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 2269 - (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 : jhr 2269 Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
18 :     Date: 2007/01/28
19 :     Description:
20 :    
21 :     Added the UTF8 structure and signature from the Moby compiler to
22 :     the SML/NJ library (and the CML library). These modules will replace
23 :     the version in the ml-lpt-lib.
24 :    
25 :     ----------------------------------------------------------------------
26 : blume 2237 Name: Matthias Blume (http://www.tti-c.org/blume/)
27 : blume 2267 Date: 2007/01/26 15:10:00 CST
28 :     Description:
29 :    
30 :     Added entries to handle ml-lpt-lib.cm in installer.
31 :    
32 :     ----------------------------------------------------------------------
33 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
34 : blume 2259 Date: 2006/12/15 11:10:00 CST
35 : blume 2237 Tag: release-110.61
36 :     Description:
37 :    
38 :     New working release (110.61). NEW BOOTFILES!
39 :    
40 : blume 2259 This supersedes the pre-brown-paper-bug release (see below).
41 :    
42 : blume 2237 ----------------------------------------------------------------------
43 : blume 2259 Name: Matthias Blume (http://www.tti-c.org/blume/)
44 :     Date: 2006/12/15 11:00:00 CST
45 :     Description:
46 :    
47 :     Fixed brown-paper-bag bug with CM's pathname handling, which made
48 :     installation fail under Win32.
49 :    
50 :     ----------------------------------------------------------------------
51 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
52 :     Date: 2006/12/14 12:00:00 CST
53 :     Description:
54 :    
55 :     Obsolete!
56 :    
57 :     [New working release (110.61). NEW BOOTFILES!]
58 :    
59 :     ----------------------------------------------------------------------
60 : jhr 2234 Name: John Reppy (http://cs.uchicago.edu/~jhr)
61 :     Date: 2006/14/06 09:50:00 CST
62 :     Description:
63 :    
64 :     Fixed the code in runtime/c-libs/posix-tty/{tcgetattr,tcsetattr}.c to
65 :     get the c_cc termios data copied correctly. Also moved the allocation
66 :     of the string to avoid problems if it caused a GC.
67 :    
68 :     Thanks to Timothy Bourke for the bug report and fix.
69 :    
70 :     ----------------------------------------------------------------------
71 : blume 2232 Name: Matthias Blume (http://www.tti-c.org/blume)
72 :     Date: 2006/12/08 09:50:00 CST
73 :     Description:
74 :    
75 :     Added code to CM's "standard shell tool" implementation which
76 :     causes it to tolerate (with a warning) the situation where the shell
77 :     command fails (e.g., due to the shell command's non-existence)
78 :     as long as all target files exist.
79 :    
80 :     This makes it possible to, e.g., build ml-yacc from sources even if
81 :     svn checkout messed up the time stamps on files in such a way that
82 :     yacc.grm is younger than yacc.grm.sml or yacc.grm.sig. (Ml-yacc would
83 :     be needed to re-process yacc.grm, but obviously it might not yet be
84 :     available at that time.)
85 :    
86 :     ----------------------------------------------------------------------
87 : jhr 2228 Name: John Reppy (http://cs.uchicago.edu/~jhr)
88 :     Date: 2006/12/06 12:06:00 CST
89 :     Description:
90 :    
91 :     Fixed the types of recvVecFrom, recvVecFrom', recvVecFromNB, and
92 :     recvVecFromNB' in the SOCKET signature. This error is actually
93 :     in the SML Basis specification too.
94 :    
95 :     ----------------------------------------------------------------------
96 : mblume 1923 Name: Matthias Blume (http://www.tti-c.org/blume/)
97 : blume 2226 Date: 2006/12/05 22:40:00 CST
98 :     Description:
99 :    
100 :     CM now reports undefined anchors as errors and aborts execution
101 :     rather than silently pressing on using bogus values.
102 :    
103 :     ----------------------------------------------------------------------
104 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
105 : blume 2224 Date: 2006/11/29 10:00:00 CST
106 :     Description:
107 :    
108 :     Use Say.vsay for printing the "[autoloading]" message, so
109 :     #set CM.Control.verbose false (or -Ccm.verbose=false) can be used
110 :     to suppress them.
111 :    
112 :     ----------------------------------------------------------------------
113 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
114 : blume 2219 Date: 2006/11/10 11:00:00 CST
115 :     Description:
116 :    
117 :     Fixed bug in CM where "with:" specifications that affect compilation
118 :     (as opposed to parsing) were ignored.
119 :    
120 :     ----------------------------------------------------------------------
121 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
122 : blume 2193 Date: 2006/11/09 14:30:00 CST
123 :     Tag: release-110.60
124 :     Description:
125 :    
126 :     New working release (110.60). NEW BOOTFILES!
127 :    
128 :     This is the first subversion-hosted release. There are also changes
129 :     to the directory layout. Here is a table:
130 :    
131 :     (All pathnames are relative to the SML/NJ "root" directory.)
132 :    
133 :     config -> config
134 :    
135 :     src/cm -> base/cm
136 :     src/compiler -> base/compiler
137 :     src/runtime -> base/runtime
138 :     src/system -> base/system
139 :    
140 :     src/cm/pgraph -> pgraph
141 :    
142 :     src/READMES -> base/READMES
143 :     src/$notefile -> base/NOTES/$notefile
144 :     for <notefile> in BOOT CVSNOTES CYGWININSTALL HISTORY INSTALL
145 :     MACOSXINSTALL WININSTALL
146 :    
147 :     src/smlnj-lib -> smlnj-lib
148 :     src/MLRISC -> MLRISC
149 :     ckit -> ckit
150 :     src/cml -> cml
151 :     src/eXene -> eXene
152 :     src/heap2asm -> heap2asm
153 :     src/lexgen -> lexgen
154 :     src/ml-burg -> ml-burg
155 :     src/ml-lex -> ml-lex
156 :     src/ml-yacc -> ml-yacc
157 :     src/ml-nlffi-lib -> nlffi/lib
158 :     src/ml-nlffigen -> nlffi/gen
159 :     src/smlnj-c -> smlnj-c
160 :     src/tools/TraceDebugProf -> trace-debug-profile
161 :    
162 :     In addition, there is also a new
163 :    
164 :     ml-lpt
165 :    
166 :     directory containing two new program generator tools: ml-ulex and
167 :     ml-antlr (a lexer generator that handles unicode and an ANTRL-inspired
168 :     LL(k) parser generator). These tools are currently "beta-quality"
169 :    
170 :     The latest versios of the sources can now be obtained anonymously
171 :     via subversion. For this, it is useful to first check out
172 :    
173 :     svn://smlnj-gforge.cs.uchicago.edu/smlnj/admin
174 :    
175 :     and put the resulting directory on your shell's PATH. This provides
176 :     access to three shell scripts: checkout-all.sh, stat-all.sh,
177 :     and refresh-all.sh.
178 :    
179 :     To create a freshly checked-out copy of the sources, do
180 :    
181 :     checkout-all.sh [dir]
182 :    
183 :     where dir is the optional SML/NJ root directory (default is ".").
184 :    
185 :     This creates the above directory layout. Each subdirectory of
186 :     the root is under individual subversion control. The stat-all.sh
187 :     and refresh-al.sh scripts apply "svn stat" or "svn update" to
188 :     each of these subtrees.
189 :    
190 :     ----------------------------------------------------------------------
191 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
192 : blume 2155 Date: 2006/11/02 12:20:00 CST
193 :     Description:
194 :    
195 :     Reorganized directory layout.
196 :    
197 :     This is a temporary solution, more reorganization is to come.
198 :    
199 :     The basic idea is to have a number of toplevel trees, each corresponding
200 :     to a well-defined part of the overall system. Each part can be
201 :     maintained individually, even in separate source repositories,
202 :     although currently we still serve everything out of the main
203 :     smlnj-gforge tree.
204 :    
205 :     Installer and and scripts have been updated to reflect the new
206 :     layout. The installer (base/src/system/smlnj/installer) is now
207 :     "scriptable" to avoid burning too much knowledge about the
208 :     layout into SML source code. The main script used by the installer
209 :     is in config/actions.
210 :    
211 :     The main change is that many of the subdirectories of what used to
212 :     be known as "src" have moved to the toplevel. The "src" directory
213 :     itself has moved down into a subtree called "base". (We may eventually
214 :     get rid of the extra level of indirection represented by "src".)
215 :    
216 :     The layout is now as follows:
217 :    
218 :     toplevel tree name | default repository
219 :     | (using gf=svn://smlnj-gforge.cs.uchicago.edu/smlnj)
220 :     -------------------+----------------------------------------------------
221 :     config | $gf/config/trunk
222 :     |
223 :     base | $gf/sml/trunk
224 :     |
225 :     smlnj-lib | $gf/smlnj-lib/trunk
226 :     MLRISC | $gf/MLRISC/trunk
227 :     |
228 :     ml-yacc | $gf/ml-yacc/trunk
229 :     ml-lex | $gf/ml-lex/trunk
230 :     ml-burg | $gf/ml-burg/trunk
231 :     lexgen | $gf/lexgen/trunk
232 :     heap2asm | $gf/heap2asm/trunk
233 :     |
234 :     cml | $gf/cml/trunk
235 :     eXene | $gf/eXene/trunk
236 :     ckit | $gf/ckit/trunk
237 :     nlffi | $gf/nlffi/trunk
238 :     |
239 :     smlnj-c | $gf/smlnj-c/trunk
240 :    
241 :     In $gf/admin there are a few useful shell scripts for checking out
242 :     and maintaining the entire collection of trees:
243 :    
244 :     admin/checkout-all.sh [dir]
245 :     -- optionally creates dir and checks out all trees from their
246 :     default repositories; if dir is missing, checkout into .
247 :     admin/refresh-all.sh [dir]
248 :     -- looks at all trees (from the above list) in dir (default: .)
249 :     and runs "svn update" if the tree exists and is under
250 :     subversion control;
251 :     non-existing or non-subversion trees are skipped
252 :     admin/stat-all.sh [dir]
253 :     -- like refresh-all.sh, but runs "svn stat" instead
254 :    
255 :     ----------------------------------------------------------------------
256 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
257 :     Date: 2006/10/05 11:30:00 CDT
258 : mblume 2066 Tag: blume-20061005-merge-amd64
259 :     Description:
260 :    
261 :     Merged code for AMD64 backend (Mike Rainey's work).
262 :     Everything is hooked up but untested.
263 :    
264 :     ----------------------------------------------------------------------
265 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
266 : mblume 1931 Date: 2006/05/17 14:00:00 CST
267 :     Tag: Release_110_59
268 :     Description:
269 :    
270 : blume 2193 New working version (110.59). NEW BOOTFILES!
271 : mblume 1931
272 :     I am freezing 110.59. Changes other than the version-number increase:
273 :    
274 :     eXene:
275 :    
276 :     * committed changes to eXene from Alley Stoughton:
277 :     "fixed bugs in X authorization and resource handling, as well
278 :     as in the pile and viewport widgets"
279 :    
280 :     Runtime:
281 :     * fixed linking problem with NetBSD 3.x.
282 :    
283 :     Lexgen:
284 :     * lexgen tool handles non-ascii characters in 7-bit mode the same
285 :     way that ml-lex does
286 :    
287 :     * lexgen propagates exceptions the same way that ml-lex does
288 :    
289 :     CML:
290 :     * Fixed a bug in the SyncVar polling functions (iGetPoll,
291 :     mTakePoll, and mGetPoll) that could lead to livelock.
292 :    
293 :     ----------------------------------------------------------------------
294 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
295 : mblume 1926 Date: 2006/05/12 15:20:00 CST
296 :     Tag: blume-20060515-makedepend
297 :     Description:
298 :    
299 :     Implemented ml-makedepend (i.e., CM.sources) in a better (more robust)
300 :     way. This should hopefully fix the ml-makedepend problem permanently.
301 :    
302 :     ----------------------------------------------------------------------
303 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
304 : mblume 1925 Date: 2006/05/12 15:15:00 CST
305 :     Tag: blume-20060512-makedepend
306 :     Description:
307 :    
308 :     Fixed long-standing bug with ml-makedepend where it would output a
309 :     spurious dependency to a non-existing file. (This is a simple fix.
310 :     It might need further looking into.)
311 :    
312 :     ----------------------------------------------------------------------
313 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
314 : mblume 1923 Date: 2006/04/20 10:30:00 CDT
315 :     Tag: blume-20060420-heap2exec
316 :     Description:
317 :    
318 :     Committed patches received from Johannes 5 Joemann (joemann@befree.free.de)
319 :     that enable heap2exec under Linux and FreeBSD.
320 :    
321 :     ----------------------------------------------------------------------
322 : jhr 1921 Name: John Reppy
323 :     Date: 2006/04/14
324 : blume 2155 Tag:
325 : jhr 1921 Description:
326 :    
327 :     MLRISC changes:
328 :     1) renamed GAS_PSEUDO_OPS to AS_PSEUDO_OPS and put it in its own file.
329 :     2) added support for NOTB and XORB operators in pseudo-op expressions
330 :     3) added DarwinPseudoOp functor that supports Darwin's assembler syntax.
331 :     4) added support for 64-bit integer literals
332 :    
333 :     ----------------------------------------------------------------------
334 : jhr 1922 Name: Matthias Blume (http://www.tti-c.org/blume/)
335 : mblume 1907 Date: yyyy/mm/dd
336 :     Tag: Release_110_58
337 :     Description:
338 :    
339 :     New working version (110.58). NEW BOOTFILES!
340 :    
341 :     ----------------------------------------------------------------------
342 : mblume 1896 Name: Matthias Blume (http://www.tti-c.org/blume/)
343 : mblume 1905 Date: 2006/03/01 22:15:00 CST
344 :     Tag: blume-20060301-lexgen
345 :     Description:
346 :    
347 :     Incorporated several bugfixes to lexgen. Compiler now compiles to
348 :     fixpoint when using lexgen instead of ml-lex.
349 :    
350 :     ----------------------------------------------------------------------
351 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
352 : mblume 1902 Date: 2006/02/28 22:45:00 CST
353 :     Tag: blume-20060228-lexgen
354 :     Description:
355 :    
356 :     Removed ml-flex and added lexgen instead, using Aaron Turon's newly
357 :     provided tarball. The generated lexers still have problems.
358 :    
359 :     ----------------------------------------------------------------------
360 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
361 : mblume 1898 Date: 2006/02/26 23:30:00 CST
362 :     Tag: blume-20060226-ml-flex
363 :     Description:
364 :    
365 :     Removed ml-flex' dependency on regexp-lib.cm. Turned installation
366 :     of ml-flex on by default.
367 :    
368 :     ----------------------------------------------------------------------
369 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
370 : mblume 1896 Date: 2006/02/24 16:30:00 CST
371 :     Tag: blume-20060224-ml-flex
372 :     Description:
373 :    
374 :     Added ml-flex sources. Partially integrated, but should not be
375 :     turned on yet! (Read: leave it commented-out in config/targets!)
376 :    
377 :     ----------------------------------------------------------------------
378 : jhr 1893 Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
379 :     Date: 2006/02/23
380 :     Tag: jhr-20060223-x86-64
381 :     Description:
382 :     Changes to support building on x86-64 systems (using the 32-bit mode).
383 :     Also cleaned up signal handling on Linux. Support for pre-2.2 Linux
384 :     kernels dropped.
385 :    
386 :     ----------------------------------------------------------------------
387 : allenleung 1890 Name: Allen Leung (leunga (at) reservoir (dot) com)
388 :     Date: 2006/02/22 15:07:00 EST
389 :     Tag: leunga-20060222-mdl-gen
390 :     Description:
391 :    
392 :     CM has changed. Updated the script for rebuilding the MLRISC generated
393 :     files.
394 :    
395 :     ----------------------------------------------------------------------
396 : mblume 1886 Name: Matthias Blume (http://www.tti-c.org/blume/)
397 :     Date: 2006/02/14 17:30:00 CST
398 :     Tag: blume-20060214-intel-mac
399 :     Description:
400 :    
401 :     Hooked code for Darwin-specific Intel ABI into main compiler.
402 :     (This is a temporary solution which relies on the fact that the
403 :     compiler itself does not use NLFFI. Eventually we need to divorce
404 :     intel mac from generic x86 unix code and make separate sets of binaries.)
405 :    
406 :     ----------------------------------------------------------------------
407 : jhr 1882 Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
408 : jhr 1885 Date: 2006/02/13
409 :     Tag: jhr-20060213-intel-mac
410 :     Description:
411 :     Changed MLRISC x86 CCalls for partial support of MacOS X ABI.
412 :    
413 :     ----------------------------------------------------------------------
414 :     Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
415 : jhr 1882 Date: 2006/02/06
416 :     Tag: jhr-20060206-intel-mac
417 :     Description:
418 :     Changes to support MacOS X on Intel hardware. The C-calls support in
419 :     MLRISC must also be updated to support the MacOS X ABI.
420 :    
421 :     ----------------------------------------------------------------------
422 : mblume 1863 Name: Matthias Blume (http://www.tti-c.org/blume/)
423 : mblume 1879 Date: 2005/12/16 00:15:00 CST
424 :     Tag: blume-20051216-cm
425 :     Description:
426 :    
427 :     Improved error reporting and handling in CM.
428 :    
429 :     ----------------------------------------------------------------------
430 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
431 : mblume 1878 Date: 2005/11/21 14:00:00 CST
432 :     Tag: blume-20051121-html-lib
433 :     Description:
434 :    
435 :     Enabling $/html-lib.cm in config/preloads did not work. This is fixed
436 :     now. (Since the anchor mapping for html-lib.cm is not yet in effect
437 : blume 2155 at the time when config/preloads is processed, the library has to
438 : mblume 1878 be referred to by another name. In 110.57 this would be
439 :     $SMLNJ-LIB/HTML/html-lib.cm. I arranged for $smlnj/smlnj-lib/html-lib.cm
440 :     to be valid as well.)
441 :    
442 :     Thanks to Todd Wilson (Fresno) for alerting me to this issue.
443 :    
444 :     ----------------------------------------------------------------------
445 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
446 : mblume 1874 Date: 2005/11/19 13:45:00 CST
447 :     Tag: Release_110_57
448 :     Description:
449 :    
450 :     Moved release tag for 110.57.
451 :    
452 :     ----------------------------------------------------------------------
453 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
454 : mblume 1873 Date: 2005/11/19 23:35:00 CST
455 :     Tag: blume-20051119-fish
456 :     Description:
457 :    
458 :     Fixed a problem in config/install.sh where it tries to "fish" the
459 :     name of the CM metadata directory from the wrong place (because
460 :     the physical location of basis.cm has changed). Also, corrected the
461 :     path anchor for $/html-lib.cm. (Thanks to M. Fluet for pointing out
462 :     these problems.)
463 :    
464 :     ----------------------------------------------------------------------
465 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
466 : mblume 1867 Date: 2005/11/16 14:30:00 CST
467 : mblume 1874 Tag: blume-orig-110_57
468 : mblume 1867 Description:
469 :    
470 :     New working version (110.57). NEW BOOTFILES!
471 :    
472 : mblume 1874 UPDATE: The release tag has been moved to include a fix for a brown-paper-
473 :     bag glitch (see above).
474 :    
475 : mblume 1867 ----------------------------------------------------------------------
476 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
477 : mblume 1866 Date: 2005/11/16 10:25:00 CST
478 :     Tag: blume-22051116-btrace
479 :     Description:
480 :    
481 :     Fixed problem with bogus exception message when using back-trace
482 :     facility.
483 :    
484 :     ----------------------------------------------------------------------
485 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
486 : mblume 1865 Date: 2005/11/15 22:35:00 CST
487 :     Tag: blume-22051115-array2copy
488 :     Description:
489 :    
490 :     Added simple implementation of Array2.copy. (Warning: mostly untested.)
491 :    
492 :     ----------------------------------------------------------------------
493 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
494 : mblume 1864 Date: 2005/11/15 21:35:00 CST
495 :     Tag: blume-22051115-versiontool
496 :     Description:
497 :    
498 :     Reversed change to src/system/smlnj/internals/versiontool.cm. This
499 :     file gets loaded as a tool -- by the equivalent of CM.make during the
500 :     run of CMB.make. Thus, CMB's path configuration is meaningless for
501 :     it. Instead, it has the status of "user code", so it should use
502 :     $/basis.cm to refer to the Basis library. (At least that's true for
503 :     the purpose of bootstrapping the previous change. In the future it
504 :     might make sense to have versiontool.cm refer to
505 :     $smlnj/basis/basis.cm, i.e., the version of the Basis that the
506 :     compiler itself uses.)
507 :    
508 :     Also patched src/system/testml to have it activate those extra anchor
509 :     bindings in config/extrapathconfig.
510 :    
511 :     ----------------------------------------------------------------------
512 :     Name: Matthias Blume (http://www.tti-c.org/blume/)
513 : mblume 1863 Date: 2005/11/15 16:35:00 CST
514 :     Tag: blume-22051115-libraries
515 :     Description:
516 :    
517 :     This change affects the way the following libraries are tied into
518 :     the system:
519 :    
520 :     $/basis.cm
521 :     $/smlnj-lib.cm
522 :     $/pp-lib.cm
523 :     $/controls-lib.cm
524 :     $/html-lib.cm
525 :     $/ml-yacc-lib.cm
526 :    
527 :     These libraries are now internally (as seen from the source code
528 :     of the implementation itself) known by the following names:
529 :    
530 :     $smlnj/basis/basis.cm
531 :     $smlnj/smlnj-lib/smlnj-lib.cm
532 :     $smlnj/smlnj-lib/pp-lib.cm
533 :     $smlnj/smlnj-lib/controls-lib.cm
534 :     $smlnj/smlnj-lib/html-lib.cm
535 :     $smlnj/ml-yacc/ml-yacc-lib.cm
536 :    
537 :     This makes it possible to work with code that requires different
538 :     versions of these libraries, and which refers to these libraries using
539 :     their "default" names (i.e., the first set of names above). In other
540 :     words, one can un-define or re-define those default names without
541 :     compromising the proper functioning of the compiler itself.
542 :    
543 :     A similar procedure had already been performed for several of the
544 :     MLRISC libraries that are linked into the compiler. I did some
545 :     cleanup on this code.
