43_1
2421 :     Description:
2422 :    
2423 :     Made a new interim version and bootfiles for developer's bootstrapping
2424 :     convenience.
2425 :    
2426 :     110.43.1 -- NEW BOOTFILES
2427 :    
2428 :     ----------------------------------------------------------------------
2429 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2430 : mblume 1383 Date: 2003/09/18 15:20:00 CDT
2431 :     Tag: blume-20030918-misc
2432 :     Description:
2433 :    
2434 :     1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
2435 :     this seems to be controversial at the moment) and used those in
2436 :     Posix.ProcEnv.times.
2437 :    
2438 :     2. Added Time.{from,to}Nanoseconds to Time.
2439 :    
2440 :     3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
2441 :     For example, fromLargeInt never needs to look at more than 3 "big
2442 :     digits" to get its 53 bits of precision.
2443 :    
2444 :     ----------------------------------------------------------------------
2445 : mblume 1384 Name: Matthias Blume (blume (at) tti - c (dot) org)
2446 : mblume 1381 Date: 2003/09/17 16:30:00 CDT
2447 :     Tag: blume-20030917-real32-slices
2448 :     Description:
2449 :    
2450 :     Added an entry to the primitive environment
2451 :     (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
2452 :     added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
2453 :    
2454 :     Removed some of the "magic" constants in real64.sml and replaced them
2455 :     with code that generates these values from their corresponding
2456 :     integer counterparts.
2457 :    
2458 :     Made all(?) the slice-related changes to the Basis and made everything
2459 :     compile again...
2460 :    
2461 :     ----------------------------------------------------------------------
2462 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2463 : mblume 1375 Date: 2003/09/15 17:45:00 CDT
2464 :     Tag: blume-20030915-rbase
2465 :     Description:
2466 :    
2467 :     Fixed bug in Real.fromLargeInt.
2468 :    
2469 :     ----------------------------------------------------------------------
2470 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2471 : mblume 1369 Date: 2003/09/13 18:11:00 CDT
2472 :     Tag: blume-20030913-libinstall
2473 :     Description:
2474 :    
2475 :     Minor bugfix in config/libinstall (set anchor with path to
2476 :     standalone tool after installing it, otherwise libraries that
2477 :     need ml-lex or ml-yacc won't compile the first time the installer
2478 :     runs).
2479 :    
2480 :     ----------------------------------------------------------------------
2481 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2482 : mblume 1368 Date: 2003/09/12 11:45:00 CDT
2483 :     Tag: blume-20030912-various
2484 :     Description:
2485 :    
2486 :     - fixed bug in Real.toLargeInt
2487 :     - fixed bug in Posix.ProcEnv.times
2488 :     - changed inputLine functions to return an option
2489 :     - minor installer improvements / bugfixes
2490 :     - changed default @SMLalloc parameter for x86/celeron to 64k
2491 :    
2492 :     ----------------------------------------------------------------------
2493 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2494 : mblume 1356 Date: 2003/09/09 22:00:00 CDT
2495 :     Tag: Release_110_43
2496 :     Description:
2497 :    
2498 :     New working release 110.43. New bootfiles.
2499 :    
2500 :     ----------------------------------------------------------------------
2501 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2502 : mblume 1354 Date: 2003/09/09 19:20:00 CDT
2503 :     Tag: blume-20030909-installer
2504 :     Description:
2505 :    
2506 :     Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
2507 :     Modified config/install.bat to take advantage of it. Also modified
2508 :     config/install.sh (and called it config/new-install.sh) to take advantage
2509 :     of it on Unix systems. (The SML code is (supposed to be) platform-
2510 :     independent.)
2511 :    
2512 :     The installer can now install everything under Win32
2513 :     as well as under *nix as long as it compiles.
2514 :    
2515 :     Other changes:
2516 :    
2517 :     - made CML compile again under Win32
2518 :     - made eXene compile under Win32 (by providing a fake structure UnixSock
2519 :     and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
2520 :     - fixed a bug in nowhere: it assumed that type OS.Process.status is the
2521 :     same as type int; under Win32 it isn't
2522 :     - fixed some slice-related problems in the win32-specific parts of CML
2523 :     - added a functor argument "sameVol" to os-path-fn.sml in the Basis
2524 :     (under Win32, the volume name is case-insensitive, and the
2525 :     OS.Path code compares volume names for equality)
2526 :    
2527 :     ----------------------------------------------------------------------
2528 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2529 : mblume 1353 Date: 2003/09/08 11:55:00 CDT
2530 :     Tag: blume-20030908-fullpath
2531 :     Description:
2532 :    
2533 :     Made Win32 version of OS.FileSys.fullPath return current directory
2534 :     when given an empty string. This is what the spec says, and incidentally,
2535 :     CM depends on it. (CM otherwise goes into an infinite loop in certain
2536 :     cases when presented with the name of a non-existing .cm file.)
2537 :    
2538 :     ----------------------------------------------------------------------
2539 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2540 : mblume 1350 Date: 2003/09/04 16:30:00 CDT
2541 : mblume 1352 Tag: blume-20030905-slices-etc
2542 : mblume 1350 Description:
2543 :    
2544 :     1. Changed interface to vectors and arrays in Basis to match
2545 :     (draft) Basis spec.
2546 :     2. Added signatures and implementations of slices according to
2547 :     Basis spec.
2548 :     3. Edited source code throughout the system to make it compile again
2549 :     under 1. and 2. (In some cases code had to be added to have it
2550 :     match the new signatures.)
2551 :     4. MLRISC should be backward-compatible: the copies of the originals
2552 :     of files that needed to change under 3. were retained, the .cm files
2553 :     check the compiler version number and use old versions when
2554 :     appropriate.
2555 :     5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
2556 :     dirstream -> string option (in accordance with Basis spec).
2557 :     6. When generating code that counts lines, ml-lex used function
2558 :     CharVector.foldli, taking advantage of its old interface.
2559 :     This has been replaced with the corresponding code from
2560 :     CharVectorSlice. (html-lex must be re-lexed!)
2561 :     7. BitArray in smlnj-lib/Util has been extended/modified to match the
2562 :     new MONO_ARRAY signature. (Do we need BitArraySlice?)
2563 :     8. Removed temporary additions (fromInternal, toInternal) from the
2564 :     (now obsolete) IntInf in smlnj-lib/Util.
2565 :     9. Cleaned up structure Byte.
2566 :     10. Added localOffset, scan, and fromString to Date (according to spec).
2567 :     Cleaned/corrected implementation of Date.
2568 :     (Still need to check for correctness; implement better canonicalizeDate.)
2569 :     11. Added "scan" to signature IEEE_REAL.
2570 :     12. Some improvements to IntInf [in particular: efficiency-hack for
2571 :     mod and rem when second operand is 2 (for parity checks).]
2572 :     13. Changed representation of type Time.time, using a single IntInf.int
2573 :     value counting microseconds. This considerably simplified the
2574 :     implementation of structure Time. We now support negative time
2575 :     values; scan and fromString handle signs.
2576 :     14. Functor PrimIO now takes two additional arguments (VectorSlice and
2577 :     ArraySlice).
2578 :    
2579 :     ----------------------------------------------------------------------
2580 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2581 : mblume 1347 Date: 2003/08/28 17:00:00 CDT
2582 :     Tag: blume-20030828-intinf
2583 :     Description:
2584 :    
2585 :     This is a major update which comes with a version number bump
2586 :     (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
2587 :     and an implementation of IntInf in the Basis.
2588 :    
2589 :     There are a fairly large number of related changes and updates throughout
2590 :     the system:
2591 :    
2592 :     Basis:
2593 :     - Implemented IntInf.
2594 :     - Made LargeInt a projection of IntInf (by filtering through INTEGER).
2595 : mblume 1348 - Added some missing Real64 operations, most notably Real.toLargeInt.
2596 : mblume 1347 - Added FixedInt as a synonym for Int32.
2597 :    
2598 :     compiler:
2599 :     * Added support for a built-in intinf type.
2600 :     - literals
2601 :     - pattern matching
2602 :     - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
2603 :     - overloading on literals and operations
2604 :    
2605 :     This required adding a primitive type intinf, some additional
2606 :     primops, and implementations for several non-trivial intinf
2607 :     operations in Core. (The intinf type is completely abstract
2608 :     to the compiler; all operations get delegated back to the Core.)
2609 :    
2610 :     * Intinf equality is handled by polyequal. However, the compiler
2611 :     does not print its usual warning in this case (since polyequal
2612 :     is the right thing to do there).
2613 :    
2614 :     * Improved the organization of structure InlineT.
2615 :    
2616 :     * A word about conversion primops:
2617 :     If conversions involving intinf do not cancel out during
2618 : mblume 1348 CPS contract, then the compiler must insert calls to Core functions.
2619 : mblume 1347 Since all core access must be resolved already during the FLINT
2620 :     translate phase, it would be too late a the time of CPS contract
2621 : mblume 1348 to add new Core calls. For this reason, conversion primops
2622 : mblume 1347 for intinf carry two arguments: 1. the numeric argument that
2623 :     they are supposed to convert, and 2. the Core function that
2624 :     can help with this conversion if necessary. If CPS contract
2625 : mblume 1348 eliminates a primop, then the associated Core function becomes
2626 :     dead and goes away. Intinf conversion primops that do not get
2627 :     eliminated by CPS contract get rewritten into calls of their
2628 :     core functions by a separate, new phase.
2629 : mblume 1347
2630 :     interactive system:
2631 :     - Control.Print.intinfDepth controls max length of intinf constants
2632 :     being printed. (Analogous to Control.Print.stringDepth.)
