Diff of /sml/trunk/HISTORY

Parent Directory | Revision Log | Patch

	revision 656, Fri Jun 9 03:39:04 2000 UTC	revision 837, Fri Jun 1 17:27:54 2001 UTC
#	Line 11	Line 11
11	Date:	Date:
12	Tag: <post-commit CVS tag>	Tag: <post-commit CVS tag>
13	Description:	Description:
14
15		----------------------------------------------------------------------
16		Name: Matthias Blume
17		Date: 2001/06/01 13:30:00 EDT
18		Tag: blume-20010601-nlffi-cleanup
19		Description:
20
21		This is mostly a cleanup of MLFFI stuff:
22
23		- some signature files have been put into a more exposed place
24		- the ugly 'f type parameter is gone (simplifies types tremendously!)
25		- ml-nlffigen changed accordingly
26		- tutorial updated
27
28		Other changes:
29
30		- author's affiliation in CM manual(s) updated
31		- some more recognized keywords added to Allen's sml.sty
32
33		----------------------------------------------------------------------
34		Name: Matthias Blume
35		Date: 2001/05/25 15:30:00 EDT
36		Tag: blume-20010525-iptr
37		Description:
38
39		- put the official 110.33-README (as it appears on the ftp server) under
40		CVS
41		- fixed a small bug related to incomplete pointer types in
42		ml-nlffigen
43		- small cosmetic change to the ml-nlffi-lib's "arr" type constructor
44		(it does not need the 'f type parameter)
45
46		----------------------------------------------------------------------
47		Name: Matthias Blume
48		Date: 2001/05/23 14:30:00 EDT
49		Tag: Release_110_33
50		Description:
51
52		New version number (110.33).  New bootfiles.
53
54		----------------------------------------------------------------------
55		Name: Matthias Blume
56		Date: 2001/05/22 18:06:00 EDT
57		Tag: blume-20010522-targets
58		Description:
59
60		Made install.sh use file config/targets.customized if it exists, falling
61		back to config/targets if it doesn't.  This way one can have a customized
62		version of the targets file without touching the "real thing", thus
63		eliminating the constant fear of accidentally checking something bogus
64		back into the CVS repository...  (File config/targets.customized must
65		not be added to the repository!)
66
67		----------------------------------------------------------------------
68		Name: Matthias Blume
69		Date: 2001/05/22 16:30:00 EDT
70		Tag: blume-20010522-minitut
71		Description:
72
73		1. Bug fix in ml-nlffigen; now (hopefully) correctly handling
74		struct returns.
75		2. Added src/ml-nlffi-lib/Doc/mini-tutorial.txt.  This is some very
76		incomplete, preliminary documentation for NLFFI.
77
78		----------------------------------------------------------------------
79		Name: Matthias Blume
80		Date: 2001/05/14 11:30:00 EDT
81		Tag: blume-20010514-script
82		Description:
83
84		Some bugs in install script fixed.
85
86		In addition to that I also made a slight change to the NLFFI API:
87		Functors generated by ml-nlffigen now take the dynamic library as a
88		straight functor argument, not as a suspended one.  (The original
89		functor code used to force the suspension right away anyway, so there
90		was nothing gained by this complication of the interface.)
91
92		----------------------------------------------------------------------
93		Name: Matthias Blume
94		Date: 2001/05/11 14:35:00 EDT
95		Tag: blume-20010511-ml-nlffi
96		Description:
97
98		I finally took the plunge and added my new FFI code to the main
99		repository.  For x86-linux it is now ready for prime-time.
100
101		There are two new subdirectories of "src":
102
103		- ml-nlffi-lib:
104		The utility library for programs using the FFI interface.
105		Here is the implementation of $/c.cm and its associated low-level
106		partners $/c-int.cm and $/memory.cm.
107		- ml-nlffigen:
108		A stand-alone program for generating ML glue code from C source
109		code.
110
111		Building ml-nlffigen requires $/ckit-lib.cm.
112
113		The config/install.sh script has been updates to do the Right Thing
114		(hopefully).
115
116		Notice that the source tree for the C-Kit will not be put under "src"
117		but directly under the installation root directory.  (This is the
118		structure that currently exists on the CVS server when you check out
119		module "sml".)  Fortunately, config/install.sh knows about this oddity.