546 :    
547 :     A new file in the config directory (named extrapathconfig) is
548 :     responsible for setting up path anchors that the compiler itself does
549 :     not need, but that are typically required by user code.
550 :    
551 :     ----------------------------------------------------------------------
552 : mblume 1862 Name: Matthias Blume
553 :     Date: 2005/11/07 17:05:00 CST
554 :     Tag: blume-20051107-slice
555 :     Description:
556 :    
557 :     Fixed erroneous out-of-bounds test in the "update" function of
558 :     various *ArraySlice modules. (Thanks to Vesa A. Norrman for pointing
559 :     out the problem.)
560 :    
561 :     Pushed some Basis changes through ML-Lex, CML, and eXene.
562 :    
563 :     ----------------------------------------------------------------------
564 : jhr 1858 Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
565 : jhr 1861 Date: 2005/11/07
566 :     Description:
567 :     Fixed a Basis incompatibility: The depreciated function Substring.all
568 :     was removed (use Substring.full instead).
569 :    
570 :     ----------------------------------------------------------------------
571 :     Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
572 : jhr 1858 Date: 2005/11/05
573 :     Description:
574 :     Tweaked interval set API in SML/NJ library; see the CHANGES file for details.
575 :    
576 :     ----------------------------------------------------------------------
577 : mblume 1856 Name: Matthias Blume (http://www.tti-c.org/blume/)
578 :     Date: 2005/11/02 16:25:00 CST
579 :     Tag: blume-20051102-hex
580 :     Description:
581 :    
582 :     * Runtime system bootstrap code now accepts hex digits in BOOTLIST
583 :     in either upper case or lower case format.
584 :     * Pushed changes to names of Pack<N>{Big,Little} structures through
585 :     CML and eXene.
586 :    
587 :     ----------------------------------------------------------------------
588 : jhr 1852 Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
589 : jhr 1855 Date: 2005/11/02
590 :     Description:
591 :     Fixed a Basis incompatibility: Pack<N>{Big,Little} structures should
592 :     be called PackWord<N>{Big,Little}.
593 :    
594 :     ----------------------------------------------------------------------
595 :     Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
596 : jhr 1852 Date: 2005/10/28
597 :     Description:
598 :     Fixed a minor Basis incompatibility: hex digits should be upper case.
599 :    
600 :     ----------------------------------------------------------------------
601 : mblume 1846 Name: Matthias Blume (blume (at) tti - c (dot) org)
602 :     Date: 2005/10/25 16:00:00 CDT
603 :     Tag: Release_110_56
604 :     Description:
605 :    
606 :     New working version (110.56). NEW BOOTFILES!
607 :    
608 :     ----------------------------------------------------------------------
609 : jhr 1838 Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
610 : jhr 1840 Date: 2005/10/25
611 :     Description:
612 :     Added interval sets to utility library (signatures INTERVAL_DOMAIN
613 :     and INTERVAL_SET, and functor IntervalSetFn).
614 :    
615 :     ----------------------------------------------------------------------
616 :     Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
617 : jhr 1838 Date: 2005/10/14
618 :     Description:
619 :     Add Zhong Shao's fix for datatype equality functions.
620 :    
621 :     ----------------------------------------------------------------------
622 : allenleung 1836 Name: Allen Leung (leunga (at) reservoir (dot) com)
623 : allenleung 1837 Date: 2005/10/14 12:30:00 EST
624 :     Tag: leunga-20051014-x86-spill
625 :     Description:
626 :    
627 :     Bug fix a bug found by Carl Hauser.
628 :     There was a typo in the reload code for FCMP in x86SpillInstr.sml.
629 :    
630 :     ----------------------------------------------------------------------
631 :     Name: Allen Leung (leunga (at) reservoir (dot) com)
632 : allenleung 1836 Date: 2005/10/14 11:15:00 EST
633 :     Tag: leunga-20051014-x86-asm
634 :     Description:
635 :    
636 :     Removed some debugging code in file x86Asm.sml.
637 :     The function emit_operand was printing out debugging output.
638 :    
639 :     ----------------------------------------------------------------------
640 : jhr 1813 Name: John Reppy
641 : jhr 1831 Date: 2005/07/27
642 :     Tag:
643 :     Description:
644 :    
645 : jhr 1832 Fixed ml-lex to recognize "\r" as representing carriage return.
646 :    
647 :     ----------------------------------------------------------------------
648 :     Name: John Reppy
649 :     Date: 2005/07/27
650 :     Tag:
651 :     Description:
652 :    
653 : jhr 1831 Fixed ml-yacc to work on files with non-native end-of-line
654 :     encodings (e.g., Windows text file on a Unix system).
655 :    
656 :     ----------------------------------------------------------------------
657 :     Name: John Reppy
658 : jhr 1824 Date: 2005/07/20
659 :     Tag:
660 :     Description:
661 :    
662 :     Added changes from Dominic Evans (oldmanuk (at) gmail (dot) com)
663 :     to support HPUX 11.
664 :    
665 :     ----------------------------------------------------------------------
666 :     Name: John Reppy
667 : jhr 1813 Date: 2005/07/06
668 :     Tag:
669 :     Description:
670 :    
671 :     Changes to the SML/NJ library. See smlnj-lib/CHANGES for details.
672 :    
673 :     ----------------------------------------------------------------------
674 : mblume 1808 Name: Matthias Blume (blume (at) tti - c (dot) org)
675 :     Date: 2005/07/06 00:45:00 CDT
676 :     Tag: blume-20050706-slice-copy
677 :     Description:
678 :    
679 :     Fixed reversed logic for deciding whether to "copy up" or "copy down"
680 :     in *-array-slice.sml.
681 :    
682 :     ----------------------------------------------------------------------
683 : allenleung 1806 Name: Allen Leung (leunga (at) reservoir (dot) com)
684 : allenleung 1807 Date: 2005/05/31 17:00:00 EST
685 :     Tag: leunga-20050531-cygwin-fault-2
686 :     Description:
687 :    
688 :     A typo in the cygwin code fixed.
689 :    
690 :     ----------------------------------------------------------------------
691 :     Name: Allen Leung (leunga (at) reservoir (dot) com)
692 : allenleung 1806 Date: 2005/05/31 16:47:00 EST
693 :     Tag: leunga-20050531-cygwin-fault
694 :     Description:
695 :    
696 : blume 2155 Updated Cygwin's fault/signal handling to match the Windows version.
697 : allenleung 1806 Updated the export list.
698 :    
699 :     ----------------------------------------------------------------------
700 : mblume 1791 Name: Matthias Blume (blume (at) tti - c (dot) org)
701 : mblume 1795 Date: 2005/05/18 13:35:00 CDT
702 :     Tag: Release_110_54
703 :     Description:
704 :    
705 :     New working version (110.54). NEW BOOTFILES!
706 :    
707 :     ----------------------------------------------------------------------
708 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
709 : mblume 1794 Date: 2005/05/18 11:58:00 CDT
710 :     Tag: blume-20050518-installer
711 :     Description:
712 :    
713 :     Added support scripts for Mac OS X PackageMaker and modified
714 :     config/install.sh so that it supports re-dumping a heap image
715 :     after customization.
716 :    
717 :     ----------------------------------------------------------------------
718 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
719 : mblume 1793 Date: 2005/05/18 10:55:00 CDT
720 :     Tag: blume-20050518-realdiv-noovld
721 :     Description:
722 :    
723 :     Un-overloaded / to work around bug in overloading resolution code.
724 :    
725 :     ----------------------------------------------------------------------
726 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
727 : mblume 1792 Date: 2005/05/16 23:50:00 CDT
728 :     Tag: blume-20050516-redump-heap
729 :     Description:
730 :    
731 :     Added mechanism for re-creating a heap file for the interactive system
732 :     after configuration variables have been changed.
733 :    
734 :     CM.redump_heap : string -> unit
735 :    
736 :     This is much like SMLofNJ.exportML, but starting from the resulting
737 : blume 2155 heap does not return to the caller of CM.redump_heap but
738 : mblume 1792 restarts the interactive system from scratch. The original call of
739 :     CM.redump_heap does not return but ends the interactive session.
740 :     Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
741 :    
742 :     Internally, redump_heap winds the dynamic execution context back to
743 :     the point where the original heap image was created and re-executes
744 :     the heap image generation code in the boot code.
745 :    
746 :     ----------------------------------------------------------------------
747 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
748 : mblume 1791 Date: 2005/05/09 21:55:00 CDT
749 :     Tag: blume-20050509-word64
750 :     Description:
751 :    
752 :     Added a hack to the existing hack known as Word64 to make fromString
753 :     behave correctly. I am still not sure whether Word64.scan will work
754 :     as specified with respect to the interaction of radix and prefix.
755 :    
756 :     ----------------------------------------------------------------------
757 : allenleung 1785 Name: Allen Leung (leunga (at) reservoir (dot) com)
758 :     Date: 2005/05/04 11:50:00 EST
759 :     Tag: leunga-20050504-checkgc
760 :     Description:
761 :    
762 : blume 2155 Added a gc protocol checking phase. This phase is enabled with
763 : allenleung 1785 the flag "check-gc". "debug-check-gc" turns on the verbose mode.
764 :    
765 :     ----------------------------------------------------------------------
766 : mblume 1741 Name: Matthias Blume (blume (at) tti - c (dot) org)
767 : mblume 1784 Date: 2005/05/04 10:45:00 CDT
768 :     Tag: blume-20050504-intinf
769 :     Description:
770 :    
771 :     Fixed a bug in the implementation of div and mod for IntInf.
772 :     Thanks to Neophytos Michael for reporting the problem.
773 :    
774 :     ----------------------------------------------------------------------
775 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
776 : mblume 1783 Date: 2005/05/04 10:35:00 CDT
777 :     Tag: blume-20050504-join
778 :     Description:
779 :    
780 :     Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
781 :    
782 :     ----------------------------------------------------------------------
783 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
784 : mblume 1770 Date: 2005/02/28 23:40:00 CST
785 :     Tag: blume-20050228-mVar
786 :     Description:
787 :    
788 :     Fixed serious bug (brown paper bag variety) in new implementation of
789 :     structure Atom in CML. (I had accidentally used a mailbox instead of
790 :     an mvar, leaving the door open for races.)
791 :    
792 :     ----------------------------------------------------------------------
793 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
794 : mblume 1761 Date: 2005/02/25 15:00:00 CST
795 :     Tag: Release_110_53
796 :     Description:
797 :    
798 :     New working version (110.53). NEW BOOTFILES!
799 :    
800 :     ----------------------------------------------------------------------
801 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
802 : mblume 1760 Date: 2005/02/25 14:50:00 CST
803 :     Tag: blume-20050225-susp
804 :     Description:
805 :    
806 :     Brought back SMLofNJ.Susp. The underlying suspension type is the one
807 :     implemented in Core -- which means that it is the same as the one used
808 :     by the lazy extension.
809 :    
810 :     ----------------------------------------------------------------------
811 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
812 : mblume 1759 Date: 2005/02/24 16:50:00 CST
813 :     Tag: blume-20050224-cml-atom
814 :     Description:
815 :    
816 :     Simpler and at the same time more general implementation of structure
817 :     Atom in CML.
818 :    
819 :     ----------------------------------------------------------------------
820 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
821 : mblume 1758 Date: 2005/02/15 17:35:00 CST
822 :     Tag: blume-20050215-tools
823 :     Description:
824 :    
825 :     Created new "tools" directory under "src" and moved "TraceDebugProf"
826 :     there.
827 :    
828 :     ----------------------------------------------------------------------
829 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
830 : mblume 1755 Date: 2005/02/10 17:55:00 CST
831 :     Tag: blume-20050210-longlong
832 :     Description:
833 :    
834 :     Implemented "long long" arguments and results for NLFFI. (Only the
835 :     PPC/MacOS implementation is complete, the other backends still need to
836 :     be updated.)
837 :    
838 :     ----------------------------------------------------------------------
839 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
840 : mblume 1751 Date: 2005/01/24 17:40:00 CST
841 :     Tag: blume-20050124-mlyacc
842 :     Description:
843 :    
844 :     Minor cleanup in ML-Yacc rule printing mechanism. This should fix a
845 :     problem with certain "as" patterns which previously got rendered
846 :     using incorrect syntax.
847 :    
848 :     ----------------------------------------------------------------------
849 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
850 : mblume 1750 Date: 2005/01/18 12:00:00 CST
851 :     Tag: blume-20050118-profile
852 :     Description:
853 :    
854 :     Made time profiling code (interrupt handler) in runtime system aware
855 :     of new array representation.
856 :    
857 :     ----------------------------------------------------------------------
858 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
859 : mblume 1747 Date: 2005/01/14 18:00:00 CST
860 :     Tag: blume-20050114-heap2exec
861 :     Description:
862 :    
863 :     Implemented new (but still experimental) heap2exec facility. This is
864 :     tested under Mac OS X and should work under Linux (will test shortly).
865 :     It will probably also work on the Sparc (will test some time later).
866 :    
867 :     - removed old "HACKED_STANDALONE" hack from runtime
868 :    
869 :     To be able to test this, uncomment the request for "heap2asm" in
870 :     config/targets prior to installation. (Notice that this is different
871 :     from "heap2exec" mentioned below. Not a typo.)
872 :    
873 :     To perform an actual test, run the command
874 :    
875 :     $ bin/heap2exec heapfile execfile
876 :    
877 :     (You can put heap2exec on your shell's path.)
878 :    
879 :     For example, run
880 :    
881 :     $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
882 :    
883 :     This will create a standalone executable called "mly" which you can
884 :     then invoke directly as a command.
885 :    
886 :     ----------------------------------------------------------------------
887 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
888 : mblume 1746 Date: 2005/01/07 11:44:00 CST
889 :     Tag: blume-20050107-mlstring
890 :     Description:
891 :    
892 :     fixed off-by-one error in ML_STRING macro (globals.c)
893 :    
894 :     ----------------------------------------------------------------------
895 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
896 : mblume 1744 Date: 2004/12/23 18:00:00 CST
897 :     Tag: blume-20041223-santa
898 :     Description:
899 :    
900 :     Made ml-build script "smarter" (but only very little).
901 :    
902 :     ----------------------------------------------------------------------
903 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
904 : mblume 1742 Date: 2004/12/21 15:05:00 CST
905 :     Tag: blume-20041221-longlong
906 :     Description:
907 :    
908 :     * Implemented access to signed and unsigned long long data in NLFFI.
909 :     (The parameter-passing part of the picture has not complete. But
910 :     data structure access seems to work.)
911 :    
912 :     * Fixed CM's incorrect assumption that the PPC is little-endian.
913 :     (On the Mac, it is big-endian. And that's currently our only
914 :     PPC platform.)
915 :    
916 :     ----------------------------------------------------------------------
917 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
918 : mblume 1741 Date: 2004/12/21 12:50:00 CST
919 :     Tag: blume-20041221-memory
920 :     Description:
921 :    
922 :     Some cleanup in the $c/memory.cm library: separated some concerns by
923 :     moving allocation code and memory access code each into their own
924 :     files.
925 :    
926 :     ----------------------------------------------------------------------
927 : allenleung 1740 Name: Allen Leung (leunga (at) reservoir (dot) com)
928 :     Date: 2004/12/17 16:12:00 EST
929 :     Tag: leunga-20041217-cygwin-smlnj-home
930 :     Description:
931 :    
932 : blume 2155 The Unix I/O library of SML/NJ on cygwin does not understand
933 :     Windows style pathname, so problems arise when SMLNJ_HOME is set
934 :     to a Windows style pathname. _run-sml now converts SMLNJ_HOME
935 :     to a POSIX pathname on cygwin.
936 : allenleung 1740
937 :     ----------------------------------------------------------------------
938 : mblume 1737 Name: Matthias Blume (blume (at) tti - c (dot) org)
939 : mblume 1736 Date: 2004/12/16 13:35:00 CST
940 :     Tag: Release_110_52
941 :     Description:
942 :    
943 :     Last-minute changes incorporated into 110.52. Release tag moved.
944 :    
945 :     The changes:
946 :     - HashString.hashString' -> HashString.hashSubstring
947 :     - bug fix in UnivariateStats
948 :    
949 :     ----------------------------------------------------------------------
950 : mblume 1702 Name: Matthias Blume (blume (at) tti - c (dot) org)
951 : mblume 1733 Date: 2004/12/15 23:40:00 CST
952 :     Tag: blume-20041215-hashSubstring
953 :     Description:
954 :    
955 :     - HashString.hashString' -> HashString.hashSubstring
956 :     - corresponding changes in atom.sml
957 :     - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
958 :     added some comments
959 :    
960 :     ----------------------------------------------------------------------
961 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
962 : mblume 1728 Date: 2004/12/15 15:30:00 CST
963 : mblume 1736 Tag: (Release_110_52)
964 : mblume 1728 Description:
965 :    
966 :     New working version (110.52). NEW BOOTFILES!
967 :    
968 :     ----------------------------------------------------------------------
969 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
970 : mblume 1727 Date: 2004/12/15 12:45:00 CST
971 :     Tag: blume-20041215-spaces
972 :     Description:
973 :    
974 :     More on the space problem (this time for Win32).
975 :    
976 :     ----------------------------------------------------------------------
977 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
978 : mblume 1725 Date: 2004/12/14 17:30:00 CST
979 :     Tag: blume-20041214-spaces
980 :     Description:
981 :    
982 :     Hacked some of the scripts (in particular: the installer) to cope with
983 :     spaces in filenames a bit better. But beware: the current "solution"
984 :     is likely still full of bugs and inherently incomplete. (We need to
985 :     do away with those shell scripts for a comprehensive solution.)
986 :    
987 :     ----------------------------------------------------------------------
988 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
989 : mblume 1724 Date: 2004/12/13 14:45:00 CST
990 :     Tag: blume-20041213-ml-makedepend
991 :     Description:
992 :    
993 :     Fixed bug in code for ml-makedepend.
994 :    
995 :     ----------------------------------------------------------------------
996 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
997 : mblume 1720 Date: 2004/12/09 16:30:00 CST
998 :     Tag: blume-20041209-statistics
999 :     Description:
1000 :    
1001 :     Added two simple but potentially useful statistics modules to SML/NJ Library.
1002 :     (See CHANGES file there.)
1003 :    
1004 :     ----------------------------------------------------------------------
1005 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1006 : mblume 1718 Date: 2004/12/01 16:50:00 CST
1007 :     Tag: blume-20041201-atom
1008 :     Description:
1009 :    
1010 :     smlnj-lib:
1011 :    
1012 :     Added function HashString.hashString' for substrings.
1013 :     Hand-inlined CharVector.foldl into HashString (for speed).
1014 :     Modified implementation of structure Atom to avoid extracting
1015 :     strings from substrings unless necessary.
1016 :     (Also see CHANGES file for smlnj-lib.)
1017 :    
1018 :     ----------------------------------------------------------------------
1019 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1020 : mblume 1717 Date: 2004/11/24 22:15:00 CST
1021 :     Tag: blume-20041124-cml
1022 :     Description:
1023 :    
1024 :     Made sure CML compiles when Position = Int64.
1025 :    
1026 :     ----------------------------------------------------------------------
1027 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1028 : mblume 1715 Date: 2004/11/24 14:30:00 CST
1029 :     Tag: blume-20041124-position
1030 :     Description:
1031 :    
1032 :     The compiler can now be compiled in a mode that makes structure Position
1033 :     equal to Int64. The default, however, is unchanged (Position = Int31)
1034 :     for the time being.
1035 :    
1036 :     To enable 64-bit positions, use the following procedure:
1037 :    
1038 :     1. Start sml
1039 :     2. Autoload $smlnj/cmb.cm (if not already autoloaded)
1040 :     3. Type
1041 :     #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
1042 :     4. Run CMB.make() as usual.
1043 :    
1044 :     This is barely tested. The only test so far was a little SML program
1045 :     counting the number of characters in an 8-gigabyte file by
1046 :     reading it character-by-character. That test was successful.
1047 :    
1048 :     In support of 64-bit positions, a number of new functions have been
1049 :     added to the runtime system.
1050 :    
1051 :     ----------------------------------------------------------------------
1052 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1053 : mblume 1716 Date: 2004/11/23 14:45:00 CST
1054 :     Tag: blume-20041123-useFile
1055 : mblume 1712 Description:
1056 :    
1057 :     Fixed a problem with unhelpful error messages related to problems with
1058 :     .cm- or .sml files that appear as part of the sml command line.
1059 :    
1060 :     ----------------------------------------------------------------------
1061 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1062 : mblume 1703 Date: 2004/11/18 15:40:00 CST
1063 :     Tag: Release_110_51
1064 :     Description:
1065 :    
1066 :     New working version (110.51). NEW BOOTFILES!
1067 :    
1068 :     ----------------------------------------------------------------------
1069 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1070 : mblume 1702 Date: 2004/11/18 15:35:00 CST
1071 :     Tag: <none>
1072 :     Description:
1073 :    
1074 :     Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
1075 :     Joemann).
1076 :    
1077 :     ----------------------------------------------------------------------
1078 : allenleung 1701 Name: Allen Leung (leunga (at) reservoir (dot) com)
1079 :     Date: 2004/11/17 16:05:21 EST 2004
1080 :     Tag: leunga-20041117-mlrisc-live-kill
1081 :     Description:
1082 :    
1083 : blume 2155 Added support for MLTree constructs LIVE and KILL
1084 : allenleung 1701 to all the architectures.
1085 :    
1086 :     ----------------------------------------------------------------------
1087 : mblume 1687 Name: Matthias Blume (blume (at) tti - c (dot) org)
1088 : mblume 1689 Date: 2004/11/13 00:20:00 CST
1089 :     Tag: blume-20041113-versiontool
1090 :     Description:
1091 :    
1092 :     - Stripped down the versiontool: It now only handles the version number.
1093 :     The date string is generated at bootstrap time (during makeml).
1094 :    
1095 :     - In a previous commit, fixed a minor issue with how polyequal is being
1096 :     translated. In particular, the code now "looks through" abstractions.
1097 :     This results in slightly fewer polyEqual warnings and hopefully slightly
1098 :     more efficient code. Important examples for where this matters are
1099 :     the new int64 and word64 types.