2633 :     - Cleanup in printutil and pputil: got rid of unused stuff and
2634 :     duplicates; replaced some of the code with code that makes better
2635 :     use of library functionality.
2636 :    
2637 :     CM:
2638 :     Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
2639 :     are no longer silent.
2640 :    
2641 :     CKIT:
2642 :     Fixed mismatched uses of Int32 and LargeInt. I always decided
2643 :     in favor of LargeInt -- which is now the same as IntInf.
2644 :     CKIT-knowledgable people should check whether this is what's
2645 :     intended and otherwise change things back to using Int32 or
2646 :     FixedInt.
2647 :    
2648 :     Throughout the code:
2649 : mblume 1348 Started using IntInf.int literals and built-in operations
2650 : mblume 1347 (e.g., comparison with 0) where this seems appropriate.
2651 :    
2652 :    
2653 :     ----------------------------------------------------------------------
2654 : macqueen 1343 Name: Dave MacQueen (dbm@cs.uchicago.edu)
2655 :     Date: 2003/08/13 11:36:00 CDT
2656 :     Tag: dbm-20030813-mcz-merge1
2657 :     Description:
2658 :    
2659 :     Merging changes from the mcz-branch development branch into trunk.
2660 :     These changes involve replacement of the emulated old prettyprinter
2661 :     interface with direct use of the SML/NJ Lib PP library, and fixing
2662 :     of a couple of bugs (895, 1186) relating to error messages. A new
2663 :     prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
2664 :     has been added.
2665 :    
2666 :     ----------------------------------------------------------------------
2667 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2668 : mblume 1342 Date: 2003/08/11 15:45:00 CDT
2669 :     Tag: blume-20030811-windows
2670 :     Description:
2671 :    
2672 :     Version number bumped to 110.42.9. NEW BOOTFILES!!!
2673 :    
2674 :     http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
2675 :    
2676 :     This patch restores SML/NJ's ability to run under win32. There are a
2677 :     number of changes, including fixes for several bugs that had gone
2678 :     unnoticed until now:
2679 :    
2680 :     - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
2681 :     trying to run the win32 version from within cygwin.)
2682 :    
2683 :     - There are a number of simple .bat scripts that substitute for their
2684 :     corresponding Unix shell-scripts. (See below.)
2685 :    
2686 :     - The internals of ml-build have been modified slightly. The main
2687 :     difference is that instead of calling ".link-sml" (or link-sml.bat)
2688 :     using OS.Process.system, the ML process delegates this task back
2689 :     to the script. Otherwise problems arise in mixed environments such
2690 :     as Cygwin where scripts look and work like Unix scripts, but
2691 :     where OS.Process.system cannot run them.
2692 :    
2693 :     - In CM, the srcpath pickler used native pathname syntax -- which
2694 :     is incorrect in the case of cross-compilation. The new pickle format
2695 :     is independent of platform-specific naming conventions.
2696 :    
2697 :     - Path configuration files (such as lib/pathconfig) can now choose
2698 :     between native and standard syntax. Placing a line of the form
2699 :    
2700 :     standard!
2701 :    
2702 :     into the file causes all subsequent paths to be interpreted using
2703 :     CM standard pathname syntax (= Unix conventions); a line
2704 :    
2705 :     native!
2706 :    
2707 :     switches back to native style. This was needed so that
2708 :     path config files can be written portably, see src/system/pathconfig.
2709 :    
2710 :     - Runtime system:
2711 :    
2712 :     - win32-filesys.c: get_file_time and set_file_time now
2713 :     access modification time, not creation time.
2714 :    
2715 :     - I/O code made aware of new array representation.
2716 :    
2717 :     - Bug fixes in X86.prim.masm.
2718 :    
2719 :     - src/system/makeml made aware of win32. (For use under cygwin
2720 :     and other Unix-environments for windows.)
2721 :    
2722 :     - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
2723 :     which caused BinIO.inputAll to fail consistently.
2724 :    
2725 :     .bat scripts:
2726 :    
2727 :     Windows .bat scripts assume that SMLNJ_HOME is defined.
2728 :    
2729 :     - sml.bat, ml-yacc.bat, ml-lex.bat: Driver scripts for standalone
2730 :     applications (sml, ml-yacc, ml-lex).
2731 :     - ml-build.bat: analogous to ml-build.
2732 :     - config\install.bat: Analogous to config/install.sh. This requires
2733 :     that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2734 :     (nmake etc. must be on the path, and vcvars32 must have been run.)
2735 :     Moreover, sources for ml-lex and ml-yacc need to exist under src,
2736 :     and the bootfile hierarchy must have been unpacked under
2737 :     sml.boot.x86-win32.
2738 :     The script is very primitive and does a poor job at error checking.
2739 :     It only installs the base system, ml-lex, and ml-yacc. No other
2740 :     libraries are being installed (i.e., you get only those that
2741 :     are part of the compiler.)
2742 :     - link-sml.bat: analogous to .link-sml, but not currently used
2743 :    
2744 :     Unrelated bug fixes:
2745 :    
2746 :     - ml-nlffigen now exports structures ST_* corresponding to incomplete
2747 :     types.
2748 :     - Added getDevice to PP/src/pp-debug-fn.sml. (Would not compile
2749 :     otherwise.)
2750 :    
2751 :     ----------------------------------------------------------------------
2752 : macqueen 1337 Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2753 :     Date: 2003/06/17
2754 :     Tag: macqueen-20030617-bug895
2755 :     Description:
2756 :    
2757 :     Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2758 :     Tag will be used for new development branch (mcz-branch) for
2759 :     use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2760 :    
2761 :     ----------------------------------------------------------------------
2762 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2763 : mblume 1335 Date: 2003/05/27 16:55:00 CDT
2764 :     Tag: blume-20030527-polyeq
2765 :     Description:
2766 :    
2767 :     Tried to eliminated most cases of polymorphic equality.
2768 :    
2769 :     ----------------------------------------------------------------------
2770 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2771 : mblume 1334 Date: 2003/05/21 17:45:00 CDT
2772 :     Tag: blume-20030517-complete
2773 :     Description:
2774 :    
2775 :     Two changes:
2776 :    
2777 :     1. Added a flag for controlling whether non-exhaustive bindings will
2778 :     be treated as errors (default is false).
2779 :     2. Cleaned up the *entire* source tree so that CMB.make goes through
2780 :     without a single non-exhaustive match- or bind warning.
2781 :    
2782 :     ----------------------------------------------------------------------
2783 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2784 : mblume 1332 Date: 2003/05/17 10:20:00 CDT
2785 :     Tag: blume-20030517-absyn
2786 :     Description:
2787 :    
2788 :     1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2789 :    
2790 :     This mainly affects the quality of error messages. However, some
2791 :     of the code is now more straightforward than before. (Treatment of
2792 :     the above four constructs in translate.sml is much simpler than
2793 :     the "macro-expansion" that was going on before. Plus, the mach-
2794 :     compiler no longer gets invoked just to be able to compile an
2795 :     if-expression.)
2796 :    
2797 :     2. The ErrorMsg.Error exception is now caught and absorbed by the
2798 :     interactive loop.
2799 :    
2800 :     ----------------------------------------------------------------------
2801 : allenleung 1327 Name: Allen Leung
2802 :     Date: 2003/05/16 13:05:00 CDT
2803 :     Tag: leunga-20030516-cygwin-runtime
2804 :     Description:
2805 :    
2806 : blume 2155 Ported the runtime system to cygwin, which uses the unix
2807 : allenleung 1327 x86-unix bin files. Missing/buggy features:
2808 :    
2809 : blume 2155 o getnetbyname, getnetbyaddr: these functions seem to be missing in
2810 : allenleung 1327 the Cygwin library.
2811 :     o Ctrl-C handling may be flaky.
2812 :     o Windows system calls and Windows I/O are not supported.
2813 :    
2814 :     A new set of binfiles is located at:
2815 :    
2816 :     http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2817 :    
2818 :     This is only needed for bootstrapping the cygwin version of smlnj.
2819 :     Other x86 versions can use the existing binfiles.
2820 :    
2821 :     ----------------------------------------------------------------------
2822 : mblume 1317 Name: Matthias Blume
2823 :     Date: 2003/04/08 15:42:00 CDT
2824 :     Tag: blume-20030408-listpair
2825 :     Description:
2826 :    
2827 :     1. Added a target 'mlrisc' to installer.
2828 :    
2829 :     2. Added missing elements to structure ListPair.
2830 :    
2831 :     ----------------------------------------------------------------------
2832 : leunga 1296 Name: Allen Leung
2833 :     Date: 2003/01/07 10:40:00 EST
2834 :     Tag: leunga-20030107-int-rem
2835 :     Description:
2836 :    
2837 :     Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2838 :     The arguments to the SUBL instruction were swapped.
2839 :    
2840 :     ----------------------------------------------------------------------
2841 : blume 1229 Name: Matthias Blume
2842 : blume 1294 Date: 2002/12/12 16:25:00 EST
2843 :     Tag: blume-20021212-risc-ra
2844 :     Description:
2845 :    
2846 :     Fixed a serious bug in the rewrite code for FP spilling/reloading that
2847 :     sent the RA into an infinite loop when floating point registers get
2848 :     spilled. (Because of this bug, e.g., nucleic stopped compiling between
2849 :     110.37 and 110.38.)
2850 :     There was another set of potential problems related to the handling of
2851 :     MLRISC annotations (but those did not yet cause real problems, apparently).
2852 :    
2853 :     ----------------------------------------------------------------------
2854 :     Name: Matthias Blume
2855 : blume 1291 Date: 2002/12/06 22:40:00 EST
2856 :     Tag: blume-20021206-cm-fileid
2857 :     Description:
2858 :    
2859 :     Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2860 :     This fixes the problem of CM getting confused by files that suddenly
2861 :     change their identity (e.g., by getting unlinked and recreated by some
2862 :     text editor such as vi). There might be a better/cheaper/cleaner way
2863 :     of doing this, but for now this will have to do.