1100 :    
1101 :     ----------------------------------------------------------------------
1102 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1103 : mblume 1687 Date: 2004/11/12 00:30:00 CST
1104 :     Tag: blume-20041112-int64
1105 :     Description:
1106 :    
1107 :     Structure Int64 fully hooked in. (The implementation is not very
1108 :     efficient, though.)
1109 :    
1110 :     ----------------------------------------------------------------------
1111 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1112 : mblume 1686 Date: 2004/11/11 17:30:00 CST
1113 :     Tag: blume-20041111-more64
1114 :     Description:
1115 :    
1116 :     All the pieces of Word64 are now there, with the exception of the
1117 :     conversions from and to LargeWord. (Eventually these need to be identities,
1118 :     but for the time being they don't even make sense because LargeWord is
1119 :     32-bit wide.)
1120 :    
1121 :     Also started to add similar support for Int64, but major pieces of that
1122 :     are still missing.
1123 :    
1124 :     ----------------------------------------------------------------------
1125 : mblume 1668 Name: Matthias Blume (blume (at) tti - c (dot) org)
1126 : mblume 1684 Date: 2004/11/11 00:15:00 CST
1127 :     Tag: blume-20041111-word64
1128 :     Description:
1129 :    
1130 :     Structure Word64 is now (almost) complete, word literals and patterns
1131 :     seem to work. There are a few odd pieces missing. In particular,
1132 :     I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
1133 :     at the moment.
1134 :    
1135 :     Making Word64 official would mean that LargeWord becomes Word64. But
1136 :     this requires extreme care because most word-word conversions have to
1137 :     go through LargeWord, so making a mistake means loss of efficiency or
1138 :     worse. Eventually there will be a solution similar to (but actually
1139 :     simpler than) what I did with IntInf.
1140 :    
1141 :     ----------------------------------------------------------------------
1142 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1143 : mblume 1683 Date: 2004/11/10 18:12:00 CST
1144 :     Tag: blume-20041110-64bit
1145 :     Description:
1146 :    
1147 :     More 64-bit hacking (but still not even half-way there yet).
1148 :     Also, some assorted improvements to the handling of 8-bit words.
1149 :    
1150 :     ----------------------------------------------------------------------
1151 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1152 : mblume 1682 Date: 2004/11/09 17:50:00 CST
1153 : mblume 1683 Tag: <none>
1154 : mblume 1682 Description:
1155 :    
1156 :     Started adding some infrastructure for supporting 64-bit int- and
1157 :     word-types. (Still in its very early stages.)
1158 :    
1159 :     ----------------------------------------------------------------------
1160 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1161 : mblume 1672 Date: 2004/10/28 10:45:00 CDT
1162 :     Tag: Release_110_50
1163 :     Description:
1164 :    
1165 :     New working version (110.50). NEW BOOTFILES!
1166 :    
1167 :     =====================
1168 :    
1169 :     Also:
1170 :    
1171 :     - Changend config/srcarchiveurl from a file just
1172 :     containing the URL string into a file containing
1173 :     shell script code. The code has access to the $VERSION variable.
1174 :     - Made corresponding changes to config/install.sh and config/unpack.
1175 :     - Default contents of config/srcarchiveurl uses $VERSION and
1176 :     normally does not have to be edited to reflect a version change.
1177 :    
1178 :     (As a result, a version change can be done by just editing
1179 :     config/version, the rest is now automatic.)
1180 :    
1181 :     ----------------------------------------------------------------------
1182 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1183 : mblume 1669 Date: 2004/10/27 17:50:00 CDT
1184 :     Tag: blume-20041027-btrace-msg
1185 :     Description:
1186 :    
1187 :     BackTrace.monitor now also reports the source of the exception that
1188 :     triggered the trace.
1189 :    
1190 :     ----------------------------------------------------------------------
1191 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1192 : mblume 1668 Date: 2004/10/27 17:20:00 CDT
1193 :     Tag: blume-20041027-x86-c-calls
1194 :     Description:
1195 :    
1196 :     This is the HISTORY entry for two earlier commits, both concerning
1197 :     the x86 c-calls code in MLRISC:
1198 :    
1199 :     - added a missing LOAD in the code that deals with struct arguments
1200 :     - made sure the caller does not add the wrong number of bytes to the
1201 :     stack pointer after a call of a function returning a struct
1202 :     (the callee already pops the implicit argument which points to
1203 :     the space reserved for the result)
1204 :    
1205 :     ----------------------------------------------------------------------
1206 : allenleung 1662 Name: Allen Leung (leunga (at) reservoir (dot) com)
1207 :     Date: 2004/10/24 14:00:00 EST
1208 :     Tag: leunga-20041024-x86-gas-fucomip
1209 :     Description:
1210 :    
1211 :     John discovered a bug in the syntax of fucomip.
1212 :     The opcodes FU?COMIP? have been changed to
1213 :    
1214 :     fu?comip? %st(i), %st
1215 :    
1216 :     ----------------------------------------------------------------------
1217 : mblume 1622 Name: Matthias Blume (blume (at) tti - c (dot) org)
1218 : mblume 1656 Date: 2004/10/20 15:06:00 CDT
1219 :     Tag: blume-20041020-standalone-backtrace
1220 :     Description:
1221 :    
1222 :     Added a mechanism for getting back-trace information from standalone
1223 :     programs. Here is how it works:
1224 :    
1225 :     1. The part of the program from which you want to get backtrace
1226 :     information (usually the whole program) should be wrapped with
1227 :     BackTrace.monitor. This is a (unit->'a)->'a function, and your
1228 :     main program could be modified from something like
1229 :    
1230 :     fun main (pgm, args) = ...
1231 :    
1232 :     to
1233 :    
1234 :     fun main (pgm, args) = BackTrace.monitor (fn () => ...)
1235 :    
1236 :     2. To be able to access BackTrace.monitor, you have to add
1237 :     library $smlnj-tdp/plugins.cm to the .cm file that contains your
1238 :     main function.
1239 :    
1240 :     3. Remove all compiled code (i.e., all the .cm/ subdirectories that
1241 :     CM might have created in the past for your project).
1242 :    
1243 :     4. Build the system using this command line:
1244 :    
1245 :     ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
1246 :     myprog.cm MyProg.main myprog
1247 :    
1248 :     instead of the usual
1249 :    
1250 :     ml-build myprog.cm MyProg.main myprog
1251 :    
1252 :     I changed a library name:
1253 :    
1254 :     $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
1255 :    
1256 :     New libraries:
1257 :    
1258 :     $smlnj-tdp/back-trace.cm
1259 :     -- when loaded causes the back-trace plugin to be installed
1260 :     $smlnj-tdp/coverage.cm
1261 :     -- when loaded causes the coverage plugin to be installed
1262 :    
1263 :     ----------------------------------------------------------------------
1264 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1265 : mblume 1655 Date: 2004/10/18 16:45:00 CDT
1266 :     Tag: blume-20041018-groupowner
1267 :     Description:
1268 :    
1269 :     Added an "obsolete" warning for the "group owner" syntax to CM's parser.
1270 :    
1271 :     Eliminated group owner specs from .cm files throughout the source tree.
1272 :    
1273 :     ----------------------------------------------------------------------
1274 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1275 : mblume 1653 Date: 2004/10/15 15:45:00 CDT
1276 :     Tag: blume-20041015-coverage
1277 :     Description:
1278 :    
1279 :     * Test coverage tool added!
1280 :     * Further reorganization of tracing-, debugging-, and profiling support:
1281 :    
1282 :     - moved original BTImp -- now called BackTrace -- into a separate
1283 :     library called $/trace-debug-profile.cm
1284 :     - eliminated all mentions of BTrace from SMLofNJ.Internals
1285 :     - only the instrumentation mechanism is now left in the compiler proper
1286 :     - BackTrace module is a plugin which is NOT plugged in by default
1287 :     - Coverage module is another such plugin
1288 :    
1289 :     To get the benefits of any of these plugin modules, the code in
1290 :     question must be compiled with tdp instrumentation turned on. This
1291 :     can be done by setting SMLofNJ.Internals.TDP.mode to true.
1292 :     (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
1293 :    
1294 :     Plugins are selected at link time. (Pre-compiled instrumented code
1295 :     can be re-loaded with different plugins in effect.) When an
1296 :     instrumented module is linked, whatever plugins are at that time
1297 :     enabled will come into effect for that module.
1298 :    
1299 :     To enable the back-trace plugin, load library $/trace-debug-profile.cm
1300 :     and invoke BackTrace.install() (e.g., from the interactive prompt).
1301 :     To enable the coverage plugin, load the same library and invoke
1302 :     Coverage.install().
1303 :    
1304 :     Back-traces are generated automatically on uncaught exceptions and
1305 :     when the code in question explicitly invokes BackTrace.trigger().
1306 :    
1307 :     Coverage (and execution frequency-) information must be queried
1308 :     explicitly by calling Coverage.not_covered and Coverage.hot_spots.
1309 :    
1310 :     ----------------------------------------------------------------------
1311 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1312 : mblume 1652 Date: 2004/10/14 17:40:00 CDT
1313 :     Tag: blume-20041014-tdp-core
1314 :     Description:
1315 :    
1316 :     Snapshot of a significant overhaul of how the trace/debug/profile support
1317 :     is hooked into the system (specifically: Core and SMLofNJ.Internals).
1318 :    
1319 :     ----------------------------------------------------------------------
1320 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1321 : mblume 1650 Date: 2004/10/13 16:34:00 CDT
1322 :     Tag: blume-20041013-tdp
1323 :     Description:
1324 :    
1325 :     Some rationalization of names:
1326 :    
1327 :     structure BTrace -> structure TDPInstrument
1328 :     etc.
1329 :    
1330 :     This is is preparation of using the original back-trace
1331 :     instrumentation for other purposes. "TDP" stands for
1332 :     Trace/Debug/Profile.
1333 :    
1334 :     The control flag controlling whether instrumentation is on or off is now
1335 :     registered under a different name, so instead of running sml as
1336 :    
1337 :     sml -Cinstrument.btrace-mode=true
1338 :    
1339 :     one has to say
1340 :    
1341 :     sml -Ctdp.instrument=true
1342 :    
1343 :     ----------------------------------------------------------------------
1344 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1345 : mblume 1645 Date: 2004/10/11 16:37:00 CDT
1346 :     Tag: blume-20041011-regions
1347 :     Description:
1348 :    
1349 :     Made some minor modifications to elabcore.sml to have source regions
1350 :     be propagated more tightly -- resulting in better (i.e., smaller)
1351 :     regions being reported in error- and debug messages.
1352 :    
1353 :     ----------------------------------------------------------------------
1354 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1355 : mblume 1644 Date: 2004/10/08 22:50:00 CDT
1356 :     Tag: blume-20041008-cmkw
1357 :     Description:
1358 :    
1359 :     Fixed handling of keywords in .cm files: After seeing "is" the lexer
1360 :     treats subsequent occurrences of "group", "library", "source", "is",
1361 :     "*", and "-" as ordinary identifiers rather than keywords.
1362 :    
1363 :     Most seriously, this fixes a problem with CM's "shell" tool. The tool
1364 :     is supposed to accept a tool argument called "source", but this did
1365 :     not work because of the clash with the keyword.
1366 :    
1367 :     ----------------------------------------------------------------------
1368 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1369 : mblume 1642 Date: 2004/10/07 16:00:00 CDT
1370 :     Tag: blume-20041007-cleanup
1371 :     Description:
1372 :    
1373 :     Assorted cleanup work:
1374 :    
1375 :     - got rid of intstrmap in favor of using the library's
1376 :     hash table implementation
1377 :     - threw out most of the pathnames stuff, as it was not used anyway
1378 :     - simplified tokentable implementation
1379 :     - fixed some minor spelling errors
1380 :    
1381 :     ----------------------------------------------------------------------
1382 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1383 : mblume 1641 Date: 2004/10/06 15:15:15 CDT
1384 :     Tag: blume-20041006-handler
1385 :     Description:
1386 :    
1387 :     Cleaned up the absyn to reflect the invariant that HANDLE always
1388 :     carries a FNexp as part of the type definition. This eliminates some
1389 :     superfluous sanity checks at runtime down the road.
1390 :    
1391 :     Some minor cleanup of the btrace code.
1392 :    
1393 :     ----------------------------------------------------------------------
1394 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1395 : mblume 1640 Date: 2004/10/01 10:20:30 CDT
1396 :     Tag: blume-20041001-slave
1397 :     Description:
1398 :    
1399 :     Added hack to make slave mode work in the presence of the version
1400 :     tool. (Still, since the master does two passes over the code for
1401 :     CMB.make, the release number gets bumped twice when slaves are
1402 :     attached. I don't know if this is worth fixing...)
1403 :    
1404 :     ----------------------------------------------------------------------
1405 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1406 : mblume 1639 Date: 2004/09/30 10:55:00 CDT
1407 :     Tag: blume-20040930-version
1408 :     Description:
1409 :    
1410 :     * Moved the "version" magic into its own little library under
1411 : mblume 1672 src/system/smlnj/internal. This avoids expensive reconstruction of
1412 :     a stable src/compiler/core.cm.
1413 : mblume 1639
1414 : mblume 1672 * At the same time, structure CompilerVersion is now known as
1415 :     structure SMLNJVersion.
1416 : mblume 1639
1417 :     * Arranged for the version tool to NOT kick in when rebuilding the system
1418 :     (makeml -rebuild, fixpt). Otherwise one would never reach a fixpoint.
1419 :     Also, loading the versiontool does not work when rebuilding the system
1420 :     because CM is not properly initialized at that time.
1421 :    
1422 :     ----------------------------------------------------------------------
1423 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1424 : mblume 1638 Date: 2004/09/29 14:00:00 CDT
1425 :     Tag: blume-20040929-autoversion
1426 :     Description:
1427 :    
1428 :     Implemented some CM magic to have
1429 :     file src/compiler/TopLevel/main/version.sml
1430 :     generated automagically.
1431 :     The version is taken from two files: config/version and config/release.
1432 :     The first is expected to contain a two-part version number such as 110.49.
1433 :     The second should contain a single number, but it may be missing.
1434 :    
1435 :     If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
1436 :     time the version tool is loaded (which is the first time you say CMB.make),
1437 :     then the tool will increment the value stored in config/release every
1438 :     time CMB.make is invoked.
1439 :    
1440 :     The binfile format is now insensitive to anything beyond the first
1441 :     two components of a version number, so bumping the release does not render
1442 :     binfiles incompatible. Auto-bumping can be used to keep track of versions
1443 :     during development without invalidating existing binfiles.
1444 :    
1445 :     In any case, every CMB.make updates the date information in version.sml.
1446 :     (This is the date that is printed in the banner.)
1447 :    
1448 :     ----------------------------------------------------------------------
1449 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1450 : mblume 1634 Date: 2004/09/28 10:53:00 CDT
1451 :     Tag: blume-20040928-controls
1452 :     Description:
1453 :    
1454 :     Some cleanup of the controls code.
1455 :    
1456 :     ----------------------------------------------------------------------
1457 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1458 : mblume 1633 Date: 2004/09/27 22:08:00 CDT
1459 :     Tag: blume-20040927-controls
1460 :     Description:
1461 :    
1462 :     Added two pieces of functionality to the Controls interface:
1463 :    
1464 :     1. val save'restore: 'a control -> unit -> unit
1465 :    
1466 :     grabs the current value of the control in stage 1 and restores it
1467 :     in stage 2.
1468 :    
1469 :     2. val set' : 'a control * 'a -> unit -> unit
1470 :    
1471 :     stores the given value into the control in stage 2 (i.e., delayed)
1472 :     but does all error checking in stage 1.
1473 :     (This is for string controls that need to do parse their argument --
1474 :     something that might fail. In some cases, notably in CM, one
1475 :     already knows the intended argument but wants to delay the actual
1476 :     assignment until a time when error recovery would be more difficult.)
1477 :    
1478 :     Changed the handling of controls in tool arguments to classes "sml" and
1479 :     "lazysml":
1480 :     - use Controls.save'restore as a more robust way of restoring the
1481 :     old value (in particular: without having to re-parse the string)
1482 :     - use controls to handle the "overload" keyword in the init group
1483 :     (I believe this change actually fixes a long-standing obscure bug.)
1484 :    
1485 :     ----------------------------------------------------------------------
1486 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1487 : mblume 1632 Date: 2004/09/27 17:00:00 CDT
1488 :     Tag: blume-20040927-lazysml
1489 :     Description:
1490 :    
1491 :     Added a new tool class called "lazysml" to CM's tool chest. The only
1492 :     difference to "sml" is that compilation is done with Control.lazysml
1493 :     set to true. A source of class "lazysml" is automatically recognized
1494 :     by a file name suffix of ".lml".
1495 :    
1496 :     In addition to the above feature, the original class "sml" now also
1497 :     supports a tool argument "lazy" which has the same effect. As a
1498 :     result, the following three lines are equivalent:
1499 :    
1500 :     foo.sml : lazysml
1501 :     foo.sml : sml (lazy)
1502 :     foo.sml (lazy)
1503 :    
1504 :     The setting goes into effect both during parsing and during
1505 :     compilation. The original setting is restored right after parsing and
1506 :     after compilation, respectively.
1507 :    
1508 :     In addition to all the above, there is also a general mechanism to set
1509 :     ANY of the "controls" that are available at the command line via
1510 :     "-C..." on a per-sml-file basis. The same rules that apply for "lazy"
1511 :     apply as well. (In fact, "lazy" is implemented as a special case of
1512 :     the general mechanism.)
1513 :    
1514 :     The .cm file syntax uses a new keyword tool argument called "with".
1515 :     There are several ways of indicating the desired settings:
1516 :    
1517 :     foo.sml (with:parser.quotations=true)
1518 :     foo.sml (with:(name:parser.quotations value:true))
1519 :     foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
1520 :     foo.sml (with:(name1=value1 name2=value2 ...))
1521 :     foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
1522 :    
1523 :     etc.
1524 :    
1525 :     Another possible abbreviation is to leave out the =v or value:v part
1526 :     if the name refers to a boolean control (in which case the value is
1527 :     taken to be true). Thus, one could get lazy sml also by saying:
1528 :    
1529 :     foo.sml (with:parser.lazy-keyword=true)
1530 :     foo.sml (with:parser.lazy-keyword)
1531 :     foo.sml (with:(name:parser.lazy-keyword value:true))
1532 :     foo.sml (with:(name:parser.lazy-keyword))
1533 :    
1534 :     ----------------------------------------------------------------------
1535 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1536 : mblume 1631 Date: 2004/09/24 16:22:00 CDT
1537 :     Tag: blume-20040924-ppc-long-branch
1538 :     Description:
1539 :    
1540 :     Turned message about "emiting long form of branch" off by default.
1541 :     Added a control flag to turn it back on when desired.
1542 :    
1543 :     ----------------------------------------------------------------------
1544 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1545 : mblume 1630 Date: 2004/09/24 16:05:00 CDT
1546 :     Tag: blume-20040924-rounding
1547 :     Description:
1548 :    
1549 :     Applied patch for setting rounding modes under Mac OS X. Thanks to
1550 :     Melissa O'Neill for providing the code!
1551 :    
1552 :     ----------------------------------------------------------------------
1553 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1554 : mblume 1629 Date: 2004/09/23 17:30:00 CDT
1555 :     Tag: blume-20040923-envvars
1556 :     Description:
1557 :    
1558 :     1. Changed definition of type ControlRegistry.registry_tree to
1559 :     include control_info (i.e., the name of the controlling environment
1560 :     variable).
1561 :    
1562 :     2. Added command-line flags -e and -E to print the names of environment
1563 :     variables that can be used to control internal settings.
1564 :     (This uses the new API mentioned in 1.)
1565 :    
1566 :     ----------------------------------------------------------------------
1567 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1568 : mblume 1622 Date: 2004/09/13 16:50:00 CDT
1569 :     Tag: Release_110_49
1570 :     Description:
1571 :    
1572 :     New working version (110.49). NEW BOOTFILES!
1573 :    
1574 :     ----------------------------------------------------------------------
1575 : mblume 1621 Name: Matthias Blume
1576 :     Date: 2004/09/13 16:20:00 CDT
1577 :     Tag: blume-20040913-config-mlrisc
1578 :     Description:
1579 :    
1580 :     Put target "mlrisc" back into the default list.
1581 :     (There is no harm in having it, and some users have expressed their
1582 :     wish to have "mlrisc" included by default.)
1583 :    
1584 :     ----------------------------------------------------------------------
1585 : jhr 1620 Name: John Reppy
1586 :     Date: 2004/09/13
1587 :     Tag: jhr-20040913-signals
1588 :     Description:
1589 :    
1590 :     Fixed the signal masking code to properly nest mask/unmask operations
1591 :     on a per-signal basis.
1592 :    
1593 :     ----------------------------------------------------------------------
1594 : mblume 1619 Name: Matthias Blume (blume (at) tti - c (dot) org)
1595 :     Date: 2004/09/08 13:20:00 CDT
1596 :     Tag: blume-20040908-heap-magic
1597 :     Description:
1598 :    
1599 :     Bumped the heap macig number to 0x09082004 to account for the changed
1600 :     layout of the ML frame under MacOS X.
1601 :    
1602 :     ----------------------------------------------------------------------
1603 : allenleung 1618 Name: Allen Leung (leunga (at) reservoir (dot) com)
1604 :     Date: 2004/09/03 11:26:00 EST
1605 :     Tag: leunga-20040903-cygwin-install
1606 :     Description:
1607 :    
1608 :     Added a patch to _arch-n-opsys to enable the Cygwin runtime.
1609 :     The Cygwin runtime is turned on by setting the environment
1610 :     variable SMLNJ_CYGWIN_RUNTIME to 1.
1611 :    
1612 :     ----------------------------------------------------------------------
1613 : mblume 1615 Name: Matthias Blume (blume (at) tti - c (dot) org)
1614 : mblume 1617 Date: 2004/08/31 17:15:00 CDT
1615 :     Tag: blume-20040831-core
1616 :     Description:
1617 :    
1618 :     Added some exports to src/compiler/core.cm upon request by J. Joemann.
1619 :    
1620 :     ----------------------------------------------------------------------
1621 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1622 : mblume 1616 Date: 2004/08/30 17:55:00 CDT
1623 :     Tag: blume-20040830-installer
1624 :     Description:
1625 :    
1626 :     Upon request by Johannes Joemann:
1627 :    
1628 :     - improved ML code of installer to fall back to coping when renaming
1629 :     fails (i.e., when source and target are on different file systems);
1630 :     the code compiles but has yet to be tested in anger
1631 :     - removed mlrisc from list of default targets (config/targets)
1632 :    
1633 :     ----------------------------------------------------------------------
1634 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1635 : mblume 1615 Date: 2004/08/27 17:20:00 CDT
1636 :     Tag: blume-20040827-ptreql
1637 :     Description:
1638 :    
1639 :     Added ptreql primop to structure InlineT (upon request from Larry
1640 :     Paulson).
1641 :    
1642 :     ----------------------------------------------------------------------
1643 : allenleung 1613 Name: Allen Leung (leunga (at) reservoir (dot) com)
1644 : allenleung 1614 Date: 2004/08/15 21:21:00 EST
1645 :     Tag: leunga-110_48-udgraph
1646 :     Description:
1647 :    
1648 :     Another bug fix from Carl Hauser:
1649 :    
1650 :     diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
1651 :     > 48c48
1652 :     > < | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
1653 :     > e::L)
1654 :     > ---
1655 :     > > | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
1656 :     > Without this, any deletion of an edge in an undirected graph does severe
1657 :     > violence to the graph.
1658 :    
1659 :     ----------------------------------------------------------------------
1660 :     Name: Allen Leung (leunga (at) reservoir (dot) com)
1661 : allenleung 1613 Date: 2004/08/10 23:35:00 EST
1662 :     Tag: leunga-110_48-ppc
1663 :     Description:
1664 :    
1665 :     The IBM/MacOS syntax switch on PPC was incorrectly swapped.
1666 :    
1667 :     ----------------------------------------------------------------------
1668 : mblume 1602 Name: Matthias Blume (blume (at) tti - c (dot) org)
1669 : mblume 1615 Date: 2004/08/10 12:00:00 CDT
1670 : mblume 1602 Tag: Release_110_48
1671 :     Description:
1672 :    
1673 :     New working version (110.48). NEW BOOTFILES!
1674 :    
1675 :     ----------------------------------------------------------------------
1676 : allenleung 1599 Name: Allen Leung (leunga (at) reservoir (dot) com)
1677 :     Date: 2004/08/09 12:21:00 EST
1678 :     Tag: leunga-110_47-dijsktra
1679 :     Description:
1680 :    
1681 :     Bug fix from Carl Hauser:
1682 :    
1683 : blume 2155 single_source_shortest_paths in dijkstra.sml was observed to get wrong
1684 : allenleung 1599 answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1685 :    
1686 : blume 2155 The problem is that following the expression A.update(dist,s,Num.zero)
1687 : allenleung 1599 it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1688 :    
1689 :     ----------------------------------------------------------------------
1690 : mblume 1595 Name: Matthias Blume (blume (at) tti - c (dot) org)
1691 :     Date: 2004/08/06 18:10:00 CDT
1692 :     Tag: blume-20040806-cmdline
1693 :     Description:
1694 :    
1695 :     Fiddled with handling of command-line options:
1696 :    
1697 :     * sml now quits after processing the command line
1698 :     if -H, -S, -h<n>, or -s<n> appears as the last
1699 :     command-line argument
1700 :     * a new option -q terminates the session when encountered on
1701 :     the command line; subsequent arguments will be ignored
1702 :     * bug fixes: short (erroneous) arguments are no longer ignored
1703 :     completely
1704 :    
1705 :     ----------------------------------------------------------------------
1706 : allenleung 1593 Name: Allen Leung (leunga (at) reservoir (dot) com)
1707 :     Date: 2004/08/04 18:17:00 EST
1708 :     Tag: leunga-110_47-ppc-ibm-asm
1709 :     Description:
1710 :    
1711 :     - Added minimal IBM assembly syntax support for PowerPC.
1712 :    
1713 : blume 2155 - Cygwin: manually changed the file cygwin.def. Some exported symbols have
1714 : allenleung 1593 been altered in the runtime. We need an automatic way to keep the file
1715 :     in sync.
1716 :    
1717 :     ----------------------------------------------------------------------
1718 : mblume 1548 Name: Matthias Blume (blume (at) tti - c (dot) org)
1719 : mblume 1581 Date: 2004/08/04 14:00:00 CDT
1720 :     Tag: Release_110_47
1721 :     Description:
1722 :    
1723 :     New working version (110.47). NEW BOOTFILES!
1724 :    
1725 :     ----------------------------------------------------------------------
1726 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1727 : mblume 1580 Date: 2004/08/03 14:25:00 CDT
1728 :     Tag: blume-20040803-callingconv
1729 :     Description:
1730 :    
1731 :     Added low-level support for choosing C calling conventions by
1732 :     twiddling the type of rawccall. (See
1733 :     src/compiler/Semant/types/cproto.sml for details.)
1734 :    
1735 :     ----------------------------------------------------------------------
1736 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1737 : mblume 1579 Date: 2004/08/02 15:55:00 CDT
1738 :     Tag: blume-20040802-backout
1739 :     Description:
1740 :    
1741 :     Backed out of change to win32-filesys.c. The earlier patch to
1742 :     get_file_time caused CM to produce files with the wrong time stamp.
1743 :    
1744 :     ----------------------------------------------------------------------
1745 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1746 : mblume 1573 Date: 2004/08/02 14:45:00 CDT
1747 :     Tag: blume-20040802-nlffi-win32
1748 :     Description:
1749 :    
1750 :     Added NLFFI support for Win32, adapted from a patch provided by David
1751 :     Hansel. This is currently completely untested. Also, the issue
1752 :     concerning stdcall vs. ccall is still unresolved.
1753 :    
1754 :     ----------------------------------------------------------------------
1755 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1756 : mblume 1572 Date: 2004/07/30 17:55:00 CDT
1757 :     Tag: blume-20040730-various
1758 :     Description:
1759 :    
1760 :     Gearing up towards 110.47...
1761 :    
1762 :     - various minor bugfixes to ml-nlffigen
1763 :     - a beginning of a manual for nlffi
1764 :    
1765 :     - eliminated 'export name=value' in config/install.sh as this does
1766 :     not work with certain versions of /bin/sh
1767 :     (Thanks to David King at Motorola for catching this.)
1768 :    
1769 :     - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1770 :     - added a test for tm==NULL to gmtime.c and localtime.c
1771 :     - applied patch for incorrect GetFileTime under win32
1772 :     - toSeconds -> toMilliseconds in Win32/win32-process.sml
1773 :    
1774 :     ----------------------------------------------------------------------
1775 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1776 : mblume 1557 Date: 2004/07/21 18:20:00 CDT
1777 :     Tag: blume-20040721-nlffigen
1778 :     Description:
1779 :    
1780 :     - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1781 :     for a typedef to an incomplete type, but leave out the "typ" member.
1782 :     (This is just for consistency.)
1783 :     - Started to produce what is supposed to become better (i.e., comprehensive)
1784 :     documentation of what ml-nlffigen does and produces.
1785 :    
1786 :     ----------------------------------------------------------------------
1787 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1788 : mblume 1548 Date: 2004/07/14 16:25:00 CDT
1789 :     Tag: blume-20040714-union
1790 :     Description:
1791 :    
1792 :     Added C_UNION to c-calls/c-types.sml and updated the machinery
1793 :     (ml-nlffigen, cproto.sml) that conveys C function interface
1794 :     information to the code generator.
1795 :    
1796 :     However, the actual architecture-specific implementation of function
1797 :     arguments and results that are C unions is still not implemented.
1798 :    
1799 :     ----------------------------------------------------------------------
1800 : allenleung 1539 Name: Allen Leung (leunga (at) reservoir (dot) com)
1801 : allenleung 1546 Date: 2004/07/14 14:38:00 EST
1802 :     Tag: leunga-110_46_1-ppc-lwzu
1803 :     Description:
1804 :    
1805 :     Added these instructions to the PowerPC architecture:
1806 :     LBZU(X), LHZU(X), LWZU(X),
1807 : blume 2155 STWU(X), STFDU, STFSU
1808 :    
1809 : allenleung 1546 etc...
1810 :    
1811 :     Note: I haven't added their instruction encoding into the description.
1812 :    
1813 :     ----------------------------------------------------------------------
1814 :     Name: Allen Leung (leunga (at) reservoir (dot) com)
1815 : allenleung 1539 Date: 2004/07/13 15:04:00 EST
1816 :     Tag: leunga-110_46_1-ppc-lwarx
1817 :     Description:
1818 :    
1819 :     Added the two instructions LWARX and STWCX to the PowerPC
1820 : blume 2155 instruction set.
1821 : allenleung 1539
1822 :     A (untested) rewrite of loop-structure.sml. The old version
1823 :     is completely broken.
1824 :    
1825 :     ----------------------------------------------------------------------
1826 : mblume 1479 Name: Matthias Blume (blume (at) tti - c (dot) org)
1827 : mblume 1537 Date: 2004/07/13 13:50:00 CDT
1828 :     Tag: blume-20040713-nlffi
1829 :     Description:
1830 :    
1831 :     - use paramAlloc to report c-calls with too many arguments
1832 :     (for PPC version where parameter area is pre-allocated)
1833 :     - added ccall_maxargspace to machspec (to implement the above)
1834 :     - made "make" commend in CM's "make" tool configurable
1835 :     - added option (default: on) for passing the name of the SML/NJ's "bin"
1836 :     directory to "make"; the call looks like this:
1837 :    
1838 :     make <options> SMLNJ_BINDIR=<dir> <target>
1839 :    
1840 :     This can be used by the Makefile to, e.g., pick the "right" version
1841 :     of ml-nlffigen.
1842 :     - minor code tweaks
1843 :    
1844 :     ----------------------------------------------------------------------
1845 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1846 : mblume 1532 Date: 2004/07/12 22:50:00 CDT
1847 :     Tag: blume-110_46_1-macosx-nlffi
1848 :     Description:
1849 :    
1850 :     NLFFI under Mac OS X now working (sort of). This is largely untested,
1851 :     though.
1852 :    
1853 :     Note: 1. You have to make a new, clean build of the runtime system.
1854 :     2. There are new BOOTFILES, you have to use them!
1855 :     (Doing the bootstrap process yourself would be *very* painful!
1856 :     If you absolutely have to do it, build the system under
1857 :     a different architecture and then cross-compile.)
1858 :    
1859 :     Version bumped to 110.46.1 to account for runtime data format changes.
1860 :    
1861 :     ----------------------------------------------------------------------
1862 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1863 : mblume 1515 Date: 2004/06/18 14:30:00 CDT
1864 :     Tag: blume-20040618-unix
1865 :     Description:
1866 :    
1867 :     Changed the implementation of structure Unix so that the same stream
1868 :     is returned every time one of the {text,bin}{In,Out}streamOf functions
1869 :     is invoked on the same proc. This is not what the spec currently
1870 :     says -- although IMO it arguably should. (See discussion below.)
1871 :    
1872 :     ----------------------------------------------------------------------
1873 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1874 : mblume 1507 Date: 2004/06/17 18:15:00 CDT
1875 :     Tag: Release_110_46
1876 :     Description:
1877 :    
1878 :     New working version (110.46). NEW BOOTFILES!
1879 :    
1880 :     ----------------------------------------------------------------------
1881 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1882 : mblume 1506 Date: 2004/06/17 17:20:00 CDT
1883 :     Tag: blume-20040617-timer-unix
1884 :     Description:
1885 :    
1886 :     Changed the interface of structures Timer and Unix to match the most
1887 :     recent Basis spec.
1888 :    
1889 :     In the case of Unix there still seems to be an open/weird issue:
1890 :    
1891 :     The {text,bin}{In,Out}streamOf functions are supposed to create
1892 :     fresh streams whenever they are called -- as opposed to have them
1893 :     return the same stream every time. This design is supposed to
1894 :     prevent space leaks caused by proc values hanging on to streams.
1895 :    
1896 :     The reap function, on the other hand, is supposed to close the
1897 :     streams. This cannot be done without having a handle on the
1898 :     stream in proc after all...
1899 :    
1900 :     I took the liberty to implement the following stopgap solution:
1901 :    
1902 :     The proc value hangs on to the most recently created stream(s).
1903 :     Reap closes those. If either or both of the two streams hadn't
1904 :     been created at all yet, then reap will close the corresponding
1905 :     file descriptors directly.
1906 :    
1907 :     PS: I don't understand the original space leak argument anymore. If
1908 :     a proc hangs on to the imperative stream, then I/O operations on those
1909 :     will advance the state of the cached stream and avoid the space leak.
1910 :    
1911 :     ----------------------------------------------------------------------
1912 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1913 : mblume 1500 Date: 2004/05/28 16:45:00 CDT
1914 :     Tag: blume-20040528-basis
1915 :     Description:
1916 :    
1917 :     Added signature PACK_REAL and exported functor PrimIO.
1918 :    
1919 :     ----------------------------------------------------------------------
1920 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1921 : mblume 1498 Date: 2004/05/25 16:00:00 CDT
1922 :     Tag: blume-20040525-group-owner
1923 :     Description:
1924 :    
1925 :     CM now ignores (but still accepts) the "owner" information in group
1926 :     descriptions. The owner of a group is its next enclosing
1927 :     library. Each group must have a unique owner. (There is a virtual
1928 : mblume 1505 "toplevel" library that owns groups which are not nested within a real
1929 : mblume 1498 library.) Previously, each group had to explicitly declare its owner,
1930 :     and CM would check that such a declaration is correct. The new scheme
1931 :     is to have CM check that for each group there is precisely one owning
1932 :     library.
1933 :    
1934 :     The advantage of the new scheme is that the programmer no longer needs
1935 :     to maintain the somewhat annoying owner information. The downside is
1936 :     that CM cannot enforce the ownership rule across multiple runs of
1937 :     CM.make. Fortunately, enclosing the same group in two different
1938 :     libraries A and B which are not part of the same program does not
1939 :     cause real problems.
1940 :    
1941 :     ----------------------------------------------------------------------
1942 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1943 : mblume 1496 Date: 2004/05/20 16:00:00 CDT
1944 :     Tag: blume-20040520-win32
1945 :     Description:
1946 :    
1947 :     Made the win32 version work again. (Strangely, a misplaced comma had
1948 :     slipped into win32-process.c which prevented the runtime from being
1949 :     compiled correctly.)
1950 :    
1951 :     Also, included a minor addition to ml-build.bat analogous to what was
1952 :     done in blume-20040519-ml-build.
1953 :    
1954 :     ----------------------------------------------------------------------
1955 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1956 : mblume 1494 Date: 2004/05/19 22:10:00 CDT
1957 :     Tag: blume-20040519-ml-build
1958 :     Description:
1959 :    
1960 :     Arranged for ml-build to clean up after itself a little bit better.
1961 :     The script generates a temporary SML source file and compiles it using
1962 :     CM, so CM generates metadata (GUID, SKEL, objectfile) for it. It now
1963 :     gets rid of those at the end, so they don't accumulate under .cm.
1964 :    
1965 :     This required a minor change to install.sh because the name of the
1966 :     metadata directory (default: .cm) is actually configurable at
1967 :     installation time.
1968 :    
1969 :     ----------------------------------------------------------------------
1970 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1971 : mblume 1491 Date: 2004/05/18 15:50:00 CDT
1972 :     Tag: blume-20040518-mkreader
1973 :     Description:
1974 :    
1975 :     Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1976 :     implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1977 :    
1978 :     ----------------------------------------------------------------------
1979 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
1980 : mblume 1489 Date: 2004/05/11 14:35:00 CDT
1981 :     Tag: blume-20040511-win32sock
1982 :     Description:
1983 :    
1984 :     Added previously missing support for many socket-related functions
1985 :     under win32. Thanks to David Hansel <hansel@reactive-systems.com>
1986 :     for the voluminous patch!
1987 :    
1988 :     (I have not tested this patch under win32 yet.)
1989 :    
1990 : mblume 1490 Here is David's e-mail:
1991 :    
1992 :     Hi,
1993 :    
1994 :     Attached to this email you find a diff against sml/nj 110.45
1995 :     that will enable socket support under Windows.
1996 :    
1997 :     To apply the patch (using unix or cygwin)
1998 :     1) gunzip runtime.diff.gz
1999 :     2) "cd" into "src/runtime" in the source tree of a fresh
2000 :     110.45 installation.
2001 :     3) patch -p 1 < [your/path/to]runtime.diff
2002 :    
2003 : blume 2155 The code compiles fine but has NOT yet been extensively tested.
2004 :     I only ran a few tests for basic socket client functionality
2005 :     (which worked fine). Especially the functions that use ioctl
2006 : mblume 1490 are not tested at all and might not work (see below).
2007 :    
2008 :     I implemented this since we want to move to a newer version of sml/nj
2009 :     but need socket support in order to use it. This is the first time I
2010 :     even had a look at the sml/nj source, so please review my changes
2011 :     before making this part of the distribution! Here are a few issues
2012 : blume 2155 that I think might be better for someone to solve who is more
2013 : mblume 1490 familiar with the sml/nj source (and socket programming):
2014 :    
2015 :     - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
2016 :     exception since I could not figure out what the windows equivalent
2017 :     of these functions is
2018 :    
2019 :     - In sockets-osdep.h there are a some #include statements that are
2020 :     only used in a few files that include sockets-osdep.h
2021 :    
2022 :     - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
2023 :     does not process its return value since I don't know how to
2024 :     report an error upwards.
2025 :    
2026 :     - It would probably be good to have a call to WSACleanup() when
2027 : blume 2155 the library is unloaded (if there is such a possibility).
2028 : mblume 1490 Otherwise I think Windows will take care of this automatically
2029 :     when the process finishes.
2030 :    
2031 :     - I used ioctlsocket() as a replacement for ioctl() but I have
2032 :     no idea if that is actually the proper replacement on Windows.
2033 :    
2034 :     - All these issues are marked in the code by "FIXME" comments.
2035 :    
2036 : blume 2155 We use sml/nj extensively in our products and are quite happy
2037 :     with it. I hope this contribution will help you.
2038 : mblume 1490
2039 :     Keep up the good work!
2040 :    
2041 :     David
2042 :    
2043 : mblume 1489 ----------------------------------------------------------------------
2044 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2045 : mblume 1488 Date: 2004/05/11 14:20:00 CDT
2046 :     Tag: blume-20040511-installml
2047 :     Description:
2048 :    
2049 :     Fixed two bugs in installml script. (Thanks to Vesa A. Norrman for
2050 :     the patch.)
2051 :    
2052 :     ----------------------------------------------------------------------
2053 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2054 : mblume 1487 Date: 2004/05/11 14:05:00 CDT
2055 :     Tag: blume-20040511-nlffi-netbsd
2056 :     Description:
2057 :    
2058 :     Added support for nlffi under netbsd. (Thanks to Vesa A. Norrman for
2059 :     the patch.)
2060 :    
2061 :     ----------------------------------------------------------------------
2062 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2063 : mblume 1486 Date: 2004/05/11 12:05:00 CDT
2064 :     Tag: blume-20040511-exports
2065 :     Description:
2066 :    
2067 :     As per request by Adam Chlipala <adam@hcoop.net>, extended various
2068 :     export lists in compiler-related .cm-files.
2069 :    
2070 :     ----------------------------------------------------------------------
2071 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2072 : mblume 1485 Date: 2004/05/11 11:35:00 CDT
2073 :     Tag: blume-20040511-allsource
2074 :     Description:
2075 :    
2076 :     The installer now honors the "src-smlnj" target again, although its meaning
2077 :     has changed from "all sources required for the compiler" to "all sources
2078 :     the installer knows about". In other words, if you enable "src-smlnj"
2079 :     in the "targets" file, then the installer will pull in sources for
2080 :     everything. (Notice that this refers to source code only. Compiled
2081 :     code is still only installed for modules that were requested explicitly
2082 :     or which are required for other modules that were requested explicitly.)
2083 :    
2084 :     ----------------------------------------------------------------------
2085 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2086 : mblume 1482 Date: 2004/04/23 17:40:00 CDT
2087 :     Tag: blume-20040423-ieee-scan
2088 :     Description:
2089 :    
2090 :     Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
2091 :     in the exponent calculation we get INF or ZERO (depending on the mantissa
2092 :     and the sign of the exponent).
2093 :    
2094 :     ----------------------------------------------------------------------
2095 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2096 : mblume 1480 Date: 2004/04/23 10:40:00 CDT
2097 :     Tag: blume-20040423-ml-build
2098 :     Description:
2099 :    
2100 :     The ml-build script now terminates with a non-0 status when something
2101 :     goes wrong.
2102 :    
2103 :     ----------------------------------------------------------------------
2104 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2105 : mblume 1479 Date: 2004/04/22 16:35:00 CDT
2106 :     Tag: blume-20040422-Option
2107 :     Description:
2108 :    
2109 :     Made exception Option to be the same as exception Option.Option
2110 :     (as it should be).
2111 :    
2112 :     ----------------------------------------------------------------------
2113 : allenleung 1468 Name: Allen Leung (leunga (at) reservoir (dot) com)
2114 :     Date: 2004/03/19 14:40:00 EST
2115 :     Tag: leunga-20040319-cygwin-nlffi
2116 :     Description:
2117 :    
2118 :     Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
2119 :     of SML/NJ. The problem is that
2120 :    
2121 :     lib = dlopen(NULL, ...)
2122 :     f = dlsym(lib, "malloc");
2123 :    
2124 : blume 2155 does not work on Windows unless we explicitly export symbols
2125 :     such as 'malloc' during linking. We fixed this by explicitly
2126 : allenleung 1468 exporting the required symbols with the magic gcc incantation:
2127 :    
2128 :     -Wl,--export-all cygwin.def
2129 :    
2130 :     where cygwin.def is a file containing all the symbols that
2131 : blume 2155 we wish to export.
2132 : allenleung 1468
2133 : blume 2155 I suspect this is a Windows problem and we'll have to
2134 :     do the same (somehow with windows compilers) when
2135 : allenleung 1468 we build the native win32 version with the system
2136 : blume 2155 calls LoadLibrary/GetProcAddress.