2864 :    
2865 :     ----------------------------------------------------------------------
2866 :     Name: Matthias Blume
2867 : blume 1288 Date: 2002/10/28 09:50:00 EST
2868 : blume 2155 Tag: blume-20021028-typecheck
2869 : blume 1288 Description:
2870 :    
2871 :     Exported structure Typecheck from $smlnj/viscomp/core.cm.
2872 :    
2873 :     ----------------------------------------------------------------------
2874 :     Name: Matthias Blume
2875 : blume 1282 Date: 2002/10/17 09:10:00 EDT
2876 : blume 1279 Tag: Release_110_42
2877 :     Description:
2878 :    
2879 : blume 1282 In good old tradition, there has been a slight hiccup so that we have
2880 :     to patch 110.42 after the fact. The old release tag has been replaced
2881 :     (see below).
2882 :    
2883 :     The change solves a problem with two competing approaches the
2884 :     configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2885 :     each other's way.
2886 :    
2887 :     This change only affects the runtime system code and the installer script.
2888 :     (No new bootfiles.)
2889 :    
2890 :     ----------------------------------------------------------------------
2891 :     Name: Matthias Blume
2892 :     Date: 2002/10/16 12:00:00 EDT
2893 :     Tag: Release_110_42_removed
2894 :     Description:
2895 :    
2896 : blume 1279 New working release. New bootfiles.
2897 :    
2898 :     ----------------------------------------------------------------------
2899 :     Name: Matthias Blume
2900 : blume 1273 Date: 2002/10/10 13:10:00 EDT
2901 :     Tag: blume-20021010-ppc-divs
2902 :     Description:
2903 :    
2904 :     The mltree operator DIVS must be implemented with an overflow check on
2905 :     the PPC because the hardware indicates divide-by-zero using "overflow" as
2906 :     well.
2907 :    
2908 :     ----------------------------------------------------------------------
2909 :     Name: Matthias Blume
2910 : blume 1261 Date: 2002/07/23 11:45:00
2911 :     Tag: blume-20020723-smlnj-home
2912 :     Description:
2913 :    
2914 :     Sml now senses the SMLNJ_HOME environment variable. If this is set,
2915 :     then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2916 :     CM_PATHCONFIG is also set), the path configuration file is assumed
2917 :     to be in $SMLNJ_HOME/lib/pathconfig. This way one can easily move
2918 :     the entire tree to some other place and everything will "just work".
2919 :    
2920 :     (Companion commands such as ml-build and ml-makedepend also sense this
2921 :     variable.)
2922 :    
2923 :     ----------------------------------------------------------------------
2924 :     Name: Matthias Blume
2925 : blume 1259 Date: 2002/07/12 21:19:00 EDT
2926 :     Tag: blume-20020712-liveness
2927 :     Description:
2928 :    
2929 :     Exported two useful "step" functions from liveness module (MLRISC).
2930 :    
2931 :     ----------------------------------------------------------------------
2932 :     Name: Matthias Blume
2933 : blume 1253 Date: 2002/07/05 16:00 EDT
2934 :     Tag: Release_110_41
2935 :     Description:
2936 :    
2937 :     New working release. New bootfiles.
2938 :    
2939 :     ----------------------------------------------------------------------
2940 :     Name: Matthias Blume
2941 : blume 1250 Date: 2002/07/05 10:25:00 EDT
2942 :     Tag: blume-20020705-btimp
2943 :     Description:
2944 :    
2945 :     Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2946 :     other clients can set up backtracing support.
2947 :    
2948 :     ----------------------------------------------------------------------
2949 :     Name: Matthias Blume
2950 : blume 1249 Date: 2002/06/25 17:23:00 EDT
2951 :     Tag: blume-20020625-fpmax
2952 :     Description:
2953 :    
2954 :     Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2955 :     case. (The sense of the isNaN test was reversed -- which made min and
2956 :     max always return their first argument.)
2957 :    
2958 :     ----------------------------------------------------------------------
2959 :     Name: Matthias Blume
2960 : blume 1240 Date: 2002/06/11
2961 :     Tag: blume-20020611-unixpath
2962 :     Description:
2963 :    
2964 :     Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2965 :    
2966 :     ----------------------------------------------------------------------
2967 :     Name: Matthias Blume
2968 : blume 1238 Date: 2002/06/10 16:35:00 EDT
2969 :     Tag: blume-20020610-ieeereal
2970 :     Description:
2971 :    
2972 :     I back-ported my implementation of IEEEReal.fromString from the
2973 :     idlbasis-devel branch so that we can test it.
2974 :    
2975 :     Another small change is that ppDec tries to give more information
2976 :     than just "<sig>" in the case of functors. However, this code is
2977 :     broken in some mysterious way if the functor's body's signature
2978 :     has not been declared by ascription but gets inferred from the
2979 :     implementation. This needs fixing...
2980 :    
2981 :     ----------------------------------------------------------------------
2982 :     Name: Matthias Blume
2983 : blume 1229 Date: 2002/05/31
2984 :     Tag: blume-20020531-btrace-mode
2985 :     Description:
2986 :    
2987 :     Resurrected SMLofNJ.Internals.BTrace.mode. (It accidentally fell by
2988 :     the wayside when I switched over to using Controls everywhere.)
2989 :    
2990 :     ----------------------------------------------------------------------
2991 : george 1222 Name: Lal George
2992 :     Date: 2002/05/23 12:21:40 EDT
2993 :     Tag: george-20020523-visual-labels
2994 :     Description:
2995 :    
2996 :     Labels are now displayed in the graphical output to make
2997 :     the fall-through and target blocks obvious.
2998 :    
2999 :     ----------------------------------------------------------------------
3000 : blume 1204 Name: Matthias Blume
3001 : blume 1220 Date: 2002/05/22 11:03:00 EDT
3002 :     Tag: blume-20020522-shrink
3003 :     Description:
3004 :    
3005 :     John tweaked yesterday's fix for 1131 to handle an out-of-memory
3006 :     situation that comes up when allocating huge arrays.
3007 :    
3008 :     ----------------------------------------------------------------------
3009 :     Name: Matthias Blume
3010 : blume 1215 Date: 2002/05/21 16:00:00 EDT
3011 :     Tag: Release_110_40
3012 :     Description:
3013 :    
3014 :     New working release (110.40). New bootfiles.
3015 :    
3016 :     [Also: John Reppy fixed GC bug 1131.]
3017 :    
3018 :     ----------------------------------------------------------------------
3019 :     Name: Matthias Blume
3020 : blume 1212 Date: 2002/05/21 12:35:00 EDT
3021 :     Tag: blume-20020521-cmdoc
3022 :     Description:
3023 :    
3024 :     CM documentation update.
3025 :    
3026 :     ----------------------------------------------------------------------
3027 :     Name: Matthias Blume
3028 : blume 1211 Date: 2002/05/21 10:55:00 EDT
3029 :     Tag: blume-20020521-misc
3030 :     Description:
3031 :    
3032 :     - John tweaked runtime to be silent on heap export (except when
3033 :     GC messages are on).
3034 :     - I added a few more things (cross-compiling versions of CMB) to
3035 :     config/preloads (as suggestions).
3036 :    
3037 :     ----------------------------------------------------------------------
3038 :     Name: Matthias Blume
3039 : blume 1208 Date: 2002/05/20 22:25:00 EDT
3040 :     Tag: blume-20020520-controls
3041 :     Description:
3042 :    
3043 :     - Added ControlUtil structure to control-lib.cm.
3044 :     - Use it throughout.
3045 :     - Used Controls facility to define MLRISC controls (as opposed to
3046 :     registering MLRISC control ref cells with Controls after the fact)
3047 :     - Fixed messed-up controls priorities.
3048 :    
3049 :     * Removed again all the stuff from config/preloads that one wouldn't
3050 :     be able to preload at the time the initial heap image is built.
3051 :     (Many libraries, e.g., CML, do not exist yet at this time. The
3052 :     only libraries that can be preloaded via config/preloads are those
3053 :     that come bundled with the bootfiles.)
3054 :    
3055 :     ----------------------------------------------------------------------
3056 :     Name: Matthias Blume
3057 : blume 1204 Date: 2002/05/20 10:59:00 EDT
3058 :     Tag: blume-20020520-preloads
3059 :     Description:
3060 :    
3061 :     Added a lot of commented-out suggestions for things to be included
3062 :     in config/preloads.
3063 :    
3064 :     ----------------------------------------------------------------------
3065 : leunga 1203 Name: Allen Leung
3066 :     Date: 2002/05/18 14:20:00 EDT
3067 :     Tag: leunga-20020518-mdl
3068 :     Description:
3069 :    
3070 :     o Made the mdl tool stuff compile and run again.
3071 :     o I've disabled all the stuff that depends on RTL specifications; they
3072 :     are all badly broken anyway.
3073 :    
3074 :     ----------------------------------------------------------------------
3075 : blume 1201 Name: Matthias Blume
3076 :     Date: 2002/05/17 16:49:00 EDT
3077 :     Tag: blume-20020517-controls
3078 :     Description:
3079 :    
3080 :     0. John Reppy made several modifications to the SML/NJ library.
3081 :     In particular, there is a shiny new controls-lib.cm.
3082 :    
3083 :     1. Pushed new controls interface through compiler so that everything
3084 :     compiles again.
3085 :    
3086 :     2. Added FormatComb and FORMAT_COMB to the CML version of the
3087 :     SML/NJ library (so that CML compiles again).