2137 : allenleung 1468
2138 :     ----------------------------------------------------------------------
2139 : mblume 1384 Name: Matthias Blume (blume (at) tti - c (dot) org)
2140 : mblume 1465 Date: 2004/03/04 16:35:00 CST
2141 :     Tag: blume-20040304-intinf-fmt
2142 :     Description:
2143 :    
2144 :     Fixed problem with IntInf.fmt (sign would show up on the right instead
2145 :     of on the left for BIN, OCT, and HEX).
2146 :    
2147 :     ----------------------------------------------------------------------
2148 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2149 : mblume 1464 Date: 2004/03/04 11:25:00 CST
2150 :     Tag: blume-20040304-symlinks
2151 :     Description:
2152 :    
2153 :     Fixed problem with installer script (unix only) where bin/ml-yacc and
2154 :     friends pointed (via symlinks) to absolute locations instead of just
2155 :     .run-sml. This was reported by Vesa A Norrman.
2156 :    
2157 :     ----------------------------------------------------------------------
2158 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2159 : mblume 1455 Date: 2004/02/13 14:50:00 CST
2160 :     Tag: Release_110_45
2161 :     Description:
2162 :    
2163 :     New working version (110.45). New bootfiles.
2164 :    
2165 :     ----------------------------------------------------------------------
2166 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2167 : mblume 1448 Date: 2004/01/26 15:15:15 CST
2168 :     Tag: blume-20040126-toplevel
2169 :     Description:
2170 :    
2171 :     Improved handling of exceptions at the interactive toplevel.
2172 :    
2173 :     ----------------------------------------------------------------------
2174 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2175 :     Date: 2004/01/26 11:25:00
2176 : mblume 1446 Tag: blume-20040126-app
2177 :     Description:
2178 :    
2179 :     Type of top-level "app" corrected.
2180 : mblume 1447 Added code for setting vp_limitPtrMask to Win32-specific runtime.
2181 : mblume 1446
2182 :     ----------------------------------------------------------------------
2183 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2184 : mblume 1427 Date: 2003/11/18 17:10 CST
2185 :     Tag: blume-20031118-basis-fiddle
2186 :     Description:
2187 :    
2188 :     - changed Timer interface to what might become the spec
2189 :     - POSIX_FLAGS -> BIT_FLAGS according to spec
2190 :     - some other minor discrepancies wrt. spec eliminated
2191 :    
2192 :     ----------------------------------------------------------------------
2193 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2194 : mblume 1414 Date: 2003/11/06 12:00:00 CST
2195 :     Tag: Release_110_44
2196 :     Description:
2197 :    
2198 :     New working version (110.44). New bootfiles.
2199 :    
2200 :     ----------------------------------------------------------------------
2201 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2202 : mblume 1411 Date: 2003/11/04 11:50:00 CST
2203 :     Tag: blume-20031104-move-libraries
2204 :     Description:
2205 :    
2206 :     Eliminated the "dont_move_libraries" directive in config/targets.
2207 :     (The mechanism was broken and could not be fixed easily. Moreover,
2208 :     there does not seem to be any reason not to move all libraries into
2209 :     lib during installation. I originally implemented this directive as a
2210 :     backward-compatibility feature when I first introduced the new CM.
2211 :     Now that things have been stable for a long time and going back to the
2212 :     old CM is not an option, there is no reason to keep it around.)
2213 :    
2214 :     ----------------------------------------------------------------------
2215 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2216 : mblume 1409 Date: 2003/11/03 16:00:00 CST
2217 :     Tag: blume-20031103-installdir
2218 :     Description:
2219 :    
2220 :     Made installer honor INSTALLDIR variable again. (Thanks to Chris
2221 :     Richards for pointing out the problem and providing the solution.)
2222 :    
2223 :     ----------------------------------------------------------------------
2224 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2225 : mblume 1400 Date: 2003/10/01 17:05:00 CDT
2226 :     Tag: blume-20031001-lal-mlrisc
2227 :     Description:
2228 :    
2229 :     MLRISC bug fix from Lal.
2230 :    
2231 :     ----------------------------------------------------------------------
2232 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2233 : mblume 1399 Date: 2003/09/30 16:10:00 CDT
2234 :     Tag: blume-20030930-primio-bat
2235 : mblume 1398 Description:
2236 :    
2237 :     1. Added openVector, nullRd, and nullWr to PRIM_IO.
2238 :     2. Improved .bat files (for Win32 port) to make things work under Win95.
2239 :     (thanks to Aaron S. Hawley for this one)
2240 :    
2241 :     ----------------------------------------------------------------------
2242 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2243 : mblume 1395 Date: 2003/09/26 16:05:00 CDT
2244 :     Tag: blume-20030926-wrappriv
2245 :     Description:
2246 :    
2247 :     Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
2248 :    
2249 :     ----------------------------------------------------------------------
2250 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2251 : mblume 1394 Date: 2003/09/26 15:00:00 CDT
2252 :     Tag: blume-20030926-110_43_3
2253 :     Description:
2254 :    
2255 :     - additional cleanup
2256 :     - version number bump, NEW BOOTFILES
2257 :    
2258 :     ----------------------------------------------------------------------
2259 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2260 : mblume 1393 Date: 2003/09/26 12:00:00 CDT
2261 :     Tag: blume-20030926-ppautoload
2262 :     Description:
2263 :    
2264 :     I modified the read-eval-print loop so that the autoloader gets
2265 :     invoked whenever the prettyprinter tries to look up a symbol that
2266 :     is not currently defined in the toplevel environment but which
2267 :     appears in CM's autoload registry. As a result, we see far fewer of
2268 :     those ?.Foo.Bar.xxx names in the prettyprinter's output.
2269 :    
2270 :     In addition to this I tried to clean up some pieces of the Basis
2271 :     implementation (e.g., Socket, Word8Array) in order to prevent other
2272 :     instances of these ?.Foo.Bar.xxx names from being printed.
2273 :    
2274 :     The mechanism that picks names for types still needs some work, though.
2275 :     (Right now it seems that if there is a type A.t which is defined to
2276 : mblume 1394 be B.u, but B is unavailable at toplevel, then A.t gets printed as
2277 :     "?.B.u" although the perhaps more sensible solution would be to use
2278 : mblume 1393 "A.t" in this case. In other words, the prettyprinter should follow
2279 :     a chain of DEFtycs not farther than there are corresponding toplevel
2280 :     names in the current environment.)
2281 :    
2282 :     ----------------------------------------------------------------------
2283 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2284 : mblume 1391 Date: 2003/09/24 16:31:00 CDT
2285 :     Tag: blume-20030924-installer
2286 :     Description:
2287 :    
2288 :     Another installer tweak: All the ML code for the installer is now
2289 :     compiled during CMB.make and put into a little library called
2290 :     $smlnj/installer.cm. The installation then simply invokes
2291 :    
2292 :     sml -m $smlnj/installer.cm
2293 :    
2294 :     and everything happens automagically.
2295 :    
2296 :     Win32: ML code senses value of environment variable SMLNJ_HOME.
2297 :     Unix: ML code senses values of environment variables ROOT, CONFIGDIR,
2298 :     and BINDIR.
2299 :    
2300 :     The new scheme guarantees that the ML code responsible for the installation
2301 :     is in sync with the APIs of the main system. Also, the installer is
2302 :     somewhat faster because the installer script is precompiled.
2303 :    
2304 :     ----------------------------------------------------------------------
2305 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2306 : mblume 1390 Date: 2003/09/24 15:35:00 CDT
2307 :     Tag: blume-20030924-synsock
2308 :     Description:
2309 :    
2310 :     Added a signature SYNCHRONOUS_SOCKET to basis.cm. This is like SOCKET
2311 :     but excludes all non-blocking operations. Defined SOCKET (in Basis)
2312 :     and CML_SOCKET in terms of SYNCHRONOUS_SOCKET. Removed superfluous
2313 :     implementations of non-blocking operations from CML's Socket
2314 :     structure.
2315 :    
2316 :     ----------------------------------------------------------------------
2317 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2318 : mblume 1389 Date: 2003/09/24 15:10:05 CDT
2319 :     Tag: blume-20030924-sockets
2320 :     Description:
2321 :    
2322 :     1. Fixed SOCKET API and implementation to match Basis spec.
2323 :     This required changing the internal representation of sockets to one
2324 :     that remembers (for each socket file descriptor) whether it is currently
2325 :     blocking or non-blocking. This state is maintained lazily (i.e., a system
2326 :     call is made only if the state actually needs to change).
2327 :    
2328 :     2. OS-specific details of sockets were moved into separate files, thus
2329 :     making it possible to unify the bulk of the socket implementations
2330 :     between Unix and Win32.
2331 :    
2332 :     3. CML's socket API changed accordingly.
2333 :     (Note that we need to remove non-blocking functions from this API
2334 :     since they are redundant in the case of CML!)
2335 :    
2336 :     4. CML's socket implementation now makes use of non-blocking functions
2337 :     provided by Basis, thus removing all OS-dependent code from this part
2338 :     of CML.
2339 :    
2340 :     5. Changed Real64.precision from 52 to 53. Minor cleanup in Real64 code.
2341 :    
2342 :     ----------------------------------------------------------------------
2343 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2344 : mblume 1387 Date: 2003/09/22 12:10:00 CDT
2345 : mblume 1388 Tag: blume-20030922-110_43_2
2346 : mblume 1387 Description:
2347 :    
2348 :     Made a new interim version and bootfiles for developer's bootstrapping
2349 :     convenience.
2350 :    
2351 :     110.43.2 -- NEW BOOTFILES
2352 :    
2353 :     ----------------------------------------------------------------------
2354 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2355 : mblume 1385 Date: 2003/09/19 15:55:00 CDT
2356 :     Tag: blume-20030919-cmdir
2357 :     Description:
2358 :    
2359 :     1. new-install.sh -> install.sh
2360 :     2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
2361 :     3. tweaked installer so that another name instead of .cm can be chosen
2362 :     at install time (by setting the CM_DIR_ARC environment variable
2363 :     during installation); once installation is complete, the name is
2364 :     fixed
2365 :    
2366 :     ----------------------------------------------------------------------
2367 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2368 : mblume 1384 Date: 2003/09/18 16:00:00 CDT
2369 :     Tag: blume-20030918-110_43_1
2370 :     Description:
2371 :    
2372 :     Made a new interim version and bootfiles for developer's bootstrapping
2373 :     convenience.
2374 :    
2375 :     110.43.1 -- NEW BOOTFILES
2376 :    
2377 :     ----------------------------------------------------------------------
2378 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2379 : mblume 1383 Date: 2003/09/18 15:20:00 CDT
2380 :     Tag: blume-20030918-misc
2381 :     Description:
2382 :    
2383 :     1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
2384 :     this seems to be controversial at the moment) and used those in
2385 :     Posix.ProcEnv.times.
2386 :    
2387 :     2. Added Time.{from,to}Nanoseconds to Time.
2388 :    
2389 :     3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
2390 :     For example, fromLargeInt never needs to look at more than 3 "big
2391 :     digits" to get its 53 bits of precision.
2392 :    
2393 :     ----------------------------------------------------------------------
2394 : mblume 1384 Name: Matthias Blume (blume (at) tti - c (dot) org)
2395 : mblume 1381 Date: 2003/09/17 16:30:00 CDT
2396 :     Tag: blume-20030917-real32-slices
2397 :     Description:
2398 :    
2399 :     Added an entry to the primitive environment
2400 :     (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
2401 :     added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
2402 :    
2403 :     Removed some of the "magic" constants in real64.sml and replaced them
2404 :     with code that generates these values from their corresponding
2405 :     integer counterparts.
2406 :    
2407 :     Made all(?) the slice-related changes to the Basis and made everything
2408 :     compile again...
2409 :    
2410 :     ----------------------------------------------------------------------
2411 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2412 : mblume 1375 Date: 2003/09/15 17:45:00 CDT
2413 :     Tag: blume-20030915-rbase
2414 :     Description:
2415 :    
2416 :     Fixed bug in Real.fromLargeInt.
2417 :    
2418 :     ----------------------------------------------------------------------
2419 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2420 : mblume 1369 Date: 2003/09/13 18:11:00 CDT
2421 :     Tag: blume-20030913-libinstall
2422 :     Description:
2423 :    
2424 :     Minor bugfix in config/libinstall (set anchor with path to
2425 :     standalone tool after installing it, otherwise libraries that
2426 :     need ml-lex or ml-yacc won't compile the first time the installer
2427 :     runs).
2428 :    
2429 :     ----------------------------------------------------------------------
2430 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2431 : mblume 1368 Date: 2003/09/12 11:45:00 CDT
2432 :     Tag: blume-20030912-various
2433 :     Description:
2434 :    
2435 :     - fixed bug in Real.toLargeInt
2436 :     - fixed bug in Posix.ProcEnv.times
2437 :     - changed inputLine functions to return an option
2438 :     - minor installer improvements / bugfixes
2439 :     - changed default @SMLalloc parameter for x86/celeron to 64k
2440 :    
2441 :     ----------------------------------------------------------------------
2442 :     Name: Matthias Blume (blume (at) tti - c (dot) org)
2443 : mblume 1356 Date: 2003/09/09 22:00:00 CDT
2444 :     Tag: Release_110_43
2445 :     Description:
2446 :    
2447 :     New working release 110.43. New bootfiles.
2448 :    
2449 :     ----------------------------------------------------------------------
2450 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2451 : mblume 1354 Date: 2003/09/09 19:20:00 CDT
2452 :     Tag: blume-20030909-installer
2453 :     Description:
2454 :    
2455 :     Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
2456 :     Modified config/install.bat to take advantage of it. Also modified
2457 :     config/install.sh (and called it config/new-install.sh) to take advantage
2458 :     of it on Unix systems. (The SML code is (supposed to be) platform-
2459 :     independent.)
2460 :    
2461 :     The installer can now install everything under Win32
2462 :     as well as under *nix as long as it compiles.
2463 :    
2464 :     Other changes:
2465 :    
2466 :     - made CML compile again under Win32
2467 :     - made eXene compile under Win32 (by providing a fake structure UnixSock
2468 :     and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
2469 :     - fixed a bug in nowhere: it assumed that type OS.Process.status is the
2470 :     same as type int; under Win32 it isn't
2471 :     - fixed some slice-related problems in the win32-specific parts of CML
2472 :     - added a functor argument "sameVol" to os-path-fn.sml in the Basis
2473 :     (under Win32, the volume name is case-insensitive, and the
2474 :     OS.Path code compares volume names for equality)
2475 :    
2476 :     ----------------------------------------------------------------------
2477 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2478 : mblume 1353 Date: 2003/09/08 11:55:00 CDT
2479 :     Tag: blume-20030908-fullpath
2480 :     Description:
2481 :    
2482 :     Made Win32 version of OS.FileSys.fullPath return current directory
2483 :     when given an empty string. This is what the spec says, and incidentally,
2484 :     CM depends on it. (CM otherwise goes into an infinite loop in certain
2485 :     cases when presented with the name of a non-existing .cm file.)
2486 :    
2487 :     ----------------------------------------------------------------------
2488 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2489 : mblume 1350 Date: 2003/09/04 16:30:00 CDT
2490 : mblume 1352 Tag: blume-20030905-slices-etc
2491 : mblume 1350 Description:
2492 :    
2493 :     1. Changed interface to vectors and arrays in Basis to match
2494 :     (draft) Basis spec.
2495 :     2. Added signatures and implementations of slices according to
2496 :     Basis spec.
2497 :     3. Edited source code throughout the system to make it compile again
2498 :     under 1. and 2. (In some cases code had to be added to have it
2499 :     match the new signatures.)
2500 :     4. MLRISC should be backward-compatible: the copies of the originals
2501 :     of files that needed to change under 3. were retained, the .cm files
2502 :     check the compiler version number and use old versions when
2503 :     appropriate.
2504 :     5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
2505 :     dirstream -> string option (in accordance with Basis spec).
2506 :     6. When generating code that counts lines, ml-lex used function
2507 :     CharVector.foldli, taking advantage of its old interface.
2508 :     This has been replaced with the corresponding code from
2509 :     CharVectorSlice. (html-lex must be re-lexed!)
2510 :     7. BitArray in smlnj-lib/Util has been extended/modified to match the
2511 :     new MONO_ARRAY signature. (Do we need BitArraySlice?)
2512 :     8. Removed temporary additions (fromInternal, toInternal) from the
2513 :     (now obsolete) IntInf in smlnj-lib/Util.
2514 :     9. Cleaned up structure Byte.
2515 :     10. Added localOffset, scan, and fromString to Date (according to spec).
2516 :     Cleaned/corrected implementation of Date.
2517 :     (Still need to check for correctness; implement better canonicalizeDate.)
2518 :     11. Added "scan" to signature IEEE_REAL.
2519 :     12. Some improvements to IntInf [in particular: efficiency-hack for
2520 :     mod and rem when second operand is 2 (for parity checks).]
2521 :     13. Changed representation of type Time.time, using a single IntInf.int
2522 :     value counting microseconds. This considerably simplified the
2523 :     implementation of structure Time. We now support negative time
2524 :     values; scan and fromString handle signs.
2525 :     14. Functor PrimIO now takes two additional arguments (VectorSlice and
2526 :     ArraySlice).
2527 :    
2528 :     ----------------------------------------------------------------------
2529 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2530 : mblume 1347 Date: 2003/08/28 17:00:00 CDT
2531 :     Tag: blume-20030828-intinf
2532 :     Description:
2533 :    
2534 :     This is a major update which comes with a version number bump
2535 :     (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
2536 :     and an implementation of IntInf in the Basis.
2537 :    
2538 :     There are a fairly large number of related changes and updates throughout
2539 :     the system:
2540 :    
2541 :     Basis:
2542 :     - Implemented IntInf.
2543 :     - Made LargeInt a projection of IntInf (by filtering through INTEGER).
2544 : mblume 1348 - Added some missing Real64 operations, most notably Real.toLargeInt.
2545 : mblume 1347 - Added FixedInt as a synonym for Int32.
2546 :    
2547 :     compiler:
2548 :     * Added support for a built-in intinf type.
2549 :     - literals
2550 :     - pattern matching
2551 :     - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
2552 :     - overloading on literals and operations
2553 :    
2554 :     This required adding a primitive type intinf, some additional
2555 :     primops, and implementations for several non-trivial intinf
2556 :     operations in Core. (The intinf type is completely abstract
2557 :     to the compiler; all operations get delegated back to the Core.)
2558 :    
2559 :     * Intinf equality is handled by polyequal. However, the compiler
2560 :     does not print its usual warning in this case (since polyequal
2561 :     is the right thing to do there).
2562 :    
2563 :     * Improved the organization of structure InlineT.
2564 :    
2565 :     * A word about conversion primops:
2566 :     If conversions involving intinf do not cancel out during
2567 : mblume 1348 CPS contract, then the compiler must insert calls to Core functions.
2568 : mblume 1347 Since all core access must be resolved already during the FLINT
2569 :     translate phase, it would be too late a the time of CPS contract
2570 : mblume 1348 to add new Core calls. For this reason, conversion primops
2571 : mblume 1347 for intinf carry two arguments: 1. the numeric argument that
2572 :     they are supposed to convert, and 2. the Core function that
2573 :     can help with this conversion if necessary. If CPS contract
2574 : mblume 1348 eliminates a primop, then the associated Core function becomes
2575 :     dead and goes away. Intinf conversion primops that do not get
2576 :     eliminated by CPS contract get rewritten into calls of their
2577 :     core functions by a separate, new phase.
2578 : mblume 1347
2579 :     interactive system:
2580 :     - Control.Print.intinfDepth controls max length of intinf constants
2581 :     being printed. (Analogous to Control.Print.stringDepth.)
2582 :     - Cleanup in printutil and pputil: got rid of unused stuff and
2583 :     duplicates; replaced some of the code with code that makes better
2584 :     use of library functionality.
2585 :    
2586 :     CM:
2587 :     Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
2588 :     are no longer silent.
2589 :    
2590 :     CKIT:
2591 :     Fixed mismatched uses of Int32 and LargeInt. I always decided
2592 :     in favor of LargeInt -- which is now the same as IntInf.
2593 :     CKIT-knowledgable people should check whether this is what's
2594 :     intended and otherwise change things back to using Int32 or
2595 :     FixedInt.
2596 :    
2597 :     Throughout the code:
2598 : mblume 1348 Started using IntInf.int literals and built-in operations
2599 : mblume 1347 (e.g., comparison with 0) where this seems appropriate.
2600 :    
2601 :    
2602 :     ----------------------------------------------------------------------
2603 : macqueen 1343 Name: Dave MacQueen (dbm@cs.uchicago.edu)
2604 :     Date: 2003/08/13 11:36:00 CDT
2605 :     Tag: dbm-20030813-mcz-merge1
2606 :     Description:
2607 :    
2608 :     Merging changes from the mcz-branch development branch into trunk.
2609 :     These changes involve replacement of the emulated old prettyprinter
2610 :     interface with direct use of the SML/NJ Lib PP library, and fixing
2611 :     of a couple of bugs (895, 1186) relating to error messages. A new
2612 :     prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
2613 :     has been added.
2614 :    
2615 :     ----------------------------------------------------------------------
2616 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2617 : mblume 1342 Date: 2003/08/11 15:45:00 CDT
2618 :     Tag: blume-20030811-windows
2619 :     Description:
2620 :    
2621 :     Version number bumped to 110.42.9. NEW BOOTFILES!!!
2622 :    
2623 :     http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
2624 :    
2625 :     This patch restores SML/NJ's ability to run under win32. There are a
2626 :     number of changes, including fixes for several bugs that had gone
2627 :     unnoticed until now:
2628 :    
2629 :     - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
2630 :     trying to run the win32 version from within cygwin.)
2631 :    
2632 :     - There are a number of simple .bat scripts that substitute for their
2633 :     corresponding Unix shell-scripts. (See below.)
2634 :    
2635 :     - The internals of ml-build have been modified slightly. The main
2636 :     difference is that instead of calling ".link-sml" (or link-sml.bat)
2637 :     using OS.Process.system, the ML process delegates this task back
2638 :     to the script. Otherwise problems arise in mixed environments such
2639 :     as Cygwin where scripts look and work like Unix scripts, but
2640 :     where OS.Process.system cannot run them.