3088 :    
3089 :     3. Modified init scripts because XXX_DEFAULT environment variables
3090 :     are no longer with us. (Boot-time initialization is now done
3091 :     using the same environment variables that are also used for
3092 :     startup-time initialization of controls.)
3093 :    
3094 :     ----------------------------------------------------------------------
3095 : george 1192 Name: Lal George
3096 :     Date: 2002/05/15 09:20:10 EDT
3097 :     Tag: george-20020515-pseudo-op-decls
3098 :     Description:
3099 :    
3100 :     All pseudo-ops emitted before the first segment declaration
3101 :     such as TEXT, DATA, and BSS directives are assumed to be global
3102 :     declarations and are emitted first in the assembly file. This is
3103 :     useful in a number of situations where one has pseudo-ops that are not
3104 :     specific to any segment, and also works around the constraint that one
3105 :     cannot have client pseudo-ops in the TEXT segment.
3106 :    
3107 :     Because no segment is associated with these declarations it is
3108 :     an error to allocate any space or objects before the first segment
3109 :     directive and an exception will be raised. However, we cannot make
3110 :     this check for client pseudo-ops.
3111 :    
3112 :     These top level declarations are a field in the CFG graph_info.
3113 :     In theory you can continue to add to this field after the CFG has been
3114 :     built -- provided you know what you are doing;-)
3115 :    
3116 :     ----------------------------------------------------------------------
3117 : blume 1175 Name: Matthias Blume
3118 : blume 1190 Date: 2002/05/13 16:40:00 EDT
3119 :     Tag: blume-20020513-pp-etc
3120 :     Description:
3121 :    
3122 :     A few minor bugfixes:
3123 :    
3124 :     - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
3125 :     (Bogus printouts for redefined bindings still occur. Compiler
3126 :     bug should no longer occur now. We need to redo the prettyprinter
3127 :     from scratch.)
3128 :    
3129 :     - CM pathname printer now also adds escape sequences for ( and )
3130 :    
3131 :     - commend and docu fixes for ml-nlffi
3132 :    
3133 :     ----------------------------------------------------------------------
3134 :     Name: Matthias Blume
3135 : blume 1189 Date: 2002/05/10 16:40:00 EDT
3136 :     Tag: blume-20020510-erg-textio
3137 :     Description:
3138 :    
3139 :     Applied the following bugfix provided by Emden Gansner:
3140 :    
3141 :     Output is corrupted when outputSubstr is used rather than output.
3142 :    
3143 : blume 2155 The problem occurs when a substring
3144 : blume 1189
3145 :     ss = (s, dataStart, dataLen)
3146 :    
3147 : blume 2155 where dataStart > 0, fills a stream buffer with avail bytes left.
3148 :     avail bytes of s, starting at index dataStart, are copied into the
3149 : blume 1189 buffer, the buffer is flushed, and then the remaining dataLen-avail
3150 : blume 2155 bytes of ss are copied into the beginning of the buffer. Instead of
3151 :     starting this copy at index dataStart+avail in s, the current code
3152 : blume 1189 starts the copy at index avail.
3153 :    
3154 :     Fix:
3155 :     In text-io-fn.sml, change line 695 from
3156 :     val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
3157 : blume 2155 to
3158 : blume 1189 val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
3159 :    
3160 :     ----------------------------------------------------------------------
3161 :     Name: Matthias Blume
3162 : blume 1186 Date: 2002/04/12 13:55:00 EDT
3163 :     Tag: blume-20020412-assyntax
3164 :     Description:
3165 :    
3166 :     1. Grabbed newer assyntax.h from the XFree86 project.
3167 :     2. Fiddled with how to compile X86.prim.asm without warnings.
3168 :     3. (Very) Minor cleanup in CM.
3169 :    
3170 :     ----------------------------------------------------------------------
3171 :     Name: Matthias Blume
3172 : blume 1185 Date: 2002/04/01 (no joke!) 17:07:00 EST
3173 :     Tag: blume-20020401-x86div
3174 :     Description:
3175 :    
3176 :     Added full support for div/mod/rem/quot on the x86, using the machine
3177 :     instruction's two results (without clumsily recomputing the remainder)
3178 :     directly where appropriate.
3179 :    
3180 :     Some more extensive power-of-two support was added to the x86 instruction
3181 :     selector (avoiding expensive divs, mods, and muls where they can be
3182 :     replaced with cheaper shifts and masks). However, this sort of thing
3183 :     ought to be done earlier, e.g., within the CPS optimizer so that
3184 :     all architectures benefit from it.
3185 :    
3186 :     The compiler compiles to a fixed point, but changes might be somewhat
3187 :     fragile nevertheless. Please, report any strange things that you might
3188 :     see wrt. div/mod/quot/rem...
3189 :    
3190 :     ----------------------------------------------------------------------
3191 :     Name: Matthias Blume
3192 : blume 1184 Date: 2002/03/29 17:22:00
3193 :     Tag: blume-20020329-div
3194 :     Description:
3195 :    
3196 :     Fixed my broken div/mod logic. Unfortunately, this means that the
3197 :     inline code for div/mod now has one more comparison than before.
3198 :     Fast paths (quotient > 0 or remainder = 0) are not affected, though.
3199 :     The problem was with quotient = 0, because that alone does not tell
3200 :     us which way the rounding went. One then has to look at whether
3201 :     remainder and divisor have the same sign... :(
3202 :    
3203 :     Anyway, I replaced the bootfiles with fresh ones...
3204 :    
3205 :     ----------------------------------------------------------------------
3206 :     Name: Matthias Blume
3207 : blume 1183 Date: 2002/03/29 14:10:00 EST
3208 :     Tag: blume-20020329-inlprims
3209 :     Description:
3210 :    
3211 :     NEW BOOTFILES!!! Version number bumped to 110.39.3.
3212 :    
3213 :     Primops have changed. This means that the bin/boot-file formats have
3214 :     changed as well.
3215 :    
3216 :     To make sure that there is no confusion, I made a new version.
3217 :    
3218 :    
3219 :     CHANGES:
3220 :    
3221 :     * removed REMT from mltree (remainder should never overflow).
3222 :    
3223 :     * added primops to deal with divisions of all flavors to the frontend
3224 :    
3225 :     * handled these primops all the way through so they map to their respective
3226 :     MLRISC support
3227 :    
3228 :     * used these primops in the implementation of Int, Int32, Word, Word32
3229 :    
3230 :     * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
3231 :    
3232 :     * parameterized INLMIN, INLMAX, and INLABS by a numkind
3233 :    
3234 :     * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
3235 :     floating point
3236 :    
3237 :     * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
3238 :     and Real (but Real.abs maps to a separate floating-point-only primop)
3239 :    
3240 :    
3241 :     TODO items:
3242 :    
3243 :     * Hacked Alpha32 instruction selection, disabling the selection of REMx
3244 :     instructions because the machine instruction encoder cannot handle
3245 :     them. (Hppa, PPC, and Sparc instruction selection did not handle
3246 :     REM in the first place, and REM is supported by the x86 machine coder.)
3247 :    
3248 :     * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
3249 :     selection phase. (The two can be streamlined because the hardware
3250 :     delivers both quotient and remainder at the same time anyway.)
3251 :    
3252 :     * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
3253 :     (Currently the behavior is inconsistent both across architectures and
3254 :     wrt. the draft Basis spec.)
3255 :    
3256 :     * Word8 should eventually be handled natively, too.
3257 :    
3258 :     * There seems to be one serious bug in mltree-gen.sml. It appears, though,
3259 :     as if there currently is no execution path that could trigger it in
3260 :     SML/NJ. (The assumptions underlying functions arith and promotable do not
3261 :     hold for things like multiplication and division.)
3262 :    
3263 :     ----------------------------------------------------------------------
3264 :     Name: Matthias Blume
3265 : blume 1181 Date: 2002/03/27 16:27:00 EST
3266 :     Tag: blume-20020327-mlrisc-divisions
3267 :     Description:
3268 :    
3269 :     Added support for all four division operations (ML's div, mod, quot,
3270 :     and rem) to MLRISC. In the course of doing so, I also rationalized
3271 :     the naming (no more annoying switch-around of DIV and QUOT), by
3272 :     parameterizing the operation by div_rounding_mode (which can be either
3273 :     DIV_TO_ZERO or DIV_TO_NEGINF).
3274 :    
3275 :     The generic MLTreeGen functor takes care of compiling all four
3276 :     operations down to only round-to-zero div.
3277 :    
3278 :     Missing pieces:
3279 :    
3280 :     * Doing something smarter than relying on MLTreeGen on architectures
3281 :     like, e.g., the x86 where hardware division delivers both quotient and
3282 :     remainder at the same time. With this, the implementation of the
3283 :     round-to-neginf operations could be further streamlined.
3284 :    
3285 :     * Remove inlining support for div/mod/rem from the frontend and replace it
3286 :     with primops that get carried through to the backend. Do this for all
3287 :     int and word types.
3288 :    
3289 :     ----------------------------------------------------------------------
3290 :     Name: Matthias Blume
3291 : blume 1180 Date: 2002/03/25 17:25:00 EST
3292 :     Tag: blume-20020325-divmod
3293 :     Description:
3294 :    
3295 :     I improved (hopefully without breaking them) the implementation of Int.div,
3296 :     Int.mod, and Int.rem. For this, the code in translate.sml now takes
3297 :     advantage of the following observations:
3298 :    
3299 :     Let q = x quot y r = x rem y
3300 :     d = x div y m = x mod y
3301 :    
3302 :     where "quot" is the round-to-zero version of integer division that
3303 :     hardware usually provides. Then we have:
3304 :    
3305 :     r = x - q * y where neither the * nor the - will overflow
3306 :     d = if q >= 0 orelse x = q * y then q else q - 1
3307 :     where neither the * nor the - will overflow
3308 :     m = if q >= 0 orelse r = 0 then r else r + y
3309 :     where the + will not overflow
3310 :    
3311 :     This results in substantial simplification of the generated code.
3312 :     The following table shows the number of CFG nodes and edges generated
3313 :     for
3314 :     fun f (x, y) = x OPER y
3315 :     (* with OPER \in div, mod, quot, rem *)
3316 :    
3317 :    
3318 :     OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
3319 :     --------------------------------------------------------
3320 :     div | 24 | 39 | 12 | 16
3321 :     mod | 41 | 71 | 12 | 16
3322 :     quot | 8 | 10 | 8 | 10
3323 :     rem | 10 | 14 | 8 | 10
3324 :    
3325 :    
3326 :     ----------------------------------------------------------------------
3327 :     Name: Matthias Blume
3328 : blume 1179 Date: 2002/03/25 22:06:00 EST
3329 :     Tag: blume-20020325-cprotobug
3330 :     Description:
3331 :    
3332 :     Fixed a bug in cproto (c prototype decoder).
3333 :    
3334 :     ----------------------------------------------------------------------
3335 :     Name: Matthias Blume
3336 : blume 1178 Date: 2002/03/25 16:00:00 EST
3337 :     Tag: blume-20020325-raw-primops
3338 :     Description:
3339 :    
3340 :     I did some cleanup to Allen's new primop code and
3341 :     replaced yesterday's bootfiles with new ones.
3342 :     (But they are stored in the same place.)
3343 :    
3344 :     ----------------------------------------------------------------------
3345 :     Name: Matthias Blume
3346 : blume 1175 Date: 2002/03/24 22:40:00 EST
3347 :     Tag: blume-20020324-bootfiles
3348 :     Description:
3349 :    
3350 :     Made the bootfiles that Allen asked for.
3351 :    
3352 :     ----------------------------------------------------------------------
3353 : leunga 1174 Name: Allen Leung
3354 :     Date: 2002/03/23 15:50:00 EST
3355 :     Tag: leunga-20020323-flint-cps-rcc-primops
3356 :     Description:
3357 :    
3358 :     1. Changes to FLINT primops:
3359 :    
3360 :     (* make a call to a C-function;
3361 :     * The primop carries C function prototype information and specifies
3362 :     * which of its (ML-) arguments are floating point. C prototype
3363 :     * information is for use by the backend, ML information is for
3364 :     * use by the CPS converter. *)
3365 :     | RAW_CCALL of { c_proto: CTypes.c_proto,
3366 :     ml_args: ccall_type list,
3367 :     ml_res_opt: ccall_type option,
3368 :     reentrant : bool
3369 :     } option
3370 :     (* Allocate uninitialized storage on the heap.
3371 :     * The record is meant to hold short-lived C objects, i.e., they
3372 : blume 2155 * are not ML pointers. With the tag, the representation is
3373 : leunga 1174 * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
3374 :     *)
3375 : blume 2155 | RAW_RECORD of {tag:bool,sz:int}
3376 : leunga 1174 and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
3377 :    
3378 :     2. These CPS primops are now overloaded:
3379 :    
3380 :     rawload of {kind:numkind}
3381 :     rawstore of {kind:numkind}
3382 :    
3383 :     The one argument form is:
3384 :    
3385 :     rawload {kind} address
3386 :    
3387 :     The two argument form is:
3388 : blume 2155
3389 : leunga 1174 rawload {kind} [ml object, byte-offset]
3390 :    
3391 :     3. RAW_CCALL/RCC now takes two extra arguments:
3392 :    
3393 :     a. The first is whether the C call is reentrant, i.e., whether
3394 :     ML state should be saved and restored.
3395 :     b. The second argument is a string argument specifying the name of
3396 :     library and the C function.
3397 :    
3398 :     These things are currently not handled in the code generator, yet.
3399 :    
3400 :     4. In CProto,
3401 :    
3402 : blume 2155 An encoding type of "bool" means "ml object" and is mapped into
3403 :     C prototype of PTR. Note that "bool" is different than "string",
3404 : leunga 1174 even though "string" is also mapped into PTR, because "bool"
3405 :     is assigned an CPS type of BOGt, while "string" is assigned INT32t.
3406 :    
3407 :     5. Pickler/unpicker
3408 :    
3409 :     Changed to handle RAW_RECORD and newest RAW_CCALL
3410 :    
3411 :     6. MLRiscGen,
3412 :    
3413 :     1. Changed to handle the new rawload/rawstore/rawrecord operators.
3414 :     2. Code for handling C Calls has been moved to a new module CPSCCalls,
3415 :     in the file CodeGen/cpscompile/cps-c-calls.sml
3416 :    
3417 :     7. Added the conditional move operator
3418 :    
3419 :     condmove of branch
3420 :    
3421 : blume 2155 to cps. Generation of this is still buggy so it is currently
3422 : leunga 1174 disabled.
3423 : blume 2155
3424 : leunga 1174 ----------------------------------------------------------------------
3425 : george 1168 Name: Lal George
3426 : blume 2155 Date: 2002/03/22 14:18:25 EST
3427 : george 1171 Tag: george-20020322-cps-branch-prob
3428 : george 1168 Description:
3429 :    
3430 : blume 2155 Implemented the Ball-Larus branch prediction-heuristics, and
3431 : george 1168 incorporated graphical viewers for control flow graphs.
3432 :    
3433 : george 1171 Ball-Larus Heuristics:
3434 : george 1168 ---------------------
3435 :     See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
3436 :    
3437 :     By design it uses the Dempster-Shafer theory for combining
3438 :     probabilities. For example, in the function:
3439 :    
3440 :     fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
3441 :    
3442 :     the ball-larus heuristics predicts that the n=0 is unlikely
3443 :     (OH-heuristic), and the 'then' branch is unlikely because of the
3444 :     RH-heuristic -- giving the 'then' branch an even lower combined
3445 : george 1171 probability using the Dempster-Shafer theory.
3446 : george 1168
3447 :     Finally, John Reppy's loop analysis in MLRISC, further lowers the
3448 :     probability of the 'then' branch because of the loop in the else
3449 :     branch.
3450 :    
3451 :    
3452 :     Graphical Viewing:
3453 : blume 2155 ------------------
3454 : george 1168 I merely plugged in Allen's graphical viewers into the compiler. The
3455 :     additional code is not much. At the top level, saying:
3456 :    
3457 :     Control.MLRISC.getFlag "cfg-graphical-view" := true;
3458 :    
3459 :     will display the graphical view of the control flow graph just before
3460 :     back-patching. daVinci must be in your path for this to work. If
3461 :     daVinci is not available, then the default viewer can be changed
3462 :     using:
3463 :    
3464 :     Control.MLRISC.getString "viewer"
3465 :    
3466 :     which can be set to "dot" or "vcg" for the corresponding viewers. Of
3467 :     course, these viewers must be in your path.
3468 :    
3469 :     The above will display the compilation unit at the level of clusters,
3470 :     many of which are small, boring, and un-interesting. Also setting:
3471 :    
3472 :     Control.MLRISC.getInt "cfg-graphical-view_size"
3473 :    
3474 :     will display clusters that are larger than the value set by the above.
3475 :    
3476 : blume 2155
3477 : george 1168 ----------------------------------------------------------------------
3478 : blume 1159 Name: Matthias Blume
3479 :     Date: 2002/03/21 22:20:00 EST
3480 :     Tag: blume-20020321-kmp-bugfix
3481 :     Description:
3482 :    
3483 :     Changed the interface to the KMP routine in PreString and fixed
3484 :     a minor bug in one place where it was used.
3485 :    
3486 :     ----------------------------------------------------------------------
3487 : leunga 1156 Name: Allen Leung
3488 : leunga 1158 Date: 2002/03/21 20:30:00 EST
3489 :     Tag: leunga-20020321-cfg
3490 :     Description:
3491 :    
3492 :     Fixed a potential problem in cfg edge splitting.
3493 :    
3494 :     ----------------------------------------------------------------------
3495 :     Name: Allen Leung
3496 : leunga 1156 Date: 2002/03/21 17:15:00 EST
3497 :     Tag: leunga-20020321-x86-fp-cfg
3498 :     Description:
3499 :    
3500 : blume 2155 1. Recoded the buggy parts of x86-fp.
3501 : leunga 1156
3502 :     a. All the block reordering code has been removed.
3503 :     We now depend on the block placement phases to do this work.
3504 :    
3505 :     b. Critical edge splitting code has been simplified and moved into the
3506 :     CFG modules, as where they belong.
3507 :    
3508 :     Both of these were quite buggy and complex. The code is now much, much
3509 :     simpler.
3510 :    
3511 :     2. X86 backend.
3512 :    
3513 :     a. Added instructions for 64-bit support. Instruction selection for
3514 :     64-bit has not been committed, however, since that
3515 : blume 2155 requires changes to MLTREE which haven't been approved by
3516 : leunga 1156 Lal and John.
3517 :    
3518 : blume 2155 b. Added support for FUCOMI and FUCOMIP when generating code for
3519 : leunga 1156 PentiumPro and above. We only generate these instructions in
3520 :     the fast-fp mode.
3521 :    
3522 : blume 2155 c. Added cases for JP and JNP in X86FreqProps.
3523 :    
3524 : leunga 1156 3. CFG
3525 :    
3526 :     CFG now has a bunch of methods for edge splitting and merging.
3527 :    
3528 :     4. Machine description.