2641 :    
2642 :     - In CM, the srcpath pickler used native pathname syntax -- which
2643 :     is incorrect in the case of cross-compilation. The new pickle format
2644 :     is independent of platform-specific naming conventions.
2645 :    
2646 :     - Path configuration files (such as lib/pathconfig) can now choose
2647 :     between native and standard syntax. Placing a line of the form
2648 :    
2649 :     standard!
2650 :    
2651 :     into the file causes all subsequent paths to be interpreted using
2652 :     CM standard pathname syntax (= Unix conventions); a line
2653 :    
2654 :     native!
2655 :    
2656 :     switches back to native style. This was needed so that
2657 :     path config files can be written portably, see src/system/pathconfig.
2658 :    
2659 :     - Runtime system:
2660 :    
2661 :     - win32-filesys.c: get_file_time and set_file_time now
2662 :     access modification time, not creation time.
2663 :    
2664 :     - I/O code made aware of new array representation.
2665 :    
2666 :     - Bug fixes in X86.prim.masm.
2667 :    
2668 :     - src/system/makeml made aware of win32. (For use under cygwin
2669 :     and other Unix-environments for windows.)
2670 :    
2671 :     - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
2672 :     which caused BinIO.inputAll to fail consistently.
2673 :    
2674 :     .bat scripts:
2675 :    
2676 :     Windows .bat scripts assume that SMLNJ_HOME is defined.
2677 :    
2678 :     - sml.bat, ml-yacc.bat, ml-lex.bat: Driver scripts for standalone
2679 :     applications (sml, ml-yacc, ml-lex).
2680 :     - ml-build.bat: analogous to ml-build.
2681 :     - config\install.bat: Analogous to config/install.sh. This requires
2682 :     that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2683 :     (nmake etc. must be on the path, and vcvars32 must have been run.)
2684 :     Moreover, sources for ml-lex and ml-yacc need to exist under src,
2685 :     and the bootfile hierarchy must have been unpacked under
2686 :     sml.boot.x86-win32.
2687 :     The script is very primitive and does a poor job at error checking.
2688 :     It only installs the base system, ml-lex, and ml-yacc. No other
2689 :     libraries are being installed (i.e., you get only those that
2690 :     are part of the compiler.)
2691 :     - link-sml.bat: analogous to .link-sml, but not currently used
2692 :    
2693 :     Unrelated bug fixes:
2694 :    
2695 :     - ml-nlffigen now exports structures ST_* corresponding to incomplete
2696 :     types.
2697 :     - Added getDevice to PP/src/pp-debug-fn.sml. (Would not compile
2698 :     otherwise.)
2699 :    
2700 :     ----------------------------------------------------------------------
2701 : macqueen 1337 Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2702 :     Date: 2003/06/17
2703 :     Tag: macqueen-20030617-bug895
2704 :     Description:
2705 :    
2706 :     Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2707 :     Tag will be used for new development branch (mcz-branch) for
2708 :     use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2709 :    
2710 :     ----------------------------------------------------------------------
2711 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2712 : mblume 1335 Date: 2003/05/27 16:55:00 CDT
2713 :     Tag: blume-20030527-polyeq
2714 :     Description:
2715 :    
2716 :     Tried to eliminated most cases of polymorphic equality.
2717 :    
2718 :     ----------------------------------------------------------------------
2719 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2720 : mblume 1334 Date: 2003/05/21 17:45:00 CDT
2721 :     Tag: blume-20030517-complete
2722 :     Description:
2723 :    
2724 :     Two changes:
2725 :    
2726 :     1. Added a flag for controlling whether non-exhaustive bindings will
2727 :     be treated as errors (default is false).
2728 :     2. Cleaned up the *entire* source tree so that CMB.make goes through
2729 :     without a single non-exhaustive match- or bind warning.
2730 :    
2731 :     ----------------------------------------------------------------------
2732 : mblume 1368 Name: Matthias Blume (blume (at) tti - c (dot) org)
2733 : mblume 1332 Date: 2003/05/17 10:20:00 CDT
2734 :     Tag: blume-20030517-absyn
2735 :     Description:
2736 :    
2737 :     1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2738 :    
2739 :     This mainly affects the quality of error messages. However, some
2740 :     of the code is now more straightforward than before. (Treatment of
2741 :     the above four constructs in translate.sml is much simpler than
2742 :     the "macro-expansion" that was going on before. Plus, the mach-
2743 :     compiler no longer gets invoked just to be able to compile an
2744 :     if-expression.)
2745 :    
2746 :     2. The ErrorMsg.Error exception is now caught and absorbed by the
2747 :     interactive loop.
2748 :    
2749 :     ----------------------------------------------------------------------
2750 : allenleung 1327 Name: Allen Leung
2751 :     Date: 2003/05/16 13:05:00 CDT
2752 :     Tag: leunga-20030516-cygwin-runtime
2753 :     Description:
2754 :    
2755 : blume 2155 Ported the runtime system to cygwin, which uses the unix
2756 : allenleung 1327 x86-unix bin files. Missing/buggy features:
2757 :    
2758 : blume 2155 o getnetbyname, getnetbyaddr: these functions seem to be missing in
2759 : allenleung 1327 the Cygwin library.
2760 :     o Ctrl-C handling may be flaky.
2761 :     o Windows system calls and Windows I/O are not supported.
2762 :    
2763 :     A new set of binfiles is located at:
2764 :    
2765 :     http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2766 :    
2767 :     This is only needed for bootstrapping the cygwin version of smlnj.
2768 :     Other x86 versions can use the existing binfiles.
2769 :    
2770 :     ----------------------------------------------------------------------
2771 : mblume 1317 Name: Matthias Blume
2772 :     Date: 2003/04/08 15:42:00 CDT
2773 :     Tag: blume-20030408-listpair
2774 :     Description:
2775 :    
2776 :     1. Added a target 'mlrisc' to installer.
2777 :    
2778 :     2. Added missing elements to structure ListPair.
2779 :    
2780 :     ----------------------------------------------------------------------
2781 : leunga 1296 Name: Allen Leung
2782 :     Date: 2003/01/07 10:40:00 EST
2783 :     Tag: leunga-20030107-int-rem
2784 :     Description:
2785 :    
2786 :     Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2787 :     The arguments to the SUBL instruction were swapped.
2788 :    
2789 :     ----------------------------------------------------------------------
2790 : blume 1229 Name: Matthias Blume
2791 : blume 1294 Date: 2002/12/12 16:25:00 EST
2792 :     Tag: blume-20021212-risc-ra
2793 :     Description:
2794 :    
2795 :     Fixed a serious bug in the rewrite code for FP spilling/reloading that
2796 :     sent the RA into an infinite loop when floating point registers get
2797 :     spilled. (Because of this bug, e.g., nucleic stopped compiling between
2798 :     110.37 and 110.38.)
2799 :     There was another set of potential problems related to the handling of
2800 :     MLRISC annotations (but those did not yet cause real problems, apparently).
2801 :    
2802 :     ----------------------------------------------------------------------
2803 :     Name: Matthias Blume
2804 : blume 1291 Date: 2002/12/06 22:40:00 EST
2805 :     Tag: blume-20021206-cm-fileid
2806 :     Description:
2807 :    
2808 :     Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2809 :     This fixes the problem of CM getting confused by files that suddenly
2810 :     change their identity (e.g., by getting unlinked and recreated by some
2811 :     text editor such as vi). There might be a better/cheaper/cleaner way
2812 :     of doing this, but for now this will have to do.
2813 :    
2814 :     ----------------------------------------------------------------------
2815 :     Name: Matthias Blume
2816 : blume 1288 Date: 2002/10/28 09:50:00 EST
2817 : blume 2155 Tag: blume-20021028-typecheck
2818 : blume 1288 Description:
2819 :    
2820 :     Exported structure Typecheck from $smlnj/viscomp/core.cm.
2821 :    
2822 :     ----------------------------------------------------------------------
2823 :     Name: Matthias Blume
2824 : blume 1282 Date: 2002/10/17 09:10:00 EDT
2825 : blume 1279 Tag: Release_110_42
2826 :     Description:
2827 :    
2828 : blume 1282 In good old tradition, there has been a slight hiccup so that we have
2829 :     to patch 110.42 after the fact. The old release tag has been replaced
2830 :     (see below).
2831 :    
2832 :     The change solves a problem with two competing approaches the
2833 :     configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2834 :     each other's way.
2835 :    
2836 :     This change only affects the runtime system code and the installer script.
2837 :     (No new bootfiles.)
2838 :    
2839 :     ----------------------------------------------------------------------
2840 :     Name: Matthias Blume
2841 :     Date: 2002/10/16 12:00:00 EDT
2842 :     Tag: Release_110_42_removed
2843 :     Description:
2844 :    
2845 : blume 1279 New working release. New bootfiles.
2846 :    
2847 :     ----------------------------------------------------------------------
2848 :     Name: Matthias Blume
2849 : blume 1273 Date: 2002/10/10 13:10:00 EDT
2850 :     Tag: blume-20021010-ppc-divs
2851 :     Description:
2852 :    
2853 :     The mltree operator DIVS must be implemented with an overflow check on
2854 :     the PPC because the hardware indicates divide-by-zero using "overflow" as
2855 :     well.
2856 :    
2857 :     ----------------------------------------------------------------------
2858 :     Name: Matthias Blume
2859 : blume 1261 Date: 2002/07/23 11:45:00
2860 :     Tag: blume-20020723-smlnj-home
2861 :     Description:
2862 :    
2863 :     Sml now senses the SMLNJ_HOME environment variable. If this is set,
2864 :     then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2865 :     CM_PATHCONFIG is also set), the path configuration file is assumed
2866 :     to be in $SMLNJ_HOME/lib/pathconfig. This way one can easily move
2867 :     the entire tree to some other place and everything will "just work".
2868 :    
2869 :     (Companion commands such as ml-build and ml-makedepend also sense this
2870 :     variable.)
2871 :    
2872 :     ----------------------------------------------------------------------
2873 :     Name: Matthias Blume
2874 : blume 1259 Date: 2002/07/12 21:19:00 EDT
2875 :     Tag: blume-20020712-liveness
2876 :     Description:
2877 :    
2878 :     Exported two useful "step" functions from liveness module (MLRISC).
2879 :    
2880 :     ----------------------------------------------------------------------
2881 :     Name: Matthias Blume
2882 : blume 1253 Date: 2002/07/05 16:00 EDT
2883 :     Tag: Release_110_41
2884 :     Description:
2885 :    
2886 :     New working release. New bootfiles.
2887 :    
2888 :     ----------------------------------------------------------------------
2889 :     Name: Matthias Blume
2890 : blume 1250 Date: 2002/07/05 10:25:00 EDT
2891 :     Tag: blume-20020705-btimp
2892 :     Description:
2893 :    
2894 :     Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2895 :     other clients can set up backtracing support.
2896 :    
2897 :     ----------------------------------------------------------------------
2898 :     Name: Matthias Blume
2899 : blume 1249 Date: 2002/06/25 17:23:00 EDT
2900 :     Tag: blume-20020625-fpmax
2901 :     Description:
2902 :    
2903 :     Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2904 :     case. (The sense of the isNaN test was reversed -- which made min and
2905 :     max always return their first argument.)
2906 :    
2907 :     ----------------------------------------------------------------------
2908 :     Name: Matthias Blume
2909 : blume 1240 Date: 2002/06/11
2910 :     Tag: blume-20020611-unixpath
2911 :     Description:
2912 :    
2913 :     Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2914 :    
2915 :     ----------------------------------------------------------------------
2916 :     Name: Matthias Blume
2917 : blume 1238 Date: 2002/06/10 16:35:00 EDT
2918 :     Tag: blume-20020610-ieeereal
2919 :     Description:
2920 :    
2921 :     I back-ported my implementation of IEEEReal.fromString from the
2922 :     idlbasis-devel branch so that we can test it.
2923 :    
2924 :     Another small change is that ppDec tries to give more information
2925 :     than just "<sig>" in the case of functors. However, this code is
2926 :     broken in some mysterious way if the functor's body's signature
2927 :     has not been declared by ascription but gets inferred from the
2928 :     implementation. This needs fixing...
2929 :    
2930 :     ----------------------------------------------------------------------
2931 :     Name: Matthias Blume
2932 : blume 1229 Date: 2002/05/31
2933 :     Tag: blume-20020531-btrace-mode
2934 :     Description:
2935 :    
2936 :     Resurrected SMLofNJ.Internals.BTrace.mode. (It accidentally fell by
2937 :     the wayside when I switched over to using Controls everywhere.)
2938 :    
2939 :     ----------------------------------------------------------------------
2940 : george 1222 Name: Lal George
2941 :     Date: 2002/05/23 12:21:40 EDT
2942 :     Tag: george-20020523-visual-labels
2943 :     Description:
2944 :    
2945 :     Labels are now displayed in the graphical output to make
2946 :     the fall-through and target blocks obvious.
2947 :    
2948 :     ----------------------------------------------------------------------
2949 : blume 1204 Name: Matthias Blume
2950 : blume 1220 Date: 2002/05/22 11:03:00 EDT
2951 :     Tag: blume-20020522-shrink
2952 :     Description:
2953 :    
2954 :     John tweaked yesterday's fix for 1131 to handle an out-of-memory
2955 :     situation that comes up when allocating huge arrays.
2956 :    
2957 :     ----------------------------------------------------------------------
2958 :     Name: Matthias Blume
2959 : blume 1215 Date: 2002/05/21 16:00:00 EDT
2960 :     Tag: Release_110_40
2961 :     Description:
2962 :    
2963 :     New working release (110.40). New bootfiles.
2964 :    
2965 :     [Also: John Reppy fixed GC bug 1131.]
2966 :    
2967 :     ----------------------------------------------------------------------
2968 :     Name: Matthias Blume
2969 : blume 1212 Date: 2002/05/21 12:35:00 EDT
2970 :     Tag: blume-20020521-cmdoc
2971 :     Description:
2972 :    
2973 :     CM documentation update.
2974 :    
2975 :     ----------------------------------------------------------------------
2976 :     Name: Matthias Blume
2977 : blume 1211 Date: 2002/05/21 10:55:00 EDT
2978 :     Tag: blume-20020521-misc
2979 :     Description:
2980 :    
2981 :     - John tweaked runtime to be silent on heap export (except when
2982 :     GC messages are on).
2983 :     - I added a few more things (cross-compiling versions of CMB) to
2984 :     config/preloads (as suggestions).
2985 :    
2986 :     ----------------------------------------------------------------------
2987 :     Name: Matthias Blume
2988 : blume 1208 Date: 2002/05/20 22:25:00 EDT
2989 :     Tag: blume-20020520-controls
2990 :     Description:
2991 :    
2992 :     - Added ControlUtil structure to control-lib.cm.
2993 :     - Use it throughout.
2994 :     - Used Controls facility to define MLRISC controls (as opposed to
2995 :     registering MLRISC control ref cells with Controls after the fact)
2996 :     - Fixed messed-up controls priorities.
2997 :    
2998 :     * Removed again all the stuff from config/preloads that one wouldn't
2999 :     be able to preload at the time the initial heap image is built.
3000 :     (Many libraries, e.g., CML, do not exist yet at this time. The
3001 :     only libraries that can be preloaded via config/preloads are those
3002 :     that come bundled with the bootfiles.)
3003 :    
3004 :     ----------------------------------------------------------------------
3005 :     Name: Matthias Blume
3006 : blume 1204 Date: 2002/05/20 10:59:00 EDT
3007 :     Tag: blume-20020520-preloads
3008 :     Description:
3009 :    
3010 :     Added a lot of commented-out suggestions for things to be included
3011 :     in config/preloads.
3012 :    
3013 :     ----------------------------------------------------------------------
3014 : leunga 1203 Name: Allen Leung
3015 :     Date: 2002/05/18 14:20:00 EDT
3016 :     Tag: leunga-20020518-mdl
3017 :     Description:
3018 :    
3019 :     o Made the mdl tool stuff compile and run again.
3020 :     o I've disabled all the stuff that depends on RTL specifications; they
3021 :     are all badly broken anyway.
3022 :    
3023 :     ----------------------------------------------------------------------
3024 : blume 1201 Name: Matthias Blume
3025 :     Date: 2002/05/17 16:49:00 EDT
3026 :     Tag: blume-20020517-controls
3027 :     Description:
3028 :    
3029 :     0. John Reppy made several modifications to the SML/NJ library.
3030 :     In particular, there is a shiny new controls-lib.cm.
3031 :    
3032 :     1. Pushed new controls interface through compiler so that everything
3033 :     compiles again.
3034 :    
3035 :     2. Added FormatComb and FORMAT_COMB to the CML version of the
3036 :     SML/NJ library (so that CML compiles again).
3037 :    
3038 :     3. Modified init scripts because XXX_DEFAULT environment variables
3039 :     are no longer with us. (Boot-time initialization is now done
3040 :     using the same environment variables that are also used for
3041 :     startup-time initialization of controls.)
3042 :    
3043 :     ----------------------------------------------------------------------
3044 : george 1192 Name: Lal George
3045 :     Date: 2002/05/15 09:20:10 EDT
3046 :     Tag: george-20020515-pseudo-op-decls
3047 :     Description:
3048 :    
3049 :     All pseudo-ops emitted before the first segment declaration
3050 :     such as TEXT, DATA, and BSS directives are assumed to be global
3051 :     declarations and are emitted first in the assembly file. This is
3052 :     useful in a number of situations where one has pseudo-ops that are not
3053 :     specific to any segment, and also works around the constraint that one
3054 :     cannot have client pseudo-ops in the TEXT segment.
3055 :    
3056 :     Because no segment is associated with these declarations it is
3057 :     an error to allocate any space or objects before the first segment
3058 :     directive and an exception will be raised. However, we cannot make
3059 :     this check for client pseudo-ops.
3060 :    
3061 :     These top level declarations are a field in the CFG graph_info.
3062 :     In theory you can continue to add to this field after the CFG has been
3063 :     built -- provided you know what you are doing;-)
3064 :    
3065 :     ----------------------------------------------------------------------
3066 : blume 1175 Name: Matthias Blume
3067 : blume 1190 Date: 2002/05/13 16:40:00 EDT
3068 :     Tag: blume-20020513-pp-etc
3069 :     Description:
3070 :    
3071 :     A few minor bugfixes:
3072 :    
3073 :     - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
3074 :     (Bogus printouts for redefined bindings still occur. Compiler
3075 :     bug should no longer occur now. We need to redo the prettyprinter
3076 :     from scratch.)
3077 :    
3078 :     - CM pathname printer now also adds escape sequences for ( and )
3079 :    
3080 :     - commend and docu fixes for ml-nlffi
3081 :    
3082 :     ----------------------------------------------------------------------
3083 :     Name: Matthias Blume
3084 : blume 1189 Date: 2002/05/10 16:40:00 EDT
3085 :     Tag: blume-20020510-erg-textio
3086 :     Description:
3087 :    
3088 :     Applied the following bugfix provided by Emden Gansner:
3089 :    
3090 :     Output is corrupted when outputSubstr is used rather than output.
3091 :    
3092 : blume 2155 The problem occurs when a substring
3093 : blume 1189
3094 :     ss = (s, dataStart, dataLen)
3095 :    
3096 : blume 2155 where dataStart > 0, fills a stream buffer with avail bytes left.
3097 :     avail bytes of s, starting at index dataStart, are copied into the
3098 : blume 1189 buffer, the buffer is flushed, and then the remaining dataLen-avail
3099 : blume 2155 bytes of ss are copied into the beginning of the buffer. Instead of
3100 :     starting this copy at index dataStart+avail in s, the current code
3101 : blume 1189 starts the copy at index avail.
3102 :    
3103 :     Fix:
3104 :     In text-io-fn.sml, change line 695 from
3105 :     val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
3106 : blume 2155 to
3107 : blume 1189 val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
3108 :    
3109 :     ----------------------------------------------------------------------
3110 :     Name: Matthias Blume
3111 : blume 1186 Date: 2002/04/12 13:55:00 EDT
3112 :     Tag: blume-20020412-assyntax
3113 :     Description:
3114 :    
3115 :     1. Grabbed newer assyntax.h from the XFree86 project.
3116 :     2. Fiddled with how to compile X86.prim.asm without warnings.
3117 :     3. (Very) Minor cleanup in CM.
3118 :    
3119 :     ----------------------------------------------------------------------
3120 :     Name: Matthias Blume
3121 : blume 1185 Date: 2002/04/01 (no joke!) 17:07:00 EST
3122 :     Tag: blume-20020401-x86div
3123 :     Description:
3124 :    
3125 :     Added full support for div/mod/rem/quot on the x86, using the machine
3126 :     instruction's two results (without clumsily recomputing the remainder)
3127 :     directly where appropriate.
3128 :    
3129 :     Some more extensive power-of-two support was added to the x86 instruction
3130 :     selector (avoiding expensive divs, mods, and muls where they can be
3131 :     replaced with cheaper shifts and masks). However, this sort of thing
3132 :     ought to be done earlier, e.g., within the CPS optimizer so that
3133 :     all architectures benefit from it.
3134 :    
3135 :     The compiler compiles to a fixed point, but changes might be somewhat
3136 :     fragile nevertheless. Please, report any strange things that you might
3137 :     see wrt. div/mod/quot/rem...
3138 :    
3139 :     ----------------------------------------------------------------------
3140 :     Name: Matthias Blume
3141 : blume 1184 Date: 2002/03/29 17:22:00
3142 :     Tag: blume-20020329-div
3143 :     Description:
3144 :    
3145 :     Fixed my broken div/mod logic. Unfortunately, this means that the
3146 :     inline code for div/mod now has one more comparison than before.
3147 :     Fast paths (quotient > 0 or remainder = 0) are not affected, though.
3148 :     The problem was with quotient = 0, because that alone does not tell
3149 :     us which way the rounding went. One then has to look at whether
3150 :     remainder and divisor have the same sign... :(
3151 :    
3152 :     Anyway, I replaced the bootfiles with fresh ones...
3153 :    
3154 :     ----------------------------------------------------------------------
3155 :     Name: Matthias Blume
3156 : blume 1183 Date: 2002/03/29 14:10:00 EST
3157 :     Tag: blume-20020329-inlprims
3158 :     Description:
3159 :    
3160 :     NEW BOOTFILES!!! Version number bumped to 110.39.3.