3529 :    
3530 :     John's simplification of MLTREE_BASIS.fcond broke a few machine
3531 :     description things:
3532 :    
3533 :     rtl-build.{sig,sml} and hppa.mdl fixed.
3534 :    
3535 : blume 2155 NOTE: the machine description stuff in the repository is still broken.
3536 :     Again, I can't put my fixes in because that involves
3537 :     changes to MLTREE.
3538 : leunga 1156
3539 :     ----------------------------------------------------------------------
3540 : blume 1145 Name: Matthias Blume
3541 : blume 1155 Date: 2002/03/20 15:55:00 EST
3542 :     Tag: blume-20020320-kmp
3543 :     Description:
3544 :    
3545 :     Implemented Knuth-Morris-Pratt string matching in PreString and used
3546 :     it for String.isSubstring, Substring.isSubstring, and
3547 :     Substring.position.
3548 :    
3549 :     (Might need some stress-testing. Simple examples worked fine.)
3550 :    
3551 :     ----------------------------------------------------------------------
3552 :     Name: Matthias Blume
3553 : blume 1152 Date: 2002/03/19 16:37:00 EST
3554 :     Tag: blume-20020319-witnesses
3555 :     Description:
3556 :    
3557 :     Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
3558 :    
3559 :     This implements a generic mechanism for changing constness qualifiers
3560 :     anywhere within big C types without resorting to outright "casts".
3561 :     (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
3562 :     modify the constness at the outermost level.)
3563 :     The implementation of "convert" is based on the idea of "witness"
3564 :     values -- values that are not used by the operation but whose types
3565 :     "testify" to their applicability. On the implementation side, "convert"
3566 :     is simply a projection (returning its second curried argument). With
3567 :     cross-module inlining, it should not result in any machine code being
3568 :     generated.
3569 :    
3570 :     ----------------------------------------------------------------------
3571 :     Name: Matthias Blume
3572 : blume 1148 Date: 2002/03/15 16:40:00 EST
3573 :     Tag: blume-20020315-basis
3574 :     Description:
3575 :    
3576 :     Provided (preliminary?) implementations for
3577 :    
3578 :     {String,Substring}.{concatWith,isSuffix,isSubstring}
3579 :    
3580 :     and
3581 :    
3582 :     Substring.full
3583 :    
3584 :     Those are in the Basis spec but they were missing in SML/NJ.
3585 :    
3586 :     ----------------------------------------------------------------------
3587 :     Name: Matthias Blume
3588 : blume 1145 Date: 2002/03/14 21:30:00 EST
3589 :     Tag: blume-20020314-controls
3590 :     Description:
3591 :    
3592 :     Controls:
3593 :     ---------
3594 :    
3595 :     1. Factored out the recently-added Controls : CONTROLS stuff and put
3596 :     it into its own library $/controls-lib.cm. The source tree for
3597 :     this is under src/smlnj-lib/Controls.
3598 :    
3599 :     2. Changed the names of types and functions in this interface, so they
3600 :     make a bit more "sense":
3601 :    
3602 :     module -> registry
3603 :     'a registry -> 'a group
3604 :    
3605 :     3. The interface now deals in ref cells only. The getter/setter interface
3606 :     is (mostly) gone.
3607 :    
3608 :     4. Added a function that lets one register an already-existing ref cell.
3609 :    
3610 :     5. Made the corresponding modifications to the rest of the code so that
3611 :     everything compiles again.
3612 :    
3613 :     6. Changed the implementation of Controls.MLRISC back to something closer
3614 :     to the original. In particular, this module (and therefore MLRISC)
3615 :     does not depend on Controls. There now is some link-time code in
3616 :     int-sys.sml that registers the MLRISC controls with the Controls
3617 :     module.
3618 :    
3619 :     CM:
3620 :     ---
3621 :    
3622 :     * One can now specify the lambda-split aggressiveness in init.cmi.
3623 :    
3624 :     ----------------------------------------------------------------------
3625 : leunga 1142 Name: Allen Leung
3626 :     Date: 2002/03/13 17:30:00 EST
3627 :     Tag: leunga-20020313-x86-fp-unary
3628 :     Description:
3629 :    
3630 :     Bug fix for:
3631 :    
3632 :     > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
3633 :     > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
3634 :     > - fun f(x,(y,z)) = Real.~ y;
3635 :     > [autoloading]
3636 :     > [autoloading done]
3637 :     > fchsl (%eax), 184(%esp)
3638 :     > Error: MLRisc bug: X86MCEmitter.emitInstr
3639 : blume 2155 >
3640 : leunga 1142 > uncaught exception Error
3641 :     > raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
3642 :    
3643 :     The problem was that the code generator did not generate any fp registers
3644 :     in this case, and the ra didn't know that it needed to run the X86FP phase to
3645 :     translate the pseudo fp instruction. This only happened with unary fp
3646 :     operators in certain situations.
3647 :    
3648 :     ----------------------------------------------------------------------
3649 : blume 1137 Name: Matthias Blume
3650 : blume 1140 Date: 2002/03/13 14:00:00 EST
3651 :     Tag: blume-20020313-overload-etc
3652 :     Description:
3653 :    
3654 :     1. Added _overload as a synonym for overload for backward compatibility.
3655 :     (Control.overloadKW must be true for either version to be accepted.)
3656 :    
3657 :     2. Fixed bug in install script that caused more things to be installed
3658 :     than what was requested in config/targets.
3659 :    
3660 :     3. Made CM aware of the (_)overload construct so that autoloading
3661 :     works.
3662 :    
3663 :     ----------------------------------------------------------------------
3664 :     Name: Matthias Blume
3665 : blume 1138 Date: 2002/03/12 22:03:00 EST
3666 :     Tag: blume-20020312-url
3667 :     Description:
3668 :    
3669 :     Forgot to update BOOT and srcarchiveurl.
3670 :    
3671 :     ----------------------------------------------------------------------
3672 :     Name: Matthias Blume
3673 : blume 1137 Date: 2002/03/12 17:30:00 EST
3674 :     Tag: blume-20020312-version110392
3675 :     Description:
3676 :    
3677 :     Yet another version number bump (because of small changes to the
3678 :     binfile format). Version number is now 110.39.2. NEW BOOTFILES!
3679 :    
3680 :     Changes:
3681 :    
3682 :     The new pid generation scheme described a few weeks ago was overly
3683 :     complicated. I implemented a new mechanism that is simpler and
3684 :     provides a bit more "stability": Once CM has seen a compilation
3685 :     unit, it keeps its identity constant (as long as you do not delete
3686 :     those crucial CM/GUID/* files). This means that when you change
3687 : blume 1145 an interface, compile, then go back to the old interface, and
3688 : blume 1137 compile again, you arrive at the original pid.
3689 :    
3690 :     There now also is a mechanism that instructs CM to use the plain
3691 :     environment hash as a module's pid (effectively making its GUID
3692 :     the empty string). For this, "noguid" must be specified as an
3693 :     option to the .sml file in question within its .cm file.
3694 :     This is most useful for code that is being generated by tools such
3695 :     as ml-nlffigen (because during development programmers tend to
3696 :     erase the tool's entire output directory tree including CM's cached
3697 :     GUIDs). "noguid" is somewhat dangerous (since it can be used to locally
3698 :     revert to the old, broken behavior of SML/NJ, but in specific cases
3699 :     where there is no danger of interface confusion, its use is ok
3700 :     (I think).
3701 :    
3702 :     ml-nlffigen by default generates "noguid" annotations. They can be
3703 :     turned off by specifying -guid in its command line.
3704 :    
3705 :     ----------------------------------------------------------------------
3706 : george 1136 Name: Lal George
3707 :     Date: 2002/03/12 12 14:42:36 EST
3708 :     Tag: george-20020312-frequency-computation
3709 :     Description:
3710 : blume 2155
3711 : george 1136 Integrated jump chaining and static block frequency into the
3712 :     compiler. More details and numbers later.
3713 : leunga 1025
3714 : george 1003 ----------------------------------------------------------------------
3715 : george 1133 Name: Lal George
3716 :     Date: 2002/03/11 11 22:38:53 EST
3717 :     Tag: george-20020311-jump-chain-elim
3718 :     Description:
3719 :    
3720 :     Tested the jump chain elimination on all architectures (except the
3721 :     hppa). This is on by default right now and is profitable for the
3722 :     alpha and x86, however, it may not be profitable for the sparc and ppc
3723 :     when compiling the compiler.
3724 :    
3725 : blume 2155 The gc test will typically jump to a label at the end of the cluster,
3726 :     where there is another jump to an external cluster containing the actual
3727 : george 1133 code to invoke gc. This is to allow factoring of common gc invocation
3728 :     sequences. That is to say, we generate:
3729 :    
3730 :     f:
3731 :     testgc
3732 :     ja L1 % jump if above to L1
3733 :    
3734 :     L1:
3735 :     jmp L2
3736 :    
3737 :    
3738 :     After jump chain elimination the 'ja L1' instructions is converted to
3739 :     'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
3740 :     up being implemented in their long form (if L2 is far away) using:
3741 :    
3742 :     jbe L3 % jump if below or equal to L3
3743 :     jmp L2
3744 :     L3:
3745 :     ...
3746 :    
3747 :    
3748 :     For large compilation units L2 may be far away.
3749 :    
3750 :    
3751 :     ----------------------------------------------------------------------
3752 : blume 1132 Name: Matthias Blume
3753 :     Date: 2002/03/11 13:30:00 EST
3754 :     Tag: blume-20020311-mltreeeval
3755 :     Description:
3756 :    
3757 :     A functor parameter was missing.
3758 :    
3759 :     ----------------------------------------------------------------------
3760 : leunga 1130 Name: Allen Leung
3761 : leunga 1131 Date: 2002/03/11 10:30:00 EST
3762 : leunga 1142 Tag: leunga-20020311-runtime-string0
3763 : leunga 1131 Description:
3764 :    
3765 : blume 2155 The representation of the empty string now points to a
3766 :     legal null terminated C string instead of unit. It is now possible
3767 :     to convert an ML string into C string with InlineT.CharVector.getData.
3768 : leunga 1131 This compiles into one single machine instruction.
3769 :    
3770 :     ----------------------------------------------------------------------
3771 :     Name: Allen Leung
3772 : leunga 1130 Date: 2002/03/10 23:55:00 EST
3773 :     Tag: leunga-20020310-x86-call
3774 :     Description:
3775 :    
3776 :     Added machine generation for CALL instruction (relative displacement mode)
3777 :    
3778 :     ----------------------------------------------------------------------
3779 : blume 1128 Name: Matthias Blume
3780 :     Date: 2002/03/08 16:05:00
3781 :     Tag: blume-20020308-entrypoints
3782 :     Description:
3783 :    
3784 :     Version number bumped to 110.39.1. NEW BOOTFILES!
3785 :    
3786 :     Entrypoints: non-zero offset into a code object where execution should begin.
3787 :    
3788 :     - Added the notion of an entrypoint to CodeObj.
3789 :     - Added reading/writing of entrypoint info to Binfile.
3790 :     - Made runtime system bootloader aware of entrypoints.
3791 :     - Use the address of the label of the first function given to mlriscGen
3792 :     as the entrypoint. This address is currently always 0, but it will
3793 :     not be 0 once we turn on block placement.
3794 :     - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3795 :     with entry points) from mlriscGen.
3796 :    
3797 :     ----------------------------------------------------------------------
3798 : leunga 1127 Name: Allen Leung
3799 :     Date: 2002/03/07 20:45:00 EST
3800 :     Tag: leunga-20020307-x86-cmov
3801 :     Description:
3802 :    
3803 :     Bug fixes for CMOVcc on x86.
3804 :    
3805 :     1. Added machine code generation for CMOVcc
3806 :     2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3807 :     3. CMOVcc cannot have an immediate operand as argument.
3808 :    
3809 :     ----------------------------------------------------------------------
3810 : blume 1126 Name: Matthias Blume
3811 :     Date: 2002/03/07 16:15:00 EST
3812 :     Tag: blume-20020307-controls
3813 :     Description:
3814 :    
3815 :     This is a very large but mostly boring patch which makes (almost)
3816 :     every tuneable compiler knob (i.e., pretty much everything under
3817 :     Control.* plus a few other things) configurable via both the command
3818 :     line and environment variables in the style CM did its configuration
3819 :     until now.
3820 :    
3821 :     Try starting sml with '-h' (or, if you are brave, '-H')
3822 :    
3823 :     To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3824 :     implements the underlying generic mechanism.
3825 :    
3826 :     The interface to some of the existing such facilities has changed somewhat.
3827 :     For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3828 :     (The getFoo interface is still there for backward-compatibility, but its
3829 :     use is deprecated.)
3830 :    
3831 :     The ml-build script passes -Cxxx=yyy command-line arguments through so
3832 :     that one can now twiddle the compiler settings when using this "batch"
3833 :     compiler.
3834 :    
3835 :     TODO items:
3836 :    
3837 :     We should go through and throw out all controls that are no longer
3838 :     connected to anything. Moreover, we should go through and provide
3839 :     meaningful (and correct!) documentation strings for those controls
3840 :     that still are connected.
3841 :    
3842 :     Currently, multiple calls to Controls.new are accepted (only the first
3843 :     has any effect). Eventually we should make sure that every control
3844 :     is being made (via Controls.new) exactly once. Future access can then
3845 :     be done using Controls.acc.
3846 :    
3847 :     Finally, it would probably be a good idea to use the getter-setter
3848 :     interface to controls rather than ref cells. For the time being, both
3849 :     styles are provided by the Controls module, but getter-setter pairs are
3850 :     better if thread-safety is of any concern because they can be wrapped.
3851 :    
3852 :     *****************************************
3853 :    
3854 :     One bug fix: The function blockPlacement in three of the MLRISC
3855 :     backpatch files used to be hard-wired to one of two possibilities at
3856 :     link time (according to the value of the placementFlag). But (I
3857 :     think) it should rather sense the flag every time.
3858 :    
3859 :     *****************************************
3860 :    
3861 :     Other assorted changes (by other people who did not supply a HISTORY entry):
3862 :    
3863 :     1. the cross-module inliner now works much better (Monnier)
3864 :     2. representation of weights, frequencies, and probabilities in MLRISC
3865 :     changed in preparation of using those for weighted block placement
3866 :     (Reppy, George)
3867 :    
3868 :     ----------------------------------------------------------------------
3869 : george 1116 Name: Lal George
3870 : george 1124 Date: 2002/03/07 14:44:24 EST 2002
3871 :     Tag: george-20020307-weighted-block-placement
3872 :    
3873 :     Tested the weighted block placement optimization on all architectures
3874 :     (except the hppa) using AMPL to generate the block and edge frequencies.
3875 :     Changes were required in the machine properties to correctly
3876 : blume 2155 categorize trap instructions. There is an MLRISC flag
3877 :     "weighted-block-placement" that can be used to enable weighted block
3878 : george 1124 placement, but this will be ineffective without block/edge
3879 :     frequencies (coming soon).
3880 :    
3881 :    
3882 :     ----------------------------------------------------------------------
3883 :     Name: Lal George
3884 : blume 2155 Date: 2002/03/05 17:24:48 EST
3885 : george 1116 Tag: george-20020305-linkage-cluster
3886 :    
3887 : george 1124 In order to support the block placement optimization, a new cluster
3888 :     is generated as the very first cluster (called the linkage cluster).
3889 : blume 2155 It contains a single jump to the 'real' entry point for the compilation
3890 : george 1124 unit. Block placement has no effect on the linkage cluster itself, but
3891 :     all the other clusters have full freedom in the manner in which they
3892 : blume 2155 reorder blocks or functions.
3893 : george 1116
3894 :     On the x86 the typical linkage code that is generated is:
3895 :     ----------------------
3896 :     .align 2
3897 :     L0:
3898 :     addl $L1-L0, 72(%esp)
3899 : george 1124 jmp L1
3900 : george 1116
3901 :    
3902 :     .align 2
3903 :     L1:
3904 :     ----------------------
3905 :    
3906 : blume 2155 72(%esp) is the memory location for the stdlink register. This
3907 : george 1116 must contain the address of the CPS function being called. In the
3908 : george 1124 above example, it contains the address of L0; before
3909 : george 1116 calling L1 (the real entry point for the compilation unit), it
3910 :     must contain the address for L1, and hence
3911 :    
3912 :     addl $L1-L0, 72(%esp)
3913 :    
3914 : george 1124 I have tested this on all architectures except the hppa.The increase
3915 :     in code size is of course negligible
3916 : george 1116
3917 :     ----------------------------------------------------------------------
3918 : leunga 1115 Name: Allen Leung
3919 :     Date: 2002/03/03 13:20:00 EST
3920 :     Tag: leunga-20020303-mlrisc-tools
3921 :    
3922 :     Added #[ ... ] expressions to mlrisc tools
3923 :    
3924 :     ----------------------------------------------------------------------
3925 : blume 1096 Name: Matthias Blume
3926 : blume 1098 Date: 2002/02/27 12:29:00 EST
3927 :     Tag: blume-20020227-cdebug
3928 :     Description:
3929 :    
3930 :     - made types in structure C and C_Debug to be equal
3931 :     - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3932 :     - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3933 :    
3934 :     ----------------------------------------------------------------------
3935 :     Name: Matthias Blume
3936 : blume 1096 Date: 2002/02/26 12:00:00 EST
3937 :     Tag: blume-20020226-ffi
3938 :     Description:
3939 :    
3940 :     1. Fixed a minor bug in CM's "noweb" tool:
3941 :     If numbering is turned off, then truly don't number (i.e., do not
3942 :     supply the -L option to noweb). The previous behavior was to supply
3943 :     -L'' -- which caused noweb to use the "default" line numbering scheme.
3944 :     Thanks to Chris Richards for pointing this out (and supplying the fix).
3945 :    
3946 :     2. Once again, I reworked some aspects of the FFI:
3947 :    
3948 :     A. The incomplete/complete type business:
3949 :    
3950 :     - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3951 :     gone!
3952 :     - ML types representing an incomplete type are now *equal* to
3953 :     ML types representing their corresponding complete types (just like
3954 :     in C). This is still safe because ml-nlffigen will not generate
3955 :     RTTI for incomplete types, nor will it generate functions that
3956 :     require access to such RTTI. But when ML code generated from both
3957 :     incomplete and complete versions of the C type meet, the ML types
3958 :     are trivially interoperable.
3959 :    
3960 :     NOTE: These changes restore the full generality of the translation
3961 :     (which was previously lost when I eliminated functorization)!
3962 :    
3963 :     B. Enum types:
3964 :    
3965 :     - Structure C now has a type constructor "enum" that is similar to
3966 :     how the "su" constructor works. However, "enum" is not a phantom
3967 :     type because each "T enum" has values (and is isomorphic to
3968 :     MLRep.Signed.int).
3969 :     - There are generic access operations for enum objects (using
3970 :     MLRep.Signed.int).
3971 :     - ml-nlffigen will generate a structure E_foo for each "enum foo".
3972 :     * The structure contains the definition of type "mlrep" (the ML-side
3973 :     representation type of the enum). Normally, mlrep is the same
3974 :     as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3975 :     then mlrep will be defined as a datatype -- thus facilitating
3976 :     pattern matching on mlrep values.