3161 :    
3162 :     Primops have changed. This means that the bin/boot-file formats have
3163 :     changed as well.
3164 :    
3165 :     To make sure that there is no confusion, I made a new version.
3166 :    
3167 :    
3168 :     CHANGES:
3169 :    
3170 :     * removed REMT from mltree (remainder should never overflow).
3171 :    
3172 :     * added primops to deal with divisions of all flavors to the frontend
3173 :    
3174 :     * handled these primops all the way through so they map to their respective
3175 :     MLRISC support
3176 :    
3177 :     * used these primops in the implementation of Int, Int32, Word, Word32
3178 :    
3179 :     * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
3180 :    
3181 :     * parameterized INLMIN, INLMAX, and INLABS by a numkind
3182 :    
3183 :     * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
3184 :     floating point
3185 :    
3186 :     * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
3187 :     and Real (but Real.abs maps to a separate floating-point-only primop)
3188 :    
3189 :    
3190 :     TODO items:
3191 :    
3192 :     * Hacked Alpha32 instruction selection, disabling the selection of REMx
3193 :     instructions because the machine instruction encoder cannot handle
3194 :     them. (Hppa, PPC, and Sparc instruction selection did not handle
3195 :     REM in the first place, and REM is supported by the x86 machine coder.)
3196 :    
3197 :     * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
3198 :     selection phase. (The two can be streamlined because the hardware
3199 :     delivers both quotient and remainder at the same time anyway.)
3200 :    
3201 :     * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
3202 :     (Currently the behavior is inconsistent both across architectures and
3203 :     wrt. the draft Basis spec.)
3204 :    
3205 :     * Word8 should eventually be handled natively, too.
3206 :    
3207 :     * There seems to be one serious bug in mltree-gen.sml. It appears, though,
3208 :     as if there currently is no execution path that could trigger it in
3209 :     SML/NJ. (The assumptions underlying functions arith and promotable do not
3210 :     hold for things like multiplication and division.)
3211 :    
3212 :     ----------------------------------------------------------------------
3213 :     Name: Matthias Blume
3214 : blume 1181 Date: 2002/03/27 16:27:00 EST
3215 :     Tag: blume-20020327-mlrisc-divisions
3216 :     Description:
3217 :    
3218 :     Added support for all four division operations (ML's div, mod, quot,
3219 :     and rem) to MLRISC. In the course of doing so, I also rationalized
3220 :     the naming (no more annoying switch-around of DIV and QUOT), by
3221 :     parameterizing the operation by div_rounding_mode (which can be either
3222 :     DIV_TO_ZERO or DIV_TO_NEGINF).
3223 :    
3224 :     The generic MLTreeGen functor takes care of compiling all four
3225 :     operations down to only round-to-zero div.
3226 :    
3227 :     Missing pieces:
3228 :    
3229 :     * Doing something smarter than relying on MLTreeGen on architectures
3230 :     like, e.g., the x86 where hardware division delivers both quotient and
3231 :     remainder at the same time. With this, the implementation of the
3232 :     round-to-neginf operations could be further streamlined.
3233 :    
3234 :     * Remove inlining support for div/mod/rem from the frontend and replace it
3235 :     with primops that get carried through to the backend. Do this for all
3236 :     int and word types.
3237 :    
3238 :     ----------------------------------------------------------------------
3239 :     Name: Matthias Blume
3240 : blume 1180 Date: 2002/03/25 17:25:00 EST
3241 :     Tag: blume-20020325-divmod
3242 :     Description:
3243 :    
3244 :     I improved (hopefully without breaking them) the implementation of Int.div,
3245 :     Int.mod, and Int.rem. For this, the code in translate.sml now takes
3246 :     advantage of the following observations:
3247 :    
3248 :     Let q = x quot y r = x rem y
3249 :     d = x div y m = x mod y
3250 :    
3251 :     where "quot" is the round-to-zero version of integer division that
3252 :     hardware usually provides. Then we have:
3253 :    
3254 :     r = x - q * y where neither the * nor the - will overflow
3255 :     d = if q >= 0 orelse x = q * y then q else q - 1
3256 :     where neither the * nor the - will overflow
3257 :     m = if q >= 0 orelse r = 0 then r else r + y
3258 :     where the + will not overflow
3259 :    
3260 :     This results in substantial simplification of the generated code.
3261 :     The following table shows the number of CFG nodes and edges generated
3262 :     for
3263 :     fun f (x, y) = x OPER y
3264 :     (* with OPER \in div, mod, quot, rem *)
3265 :    
3266 :    
3267 :     OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
3268 :     --------------------------------------------------------
3269 :     div | 24 | 39 | 12 | 16
3270 :     mod | 41 | 71 | 12 | 16
3271 :     quot | 8 | 10 | 8 | 10
3272 :     rem | 10 | 14 | 8 | 10
3273 :    
3274 :    
3275 :     ----------------------------------------------------------------------
3276 :     Name: Matthias Blume
3277 : blume 1179 Date: 2002/03/25 22:06:00 EST
3278 :     Tag: blume-20020325-cprotobug
3279 :     Description:
3280 :    
3281 :     Fixed a bug in cproto (c prototype decoder).
3282 :    
3283 :     ----------------------------------------------------------------------
3284 :     Name: Matthias Blume
3285 : blume 1178 Date: 2002/03/25 16:00:00 EST
3286 :     Tag: blume-20020325-raw-primops
3287 :     Description:
3288 :    
3289 :     I did some cleanup to Allen's new primop code and
3290 :     replaced yesterday's bootfiles with new ones.
3291 :     (But they are stored in the same place.)
3292 :    
3293 :     ----------------------------------------------------------------------
3294 :     Name: Matthias Blume
3295 : blume 1175 Date: 2002/03/24 22:40:00 EST
3296 :     Tag: blume-20020324-bootfiles
3297 :     Description:
3298 :    
3299 :     Made the bootfiles that Allen asked for.
3300 :    
3301 :     ----------------------------------------------------------------------
3302 : leunga 1174 Name: Allen Leung
3303 :     Date: 2002/03/23 15:50:00 EST
3304 :     Tag: leunga-20020323-flint-cps-rcc-primops
3305 :     Description:
3306 :    
3307 :     1. Changes to FLINT primops:
3308 :    
3309 :     (* make a call to a C-function;
3310 :     * The primop carries C function prototype information and specifies
3311 :     * which of its (ML-) arguments are floating point. C prototype
3312 :     * information is for use by the backend, ML information is for
3313 :     * use by the CPS converter. *)
3314 :     | RAW_CCALL of { c_proto: CTypes.c_proto,
3315 :     ml_args: ccall_type list,
3316 :     ml_res_opt: ccall_type option,
3317 :     reentrant : bool
3318 :     } option
3319 :     (* Allocate uninitialized storage on the heap.
3320 :     * The record is meant to hold short-lived C objects, i.e., they
3321 : blume 2155 * are not ML pointers. With the tag, the representation is
3322 : leunga 1174 * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
3323 :     *)
3324 : blume 2155 | RAW_RECORD of {tag:bool,sz:int}
3325 : leunga 1174 and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
3326 :    
3327 :     2. These CPS primops are now overloaded:
3328 :    
3329 :     rawload of {kind:numkind}
3330 :     rawstore of {kind:numkind}
3331 :    
3332 :     The one argument form is:
3333 :    
3334 :     rawload {kind} address
3335 :    
3336 :     The two argument form is:
3337 : blume 2155
3338 : leunga 1174 rawload {kind} [ml object, byte-offset]
3339 :    
3340 :     3. RAW_CCALL/RCC now takes two extra arguments:
3341 :    
3342 :     a. The first is whether the C call is reentrant, i.e., whether
3343 :     ML state should be saved and restored.
3344 :     b. The second argument is a string argument specifying the name of
3345 :     library and the C function.
3346 :    
3347 :     These things are currently not handled in the code generator, yet.
3348 :    
3349 :     4. In CProto,
3350 :    
3351 : blume 2155 An encoding type of "bool" means "ml object" and is mapped into
3352 :     C prototype of PTR. Note that "bool" is different than "string",
3353 : leunga 1174 even though "string" is also mapped into PTR, because "bool"
3354 :     is assigned an CPS type of BOGt, while "string" is assigned INT32t.
3355 :    
3356 :     5. Pickler/unpicker
3357 :    
3358 :     Changed to handle RAW_RECORD and newest RAW_CCALL
3359 :    
3360 :     6. MLRiscGen,
3361 :    
3362 :     1. Changed to handle the new rawload/rawstore/rawrecord operators.
3363 :     2. Code for handling C Calls has been moved to a new module CPSCCalls,
3364 :     in the file CodeGen/cpscompile/cps-c-calls.sml
3365 :    
3366 :     7. Added the conditional move operator
3367 :    
3368 :     condmove of branch
3369 :    
3370 : blume 2155 to cps. Generation of this is still buggy so it is currently
3371 : leunga 1174 disabled.
3372 : blume 2155
3373 : leunga 1174 ----------------------------------------------------------------------
3374 : george 1168 Name: Lal George
3375 : blume 2155 Date: 2002/03/22 14:18:25 EST
3376 : george 1171 Tag: george-20020322-cps-branch-prob
3377 : george 1168 Description:
3378 :    
3379 : blume 2155 Implemented the Ball-Larus branch prediction-heuristics, and
3380 : george 1168 incorporated graphical viewers for control flow graphs.
3381 :    
3382 : george 1171 Ball-Larus Heuristics:
3383 : george 1168 ---------------------
3384 :     See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
3385 :    
3386 :     By design it uses the Dempster-Shafer theory for combining
3387 :     probabilities. For example, in the function:
3388 :    
3389 :     fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
3390 :    
3391 :     the ball-larus heuristics predicts that the n=0 is unlikely
3392 :     (OH-heuristic), and the 'then' branch is unlikely because of the
3393 :     RH-heuristic -- giving the 'then' branch an even lower combined
3394 : george 1171 probability using the Dempster-Shafer theory.
3395 : george 1168
3396 :     Finally, John Reppy's loop analysis in MLRISC, further lowers the
3397 :     probability of the 'then' branch because of the loop in the else
3398 :     branch.
3399 :    
3400 :    
3401 :     Graphical Viewing:
3402 : blume 2155 ------------------
3403 : george 1168 I merely plugged in Allen's graphical viewers into the compiler. The
3404 :     additional code is not much. At the top level, saying:
3405 :    
3406 :     Control.MLRISC.getFlag "cfg-graphical-view" := true;
3407 :    
3408 :     will display the graphical view of the control flow graph just before
3409 :     back-patching. daVinci must be in your path for this to work. If
3410 :     daVinci is not available, then the default viewer can be changed
3411 :     using:
3412 :    
3413 :     Control.MLRISC.getString "viewer"
3414 :    
3415 :     which can be set to "dot" or "vcg" for the corresponding viewers. Of
3416 :     course, these viewers must be in your path.
3417 :    
3418 :     The above will display the compilation unit at the level of clusters,
3419 :     many of which are small, boring, and un-interesting. Also setting:
3420 :    
3421 :     Control.MLRISC.getInt "cfg-graphical-view_size"
3422 :    
3423 :     will display clusters that are larger than the value set by the above.
3424 :    
3425 : blume 2155
3426 : george 1168 ----------------------------------------------------------------------
3427 : blume 1159 Name: Matthias Blume
3428 :     Date: 2002/03/21 22:20:00 EST
3429 :     Tag: blume-20020321-kmp-bugfix
3430 :     Description:
3431 :    
3432 :     Changed the interface to the KMP routine in PreString and fixed
3433 :     a minor bug in one place where it was used.
3434 :    
3435 :     ----------------------------------------------------------------------
3436 : leunga 1156 Name: Allen Leung
3437 : leunga 1158 Date: 2002/03/21 20:30:00 EST
3438 :     Tag: leunga-20020321-cfg
3439 :     Description:
3440 :    
3441 :     Fixed a potential problem in cfg edge splitting.
3442 :    
3443 :     ----------------------------------------------------------------------
3444 :     Name: Allen Leung
3445 : leunga 1156 Date: 2002/03/21 17:15:00 EST
3446 :     Tag: leunga-20020321-x86-fp-cfg
3447 :     Description:
3448 :    
3449 : blume 2155 1. Recoded the buggy parts of x86-fp.
3450 : leunga 1156
3451 :     a. All the block reordering code has been removed.
3452 :     We now depend on the block placement phases to do this work.
3453 :    
3454 :     b. Critical edge splitting code has been simplified and moved into the
3455 :     CFG modules, as where they belong.
3456 :    
3457 :     Both of these were quite buggy and complex. The code is now much, much
3458 :     simpler.
3459 :    
3460 :     2. X86 backend.
3461 :    
3462 :     a. Added instructions for 64-bit support. Instruction selection for
3463 :     64-bit has not been committed, however, since that
3464 : blume 2155 requires changes to MLTREE which haven't been approved by
3465 : leunga 1156 Lal and John.
3466 :    
3467 : blume 2155 b. Added support for FUCOMI and FUCOMIP when generating code for
3468 : leunga 1156 PentiumPro and above. We only generate these instructions in
3469 :     the fast-fp mode.
3470 :    
3471 : blume 2155 c. Added cases for JP and JNP in X86FreqProps.
3472 :    
3473 : leunga 1156 3. CFG
3474 :    
3475 :     CFG now has a bunch of methods for edge splitting and merging.
3476 :    
3477 :     4. Machine description.
3478 :    
3479 :     John's simplification of MLTREE_BASIS.fcond broke a few machine
3480 :     description things:
3481 :    
3482 :     rtl-build.{sig,sml} and hppa.mdl fixed.
3483 :    
3484 : blume 2155 NOTE: the machine description stuff in the repository is still broken.
3485 :     Again, I can't put my fixes in because that involves
3486 :     changes to MLTREE.
3487 : leunga 1156
3488 :     ----------------------------------------------------------------------
3489 : blume 1145 Name: Matthias Blume
3490 : blume 1155 Date: 2002/03/20 15:55:00 EST
3491 :     Tag: blume-20020320-kmp
3492 :     Description:
3493 :    
3494 :     Implemented Knuth-Morris-Pratt string matching in PreString and used
3495 :     it for String.isSubstring, Substring.isSubstring, and
3496 :     Substring.position.
3497 :    
3498 :     (Might need some stress-testing. Simple examples worked fine.)
3499 :    
3500 :     ----------------------------------------------------------------------
3501 :     Name: Matthias Blume
3502 : blume 1152 Date: 2002/03/19 16:37:00 EST
3503 :     Tag: blume-20020319-witnesses
3504 :     Description:
3505 :    
3506 :     Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
3507 :    
3508 :     This implements a generic mechanism for changing constness qualifiers
3509 :     anywhere within big C types without resorting to outright "casts".
3510 :     (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
3511 :     modify the constness at the outermost level.)
3512 :     The implementation of "convert" is based on the idea of "witness"
3513 :     values -- values that are not used by the operation but whose types
3514 :     "testify" to their applicability. On the implementation side, "convert"
3515 :     is simply a projection (returning its second curried argument). With
3516 :     cross-module inlining, it should not result in any machine code being
3517 :     generated.
3518 :    
3519 :     ----------------------------------------------------------------------
3520 :     Name: Matthias Blume
3521 : blume 1148 Date: 2002/03/15 16:40:00 EST
3522 :     Tag: blume-20020315-basis
3523 :     Description:
3524 :    
3525 :     Provided (preliminary?) implementations for
3526 :    
3527 :     {String,Substring}.{concatWith,isSuffix,isSubstring}
3528 :    
3529 :     and
3530 :    
3531 :     Substring.full
3532 :    
3533 :     Those are in the Basis spec but they were missing in SML/NJ.
3534 :    
3535 :     ----------------------------------------------------------------------
3536 :     Name: Matthias Blume
3537 : blume 1145 Date: 2002/03/14 21:30:00 EST
3538 :     Tag: blume-20020314-controls
3539 :     Description:
3540 :    
3541 :     Controls:
3542 :     ---------
3543 :    
3544 :     1. Factored out the recently-added Controls : CONTROLS stuff and put
3545 :     it into its own library $/controls-lib.cm. The source tree for
3546 :     this is under src/smlnj-lib/Controls.
3547 :    
3548 :     2. Changed the names of types and functions in this interface, so they
3549 :     make a bit more "sense":
3550 :    
3551 :     module -> registry
3552 :     'a registry -> 'a group
3553 :    
3554 :     3. The interface now deals in ref cells only. The getter/setter interface
3555 :     is (mostly) gone.
3556 :    
3557 :     4. Added a function that lets one register an already-existing ref cell.
3558 :    
3559 :     5. Made the corresponding modifications to the rest of the code so that
3560 :     everything compiles again.
3561 :    
3562 :     6. Changed the implementation of Controls.MLRISC back to something closer
3563 :     to the original. In particular, this module (and therefore MLRISC)
3564 :     does not depend on Controls. There now is some link-time code in
3565 :     int-sys.sml that registers the MLRISC controls with the Controls
3566 :     module.
3567 :    
3568 :     CM:
3569 :     ---
3570 :    
3571 :     * One can now specify the lambda-split aggressiveness in init.cmi.
3572 :    
3573 :     ----------------------------------------------------------------------
3574 : leunga 1142 Name: Allen Leung
3575 :     Date: 2002/03/13 17:30:00 EST
3576 :     Tag: leunga-20020313-x86-fp-unary
3577 :     Description:
3578 :    
3579 :     Bug fix for:
3580 :    
3581 :     > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
3582 :     > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
3583 :     > - fun f(x,(y,z)) = Real.~ y;
3584 :     > [autoloading]
3585 :     > [autoloading done]
3586 :     > fchsl (%eax), 184(%esp)
3587 :     > Error: MLRisc bug: X86MCEmitter.emitInstr
3588 : blume 2155 >
3589 : leunga 1142 > uncaught exception Error
3590 :     > raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
3591 :    
3592 :     The problem was that the code generator did not generate any fp registers
3593 :     in this case, and the ra didn't know that it needed to run the X86FP phase to
3594 :     translate the pseudo fp instruction. This only happened with unary fp
3595 :     operators in certain situations.
3596 :    
3597 :     ----------------------------------------------------------------------
3598 : blume 1137 Name: Matthias Blume
3599 : blume 1140 Date: 2002/03/13 14:00:00 EST
3600 :     Tag: blume-20020313-overload-etc
3601 :     Description:
3602 :    
3603 :     1. Added _overload as a synonym for overload for backward compatibility.
3604 :     (Control.overloadKW must be true for either version to be accepted.)
3605 :    
3606 :     2. Fixed bug in install script that caused more things to be installed
3607 :     than what was requested in config/targets.
3608 :    
3609 :     3. Made CM aware of the (_)overload construct so that autoloading
3610 :     works.
3611 :    
3612 :     ----------------------------------------------------------------------
3613 :     Name: Matthias Blume
3614 : blume 1138 Date: 2002/03/12 22:03:00 EST
3615 :     Tag: blume-20020312-url
3616 :     Description:
3617 :    
3618 :     Forgot to update BOOT and srcarchiveurl.
3619 :    
3620 :     ----------------------------------------------------------------------
3621 :     Name: Matthias Blume
3622 : blume 1137 Date: 2002/03/12 17:30:00 EST
3623 :     Tag: blume-20020312-version110392
3624 :     Description:
3625 :    
3626 :     Yet another version number bump (because of small changes to the
3627 :     binfile format). Version number is now 110.39.2. NEW BOOTFILES!
3628 :    
3629 :     Changes:
3630 :    
3631 :     The new pid generation scheme described a few weeks ago was overly
3632 :     complicated. I implemented a new mechanism that is simpler and
3633 :     provides a bit more "stability": Once CM has seen a compilation
3634 :     unit, it keeps its identity constant (as long as you do not delete
3635 :     those crucial CM/GUID/* files). This means that when you change
3636 : blume 1145 an interface, compile, then go back to the old interface, and
3637 : blume 1137 compile again, you arrive at the original pid.
3638 :    
3639 :     There now also is a mechanism that instructs CM to use the plain
3640 :     environment hash as a module's pid (effectively making its GUID
3641 :     the empty string). For this, "noguid" must be specified as an
3642 :     option to the .sml file in question within its .cm file.
3643 :     This is most useful for code that is being generated by tools such
3644 :     as ml-nlffigen (because during development programmers tend to
3645 :     erase the tool's entire output directory tree including CM's cached
3646 :     GUIDs). "noguid" is somewhat dangerous (since it can be used to locally
3647 :     revert to the old, broken behavior of SML/NJ, but in specific cases
3648 :     where there is no danger of interface confusion, its use is ok
3649 :     (I think).
3650 :    
3651 :     ml-nlffigen by default generates "noguid" annotations. They can be
3652 :     turned off by specifying -guid in its command line.
3653 :    
3654 :     ----------------------------------------------------------------------
3655 : george 1136 Name: Lal George
3656 :     Date: 2002/03/12 12 14:42:36 EST
3657 :     Tag: george-20020312-frequency-computation
3658 :     Description:
3659 : blume 2155
3660 : george 1136 Integrated jump chaining and static block frequency into the
3661 :     compiler. More details and numbers later.
3662 : leunga 1025
3663 : george 1003 ----------------------------------------------------------------------
3664 : george 1133 Name: Lal George
3665 :     Date: 2002/03/11 11 22:38:53 EST
3666 :     Tag: george-20020311-jump-chain-elim
3667 :     Description:
3668 :    
3669 :     Tested the jump chain elimination on all architectures (except the
3670 :     hppa). This is on by default right now and is profitable for the
3671 :     alpha and x86, however, it may not be profitable for the sparc and ppc
3672 :     when compiling the compiler.
3673 :    
3674 : blume 2155 The gc test will typically jump to a label at the end of the cluster,
3675 :     where there is another jump to an external cluster containing the actual
3676 : george 1133 code to invoke gc. This is to allow factoring of common gc invocation
3677 :     sequences. That is to say, we generate:
3678 :    
3679 :     f:
3680 :     testgc
3681 :     ja L1 % jump if above to L1
3682 :    
3683 :     L1:
3684 :     jmp L2
3685 :    
3686 :    
3687 :     After jump chain elimination the 'ja L1' instructions is converted to
3688 :     'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
3689 :     up being implemented in their long form (if L2 is far away) using:
3690 :    
3691 :     jbe L3 % jump if below or equal to L3
3692 :     jmp L2
3693 :     L3:
3694 :     ...