3977 :     ("-ec" will be suppressed if there are duplicate values in an
3978 :     enumeration.)
3979 :     * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3980 :     will be generated for each C enum constant xxx.
3981 :     * Conversion functions m2i and i2m convert between mlrep and
3982 :     MLRep.Signed.int. (Without "-ec", these functions are identities.)
3983 :     * Coversion functions c and ml convert between mlrep and "tag enum".
3984 :     * Access functions (get/set) fetch and store mlrep values.
3985 :     - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3986 :     enumerations are merged into one single enumeration represented by
3987 :     structure E_'.
3988 :    
3989 :     ----------------------------------------------------------------------
3990 : leunga 1080 Name: Allen Leung
3991 : leunga 1094 Date: 2002/02/25 04:45:00 EST
3992 :     Tag: leunga-20020225-cps-spill
3993 :    
3994 :     This is a new implementation of the CPS spill phase.
3995 :     The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3996 :     In case of problems, replace it with the old file spill.sml
3997 :    
3998 : blume 2155 The current compiler runs into some serious performance problems when
3999 : leunga 1094 constructing a large record. This can happen when we try to compile a
4000 :     structure with many items. Even a very simple structure like the following
4001 :     makes the compiler slow down.
4002 :    
4003 :     structure Foo = struct
4004 :     val x_1 = 0w1 : Word32.int
4005 :     val x_2 = 0w2 : Word32.int
4006 :     val x_3 = 0w3 : Word32.int
4007 :     ...
4008 :     val x_N = 0wN : Word32.int
4009 :     end
4010 :    
4011 : blume 2155 The following table shows the compile time, from N=1000 to N=4000,
4012 : leunga 1094 with the old compiler:
4013 :    
4014 :     N
4015 : blume 2155 1000 CPS 100 spill 0.04u 0.00s 0.00g
4016 :     MLRISC ra 0.06u 0.00s 0.05g
4017 : leunga 1094 (spills = 0 reloads = 0)
4018 : blume 2155 TOTAL 0.63u 0.07s 0.21g
4019 : leunga 1094
4020 : blume 2155 1100 CPS 100 spill 8.25u 0.32s 0.64g
4021 :     MLRISC ra 5.68u 0.59s 3.93g
4022 : leunga 1094 (spills = 0 reloads = 0)
4023 : blume 2155 TOTAL 14.71u 0.99s 4.81g
4024 : leunga 1094
4025 : blume 2155 1500 CPS 100 spill 58.55u 2.34s 1.74g
4026 :     MLRISC ra 5.54u 0.65s 3.91g
4027 : leunga 1094 (spills = 543 reloads = 1082)
4028 : blume 2155 TOTAL 65.40u 3.13s 6.00g
4029 : leunga 1094
4030 : blume 2155 2000 CPS 100 spill 126.69u 4.84s 3.08g
4031 :     MLRISC ra 0.80u 0.10s 0.55g
4032 : leunga 1094 (spills = 42 reloads = 84)
4033 : blume 2155 TOTAL 129.42u 5.10s 4.13g
4034 : leunga 1094
4035 : blume 2155 3000 CPS 100 spill 675.59u 19.03s 11.64g
4036 :     MLRISC ra 2.69u 0.27s 1.38g
4037 : leunga 1094 (spills = 62 reloads = 124)
4038 : blume 2155 TOTAL 682.48u 19.61s 13.99g
4039 : leunga 1094
4040 : blume 2155 4000 CPS 100 spill 2362.82u 56.28s 43.60g
4041 :     MLRISC ra 4.96u 0.27s 2.72g
4042 : leunga 1094 (spills = 85 reloads = 170)
4043 : blume 2155 TOTAL 2375.26u 57.21s 48.00g
4044 : leunga 1094
4045 : blume 2155 As you can see the old cps spill module suffers from some serious
4046 :     performance problem. But since I cannot decipher the old code fully,
4047 :     instead of patching the problems up, I'm reimplementing it
4048 :     with a different algorithm. The new code is more modular,
4049 :     smaller when compiled, and substantially faster
4050 :     (O(n log n) time and O(n) space). Timing of the new spill module:
4051 : leunga 1094
4052 : blume 2155 4000 CPS 100 spill 0.02u 0.00s 0.00g
4053 :     MLRISC ra 0.25u 0.02s 0.15g
4054 :     (spills=1 reloads=3)
4055 :     TOTAL 7.74u 0.34s 1.62g
4056 : leunga 1094
4057 :     Implementation details:
4058 :    
4059 : blume 2155 As far as I can tell, the purpose of the CPS spill module is to make sure the
4060 :     number of live variables at any program point (the bandwidth)
4061 :     does not exceed a certain limit, which is determined by the
4062 :     size of the spill area.
4063 : leunga 1094
4064 : blume 2155 When the bandwidth is too large, we decrease the register pressure by
4065 : leunga 1094 packing live variables into spill records. How we achieve this is
4066 :     completely different than what we did in the old code.
4067 :    
4068 : blume 2155 First, there is something about the MLRiscGen code generator
4069 : leunga 1094 that we should be aware of:
4070 :    
4071 :     o MLRiscGen performs code motion!
4072 : blume 2155
4073 : leunga 1094 In particular, it will move floating point computations and
4074 : blume 2155 address computations involving only the heap pointer to
4075 :     their use sites (if there is only a single use).
4076 : leunga 1094 What this means is that if we have a CPS record construction
4077 :     statement
4078 : blume 2155
4079 : leunga 1094 RECORD(k,vl,w,e)
4080 : blume 2155
4081 :     we should never count the new record address w as live if w
4082 : leunga 1094 has only one use (which is often the case).
4083 : blume 2155
4084 : leunga 1094 We should do something similar to floating point, but the transformation
4085 :     there is much more complex, so I won't deal with that.
4086 :    
4087 :     Secondly, there are now two new cps primops at our disposal:
4088 :    
4089 : blume 2155 1. rawrecord of record_kind option
4090 : leunga 1094 This pure operator allocates some uninitialized storage from the heap.
4091 :     There are two forms:
4092 : blume 2155
4093 : leunga 1094 rawrecord NONE [INT n] allocates a tagless record of length n
4094 :     rawrecord (SOME rk) [INT n] allocates a tagged record of length n
4095 :     and initializes the tag.
4096 : blume 2155
4097 : leunga 1094 2. rawupdate of cty
4098 : blume 2155 rawupdate cty (v,i,x)
4099 : leunga 1094 Assigns to x to the ith component of record v.
4100 :     The storelist is not updated.
4101 : blume 2155
4102 : leunga 1094 We use these new primops for both spilling and increment record construction.
4103 : blume 2155
4104 : leunga 1094 1. Spilling.
4105 : blume 2155
4106 : leunga 1094 This is implemented with a linear scan algorithm (but generalized
4107 :     to trees). The algorithm will create a single spill record at the
4108 :     beginning of the cps function and use rawupdate to spill to it,
4109 :     and SELECT or SELp to reload from it. So both spills and reloads
4110 : blume 2155 are fine-grain operations. In contrast, in the old algorithm
4111 :     "spills" have to be bundled together in records.
4112 :    
4113 : leunga 1094 Ideally, we should sink the spill record construction to where
4114 :     it is needed. We can even split the spill record into multiple ones
4115 :     at the places where they are needed. But CPS is not a good
4116 :     representation for global code motion, so I'll keep it simple and
4117 :     am not attempting this.
4118 : blume 2155
4119 : leunga 1094 2. Incremental record construction (aka record splitting).
4120 :    
4121 :     Long records with many component values which are simulatenously live
4122 : blume 2155 (recall that single use record addresses are not considered to
4123 : leunga 1094 be live) are constructed with rawrecord and rawupdate.
4124 :     We allocate space on the heap with rawrecord first, then gradually
4125 :     fill it in with rawupdate. This is the technique suggested to me
4126 :     by Matthias.
4127 : blume 2155
4128 : leunga 1094 Some restrictions on when this is applicable:
4129 :     1. It is not a VECTOR record. The code generator currently does not handle
4130 :     this case. VECTOR record uses double indirection like arrays.
4131 : blume 2155 2. All the record component values are defined in the same "basic block"
4132 :     as the record constructor. This is to prevent speculative
4133 :     record construction.
4134 : leunga 1094
4135 :     ----------------------------------------------------------------------
4136 :     Name: Allen Leung
4137 : leunga 1086 Date: 2002/02/22 01:02:00 EST
4138 :     Tag: leunga-20020222-mlrisc-tools
4139 :    
4140 :     Minor bug fixes in the parser and rewriter
4141 :    
4142 :     ----------------------------------------------------------------------
4143 :     Name: Allen Leung
4144 : leunga 1085 Date: 2002/02/21 20:20:00 EST
4145 :     Tag: leunga-20020221-peephole
4146 :    
4147 :     Regenerated the peephole files. Some contained typos in the specification
4148 :     and some didn't compile because of pretty printing bugs in the old version
4149 :     of 'nowhere'.
4150 :    
4151 :     ----------------------------------------------------------------------
4152 :     Name: Allen Leung
4153 : leunga 1079 Date: 2002/02/19 20:20:00 EST
4154 :     Tag: leunga-20020219-mlrisc-tools
4155 :     Description:
4156 :    
4157 :     Minor bug fixes to the mlrisc-tools library:
4158 :    
4159 :     1. Fixed up parsing colon suffixed keywords
4160 :     2. Added the ability to shut the error messages up
4161 :     3. Reimplemented the pretty printer and fixed up/improved
4162 :     the pretty printing of handle and -> types.
4163 :     4. Fixed up generation of literal symbols in the nowhere tool.
4164 :