3695 :    
3696 :    
3697 :     For large compilation units L2 may be far away.
3698 :    
3699 :    
3700 :     ----------------------------------------------------------------------
3701 : blume 1132 Name: Matthias Blume
3702 :     Date: 2002/03/11 13:30:00 EST
3703 :     Tag: blume-20020311-mltreeeval
3704 :     Description:
3705 :    
3706 :     A functor parameter was missing.
3707 :    
3708 :     ----------------------------------------------------------------------
3709 : leunga 1130 Name: Allen Leung
3710 : leunga 1131 Date: 2002/03/11 10:30:00 EST
3711 : leunga 1142 Tag: leunga-20020311-runtime-string0
3712 : leunga 1131 Description:
3713 :    
3714 : blume 2155 The representation of the empty string now points to a
3715 :     legal null terminated C string instead of unit. It is now possible
3716 :     to convert an ML string into C string with InlineT.CharVector.getData.
3717 : leunga 1131 This compiles into one single machine instruction.
3718 :    
3719 :     ----------------------------------------------------------------------
3720 :     Name: Allen Leung
3721 : leunga 1130 Date: 2002/03/10 23:55:00 EST
3722 :     Tag: leunga-20020310-x86-call
3723 :     Description:
3724 :    
3725 :     Added machine generation for CALL instruction (relative displacement mode)
3726 :    
3727 :     ----------------------------------------------------------------------
3728 : blume 1128 Name: Matthias Blume
3729 :     Date: 2002/03/08 16:05:00
3730 :     Tag: blume-20020308-entrypoints
3731 :     Description:
3732 :    
3733 :     Version number bumped to 110.39.1. NEW BOOTFILES!
3734 :    
3735 :     Entrypoints: non-zero offset into a code object where execution should begin.
3736 :    
3737 :     - Added the notion of an entrypoint to CodeObj.
3738 :     - Added reading/writing of entrypoint info to Binfile.
3739 :     - Made runtime system bootloader aware of entrypoints.
3740 :     - Use the address of the label of the first function given to mlriscGen
3741 :     as the entrypoint. This address is currently always 0, but it will
3742 :     not be 0 once we turn on block placement.
3743 :     - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3744 :     with entry points) from mlriscGen.
3745 :    
3746 :     ----------------------------------------------------------------------
3747 : leunga 1127 Name: Allen Leung
3748 :     Date: 2002/03/07 20:45:00 EST
3749 :     Tag: leunga-20020307-x86-cmov
3750 :     Description:
3751 :    
3752 :     Bug fixes for CMOVcc on x86.
3753 :    
3754 :     1. Added machine code generation for CMOVcc
3755 :     2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3756 :     3. CMOVcc cannot have an immediate operand as argument.
3757 :    
3758 :     ----------------------------------------------------------------------
3759 : blume 1126 Name: Matthias Blume
3760 :     Date: 2002/03/07 16:15:00 EST
3761 :     Tag: blume-20020307-controls
3762 :     Description:
3763 :    
3764 :     This is a very large but mostly boring patch which makes (almost)
3765 :     every tuneable compiler knob (i.e., pretty much everything under
3766 :     Control.* plus a few other things) configurable via both the command
3767 :     line and environment variables in the style CM did its configuration
3768 :     until now.
3769 :    
3770 :     Try starting sml with '-h' (or, if you are brave, '-H')
3771 :    
3772 :     To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3773 :     implements the underlying generic mechanism.
3774 :    
3775 :     The interface to some of the existing such facilities has changed somewhat.
3776 :     For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3777 :     (The getFoo interface is still there for backward-compatibility, but its
3778 :     use is deprecated.)
3779 :    
3780 :     The ml-build script passes -Cxxx=yyy command-line arguments through so
3781 :     that one can now twiddle the compiler settings when using this "batch"
3782 :     compiler.
3783 :    
3784 :     TODO items:
3785 :    
3786 :     We should go through and throw out all controls that are no longer
3787 :     connected to anything. Moreover, we should go through and provide
3788 :     meaningful (and correct!) documentation strings for those controls
3789 :     that still are connected.
3790 :    
3791 :     Currently, multiple calls to Controls.new are accepted (only the first
3792 :     has any effect). Eventually we should make sure that every control
3793 :     is being made (via Controls.new) exactly once. Future access can then
3794 :     be done using Controls.acc.
3795 :    
3796 :     Finally, it would probably be a good idea to use the getter-setter
3797 :     interface to controls rather than ref cells. For the time being, both
3798 :     styles are provided by the Controls module, but getter-setter pairs are
3799 :     better if thread-safety is of any concern because they can be wrapped.
3800 :    
3801 :     *****************************************
3802 :    
3803 :     One bug fix: The function blockPlacement in three of the MLRISC
3804 :     backpatch files used to be hard-wired to one of two possibilities at
3805 :     link time (according to the value of the placementFlag). But (I
3806 :     think) it should rather sense the flag every time.
3807 :    
3808 :     *****************************************
3809 :    
3810 :     Other assorted changes (by other people who did not supply a HISTORY entry):
3811 :    
3812 :     1. the cross-module inliner now works much better (Monnier)
3813 :     2. representation of weights, frequencies, and probabilities in MLRISC
3814 :     changed in preparation of using those for weighted block placement
3815 :     (Reppy, George)
3816 :    
3817 :     ----------------------------------------------------------------------
3818 : george 1116 Name: Lal George
3819 : george 1124 Date: 2002/03/07 14:44:24 EST 2002
3820 :     Tag: george-20020307-weighted-block-placement
3821 :    
3822 :     Tested the weighted block placement optimization on all architectures
3823 :     (except the hppa) using AMPL to generate the block and edge frequencies.
3824 :     Changes were required in the machine properties to correctly
3825 : blume 2155 categorize trap instructions. There is an MLRISC flag
3826 :     "weighted-block-placement" that can be used to enable weighted block
3827 : george 1124 placement, but this will be ineffective without block/edge
3828 :     frequencies (coming soon).
3829 :    
3830 :    
3831 :     ----------------------------------------------------------------------
3832 :     Name: Lal George
3833 : blume 2155 Date: 2002/03/05 17:24:48 EST
3834 : george 1116 Tag: george-20020305-linkage-cluster
3835 :    
3836 : george 1124 In order to support the block placement optimization, a new cluster
3837 :     is generated as the very first cluster (called the linkage cluster).
3838 : blume 2155 It contains a single jump to the 'real' entry point for the compilation
3839 : george 1124 unit. Block placement has no effect on the linkage cluster itself, but
3840 :     all the other clusters have full freedom in the manner in which they
3841 : blume 2155 reorder blocks or functions.
3842 : george 1116
3843 :     On the x86 the typical linkage code that is generated is:
3844 :     ----------------------
3845 :     .align 2
3846 :     L0:
3847 :     addl $L1-L0, 72(%esp)
3848 : george 1124 jmp L1
3849 : george 1116
3850 :    
3851 :     .align 2
3852 :     L1:
3853 :     ----------------------
3854 :    
3855 : blume 2155 72(%esp) is the memory location for the stdlink register. This
3856 : george 1116 must contain the address of the CPS function being called. In the
3857 : george 1124 above example, it contains the address of L0; before
3858 : george 1116 calling L1 (the real entry point for the compilation unit), it
3859 :     must contain the address for L1, and hence
3860 :    
3861 :     addl $L1-L0, 72(%esp)
3862 :    
3863 : george 1124 I have tested this on all architectures except the hppa.The increase
3864 :     in code size is of course negligible
3865 : george 1116
3866 :     ----------------------------------------------------------------------
3867 : leunga 1115 Name: Allen Leung
3868 :     Date: 2002/03/03 13:20:00 EST
3869 :     Tag: leunga-20020303-mlrisc-tools
3870 :    
3871 :     Added #[ ... ] expressions to mlrisc tools
3872 :    
3873 :     ----------------------------------------------------------------------
3874 : blume 1096 Name: Matthias Blume
3875 : blume 1098 Date: 2002/02/27 12:29:00 EST
3876 :     Tag: blume-20020227-cdebug
3877 :     Description:
3878 :    
3879 :     - made types in structure C and C_Debug to be equal
3880 :     - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3881 :     - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3882 :    
3883 :     ----------------------------------------------------------------------
3884 :     Name: Matthias Blume
3885 : blume 1096 Date: 2002/02/26 12:00:00 EST
3886 :     Tag: blume-20020226-ffi
3887 :     Description:
3888 :    
3889 :     1. Fixed a minor bug in CM's "noweb" tool:
3890 :     If numbering is turned off, then truly don't number (i.e., do not
3891 :     supply the -L option to noweb). The previous behavior was to supply
3892 :     -L'' -- which caused noweb to use the "default" line numbering scheme.
3893 :     Thanks to Chris Richards for pointing this out (and supplying the fix).
3894 :    
3895 :     2. Once again, I reworked some aspects of the FFI:
3896 :    
3897 :     A. The incomplete/complete type business:
3898 :    
3899 :     - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3900 :     gone!
3901 :     - ML types representing an incomplete type are now *equal* to
3902 :     ML types representing their corresponding complete types (just like
3903 :     in C). This is still safe because ml-nlffigen will not generate
3904 :     RTTI for incomplete types, nor will it generate functions that
3905 :     require access to such RTTI. But when ML code generated from both
3906 :     incomplete and complete versions of the C type meet, the ML types
3907 :     are trivially interoperable.
3908 :    
3909 :     NOTE: These changes restore the full generality of the translation
3910 :     (which was previously lost when I eliminated functorization)!
3911 :    
3912 :     B. Enum types:
3913 :    
3914 :     - Structure C now has a type constructor "enum" that is similar to
3915 :     how the "su" constructor works. However, "enum" is not a phantom
3916 :     type because each "T enum" has values (and is isomorphic to
3917 :     MLRep.Signed.int).
3918 :     - There are generic access operations for enum objects (using
3919 :     MLRep.Signed.int).
3920 :     - ml-nlffigen will generate a structure E_foo for each "enum foo".
3921 :     * The structure contains the definition of type "mlrep" (the ML-side
3922 :     representation type of the enum). Normally, mlrep is the same
3923 :     as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3924 :     then mlrep will be defined as a datatype -- thus facilitating
3925 :     pattern matching on mlrep values.
3926 :     ("-ec" will be suppressed if there are duplicate values in an
3927 :     enumeration.)
3928 :     * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3929 :     will be generated for each C enum constant xxx.
3930 :     * Conversion functions m2i and i2m convert between mlrep and
3931 :     MLRep.Signed.int. (Without "-ec", these functions are identities.)
3932 :     * Coversion functions c and ml convert between mlrep and "tag enum".
3933 :     * Access functions (get/set) fetch and store mlrep values.
3934 :     - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3935 :     enumerations are merged into one single enumeration represented by
3936 :     structure E_'.
3937 :    
3938 :     ----------------------------------------------------------------------
3939 : leunga 1080 Name: Allen Leung
3940 : leunga 1094 Date: 2002/02/25 04:45:00 EST
3941 :     Tag: leunga-20020225-cps-spill
3942 :    
3943 :     This is a new implementation of the CPS spill phase.
3944 :     The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3945 :     In case of problems, replace it with the old file spill.sml
3946 :    
3947 : blume 2155 The current compiler runs into some serious performance problems when
3948 : leunga 1094 constructing a large record. This can happen when we try to compile a
3949 :     structure with many items. Even a very simple structure like the following
3950 :     makes the compiler slow down.
3951 :    
3952 :     structure Foo = struct
3953 :     val x_1 = 0w1 : Word32.int
3954 :     val x_2 = 0w2 : Word32.int
3955 :     val x_3 = 0w3 : Word32.int
3956 :     ...
3957 :     val x_N = 0wN : Word32.int
3958 :     end
3959 :    
3960 : blume 2155 The following table shows the compile time, from N=1000 to N=4000,
3961 : leunga 1094 with the old compiler:
3962 :    
3963 :     N
3964 : blume 2155 1000 CPS 100 spill 0.04u 0.00s 0.00g
3965 :     MLRISC ra 0.06u 0.00s 0.05g
3966 : leunga 1094 (spills = 0 reloads = 0)
3967 : blume 2155 TOTAL 0.63u 0.07s 0.21g
3968 : leunga 1094
3969 : blume 2155 1100 CPS 100 spill 8.25u 0.32s 0.64g
3970 :     MLRISC ra 5.68u 0.59s 3.93g
3971 : leunga 1094 (spills = 0 reloads = 0)
3972 : blume 2155 TOTAL 14.71u 0.99s 4.81g
3973 : leunga 1094
3974 : blume 2155 1500 CPS 100 spill 58.55u 2.34s 1.74g
3975 :     MLRISC ra 5.54u 0.65s 3.91g
3976 : leunga 1094 (spills = 543 reloads = 1082)
3977 : blume 2155 TOTAL 65.40u 3.13s 6.00g
3978 : leunga 1094
3979 : blume 2155 2000 CPS 100 spill 126.69u 4.84s 3.08g
3980 :     MLRISC ra 0.80u 0.10s 0.55g
3981 : leunga 1094 (spills = 42 reloads = 84)
3982 : blume 2155 TOTAL 129.42u 5.10s 4.13g
3983 : leunga 1094
3984 : blume 2155 3000 CPS 100 spill 675.59u 19.03s 11.64g
3985 :     MLRISC ra 2.69u 0.27s 1.38g
3986 : leunga 1094 (spills = 62 reloads = 124)
3987 : blume 2155 TOTAL 682.48u 19.61s 13.99g
3988 : leunga 1094
3989 : blume 2155 4000 CPS 100 spill 2362.82u 56.28s 43.60g
3990 :     MLRISC ra 4.96u 0.27s 2.72g
3991 : leunga 1094 (spills = 85 reloads = 170)
3992 : blume 2155 TOTAL 2375.26u 57.21s 48.00g
3993 : leunga 1094
3994 : blume 2155 As you can see the old cps spill module suffers from some serious
3995 :     performance problem. But since I cannot decipher the old code fully,
3996 :     instead of patching the problems up, I'm reimplementing it
3997 :     with a different algorithm. The new code is more modular,
3998 :     smaller when compiled, and substantially faster
3999 :     (O(n log n) time and O(n) space). Timing of the new spill module:
4000 : leunga 1094
4001 : blume 2155 4000 CPS 100 spill 0.02u 0.00s 0.00g
4002 :     MLRISC ra 0.25u 0.02s 0.15g
4003 :     (spills=1 reloads=3)
4004 :     TOTAL 7.74u 0.34s 1.62g
4005 : leunga 1094
4006 :     Implementation details:
4007 :    
4008 : blume 2155 As far as I can tell, the purpose of the CPS spill module is to make sure the
4009 :     number of live variables at any program point (the bandwidth)
4010 :     does not exceed a certain limit, which is determined by the
4011 :     size of the spill area.
4012 : leunga 1094
4013 : blume 2155 When the bandwidth is too large, we decrease the register pressure by
4014 : leunga 1094 packing live variables into spill records. How we achieve this is
4015 :     completely different than what we did in the old code.
4016 :    
4017 : blume 2155 First, there is something about the MLRiscGen code generator
4018 : leunga 1094 that we should be aware of:
4019 :    
4020 :     o MLRiscGen performs code motion!
4021 : blume 2155
4022 : leunga 1094 In particular, it will move floating point computations and
4023 : blume 2155 address computations involving only the heap pointer to
4024 :     their use sites (if there is only a single use).
4025 : leunga 1094 What this means is that if we have a CPS record construction
4026 :     statement
4027 : blume 2155
4028 : leunga 1094 RECORD(k,vl,w,e)
4029 : blume 2155
4030 :     we should never count the new record address w as live if w
4031 : leunga 1094 has only one use (which is often the case).
4032 : blume 2155
4033 : leunga 1094 We should do something similar to floating point, but the transformation
4034 :     there is much more complex, so I won't deal with that.
4035 :    
4036 :     Secondly, there are now two new cps primops at our disposal:
4037 :    
4038 : blume 2155 1. rawrecord of record_kind option
4039 : leunga 1094 This pure operator allocates some uninitialized storage from the heap.
4040 :     There are two forms:
4041 : blume 2155
4042 : leunga 1094 rawrecord NONE [INT n] allocates a tagless record of length n
4043 :     rawrecord (SOME rk) [INT n] allocates a tagged record of length n
4044 :     and initializes the tag.
4045 : blume 2155
4046 : leunga 1094 2. rawupdate of cty
4047 : blume 2155 rawupdate cty (v,i,x)
4048 : leunga 1094 Assigns to x to the ith component of record v.
4049 :     The storelist is not updated.
4050 : blume 2155
4051 : leunga 1094 We use these new primops for both spilling and increment record construction.
4052 : blume 2155
4053 : leunga 1094 1. Spilling.
4054 : blume 2155
4055 : leunga 1094 This is implemented with a linear scan algorithm (but generalized
4056 :     to trees). The algorithm will create a single spill record at the
4057 :     beginning of the cps function and use rawupdate to spill to it,
4058 :     and SELECT or SELp to reload from it. So both spills and reloads
4059 : blume 2155 are fine-grain operations. In contrast, in the old algorithm
4060 :     "spills" have to be bundled together in records.
4061 :    
4062 : leunga 1094 Ideally, we should sink the spill record construction to where
4063 :     it is needed. We can even split the spill record into multiple ones
4064 :     at the places where they are needed. But CPS is not a good
4065 :     representation for global code motion, so I'll keep it simple and
4066 :     am not attempting this.
4067 : blume 2155
4068 : leunga 1094 2. Incremental record construction (aka record splitting).
4069 :    
4070 :     Long records with many component values which are simulatenously live
4071 : blume 2155 (recall that single use record addresses are not considered to
4072 : leunga 1094 be live) are constructed with rawrecord and rawupdate.
4073 :     We allocate space on the heap with rawrecord first, then gradually
4074 :     fill it in with rawupdate. This is the technique suggested to me
4075 :     by Matthias.
4076 : blume 2155
4077 : leunga 1094 Some restrictions on when this is applicable:
4078 :     1. It is not a VECTOR record. The code generator currently does not handle
4079 :     this case. VECTOR record uses double indirection like arrays.
4080 : blume 2155 2. All the record component values are defined in the same "basic block"
4081 :     as the record constructor. This is to prevent speculative
4082 :     record construction.
4083 : leunga 1094
4084 :     ----------------------------------------------------------------------
4085 :     Name: Allen Leung
4086 : leunga 1086 Date: 2002/02/22 01:02:00 EST
4087 :     Tag: leunga-20020222-mlrisc-tools
4088 :    
4089 :     Minor bug fixes in the parser and rewriter
4090 :    
4091 :     ----------------------------------------------------------------------
4092 :     Name: Allen Leung
4093 : leunga 1085 Date: 2002/02/21 20:20:00 EST
4094 :     Tag: leunga-20020221-peephole
4095 :    
4096 :     Regenerated the peephole files. Some contained typos in the specification
4097 :     and some didn't compile because of pretty printing bugs in the old version
4098 :     of 'nowhere'.
4099 :    
4100 :     ----------------------------------------------------------------------
4101 :     Name: Allen Leung
4102 : leunga 1079 Date: 2002/02/19 20:20:00 EST
4103 :     Tag: leunga-20020219-mlrisc-tools
4104 :     Description:
4105 :    
4106 :     Minor bug fixes to the mlrisc-tools library:
4107 :    
4108 :     1. Fixed up parsing colon suffixed keywords
4109 :     2. Added the ability to shut the error messages up
4110 :     3. Reimplemented the pretty printer and fixed up/improved
4111 :     the pretty printing of handle and -> types.
4112 :     4. Fixed up generation of literal symbols in the nowhere tool.
4113 :     5. Added some SML keywords to to sml.sty
4114 :    
4115 :     ----------------------------------------------------------------------
4116 :     Name: Matthias Blume
4117 : blume 1078 Date: 2002/02/19 16:20:00 EST
4118 :     Tag: blume-20020219-cmffi
4119 :     Description:
4120 :    
4121 :     A wild mix of changes, some minor, some major:
4122 :    
4123 :     * All C FFI-related libraries are now anchored under $c:
4124 :     $/c.cm --> $c/c.cm
4125 :     $/c-int.cm --> $c/internals/c-int.cm
4126 :     $/memory.cm --> $c/memory/memory.cm
4127 :    
4128 :     * "make" tool (in CM) now treats its argument pathname slightly
4129 :     differently:
4130 :     1. If the native expansion is an absolute name, then before invoking
4131 :     the "make" command on it, CM will apply OS.Path.mkRelative
4132 :     (with relativeTo = OS.FileSys.getDir()) to it.
4133 :     2. The argument will be passed through to subsequent phases of CM
4134 :     processing without "going native". In particular, if the argument
4135 :     was an anchored path, then "make" will not lose track of that anchor.
4136 :    
4137 :     * Compiler backends now "know" their respective C calling conventions
4138 :     instead of having to be told about it by ml-nlffigen. This relieves
4139 :     ml-nlffigen from one of its burdens.
4140 :    
4141 :     * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
4142 :    
4143 :     * Export C_DEBUG and C_Debug from $c/c.cm.
4144 :    
4145 :     * C type encoding in ml-nlffi-lib has been improved to model the conceptual
4146 :     subtyping relationship between incomplete pointers and their complete
4147 :     counterparts. For this, ('t, 'c) ptr has been changed to 'o ptr --
4148 :     with the convention of instantiating 'o with ('t, 'c) obj whenever
4149 :     the pointer target type is complete. In the incomplete case, 'o
4150 :     will be instantiated with some "'c iobj" -- a type obtained by
4151 :     using one of the functors PointerToIncompleteType or PointerToCompleteType.
4152 :    
4153 :     Operations that work on both incomplete and complete pointer types are
4154 :     typed as taking an 'o ptr while operations that require the target to
4155 :     be known are typed as taking some ('t, 'c) obj ptr.
4156 :    
4157 :     voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
4158 :     where void is an eqtype without any values. This makes it possible
4159 :     to work on voidptr values using functions meant to operate on light
4160 :     incomplete pointers.
4161 :    
4162 :     * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
4163 :     been vastly simplified.
4164 :    
4165 :