120
121		Bugs: No documentation yet.
122
123		----------------------------------------------------------------------
124		Name: Matthias Blume
125		Date: 2001/05/09 16:35:00 EDT
126		Tag: blume-20010509-cpscontract
127		Description:
128
129		Fixed a bug in the accounting code in cpsopt/contract.sml.  (The
130		wrapper/unwrapper elimination did not decrement usage counts and some
131		dead variables got overlooked by the dead-up logic.)
132
133		----------------------------------------------------------------------
134		Name: Lal George
135		Date: 2001/05/08  17:26:09 EDT
136		Tag: george-20010508-omit-frameptr
137		Description:
138
139		Changes to implement the omit-frame-pointer optimization to support
140		raw C calls. For now, there is only support on the Intel x86, but
141		other architectures will follow as more experience is gained with this.
142
143
144		----------------------------------------------------------------------
145		Name: Matthias Blume
146		Date: 2001/05/07 14:40:00 EDT
147		Tag: blume-20010507-proxies
148		Description:
149
150		I made into "proxy libraries" all libraries that qualify for such a
151		change.  (A qualifying library is a library that has another library or
152		groups as its sole member and repeats that member's export list
153		verbatim.  A proxy library avoids this repetition by omitting its export
154		list, effectively inheriting the list that its (only) member exports.
155		See the CM manual for more explanation.)
156		The main effect is that explicit export lists for these libraries
157		do not have to be kepts in sync, making maintenance a bit easier.
158
159		I also added copyright notices to many .cm-files.
160
161		Last but not least, I made a new set of bootfiles.
162
163		----------------------------------------------------------------------
164		Name: Matthias Blume
165		Date: 2001/05/04 17:00:00 EDT
166		Tag: blume-20010504-cm-lsplit
167		Description:
168
169		0. John merged pending changes to $/smlnj-lib.cm
170
171		1. Allen's previous change accidentally backed out of one of Lal's
172		earlier changes.  I undid this mistake (re-introducing Lal's change).
173
174		2. I used the new topOrder' function from graph-scc.sml (from $/smlnj-lib.cm)
175		within the compiler where applicable.  There is some code simplification
176		because of that.
177
178		3. The "split" phase (in FLINT) is now part of the default list of phases.
179		Compiler.Control.LambdaSplitting.* can be used to globally control the
180		lambda-splitting (cross-module-inlining) engine.  In addition to that,
181		it can now also be controlled on a per-source basis: CM has been taught
182		a new tool parameter applicable to ML source files.
183
184		- To turn lambda-splitting off completely:
185		local open Compiler.Control.LambdaSplitting in
186		val _ = set Off
187		end
188		- To make "no lambda-splitting" the global default (but allow per-source
189		overriding); this is the initial setting:
190		local open Compiler.Control.LambdaSplitting in
191		val _ = set (Default NONE)
192		end
193		- To make "lambda-splitting with aggressiveness a" the global default
194		(and allow per-source overriding):
195		local open Compiler.Control.LambdaSplitting in
196		val _ = set (Default (SOME a))
197		end
198
199		- To turn lambda-splitting off for a given ML souce file (say: a.sml)
200		write (in the respective .cm-file):
201		a.sml (lambdasplitting:off)
202		- To turn lambda-splitting for a.sml on with minimal aggressiveness:
203		a.sml (lambdasplitting:on)
204		- To turn lambda-splitting for a.sml on with aggressiveness <a> (where
205		<a> is a decimal non-negative integer):
206		a.sml (lambdasplitting:<a>)
207		- To turn lambda-splitting for a.sml on with maximal aggressiveness:
208		a.sml (lambdasplitting:infinity)
209		- To use the global default for a.sml:
210		a.sml (lambdasplitting:default)
211		or simply
212		a.sml
213
214		----------------------------------------------------------------------
215		Name: Allen Leung
216		Date: 2001/05/04 01:57:00 EDT
217		Tag: leunga-20010504-sync
218		Description:
219
220		MLRISC features.
221
222		1. Fix to CMPXCHG instructions.
223		2. Changed RA interface to allow annotations in callbacks.
224		3. Added a new method to the stream interface to allow annotations updates.
225
226		----------------------------------------------------------------------
227		Name: Matthias Blume
228		Date: 2001/05/01 11:45:00 EDT
229		Tag: blume-20010501-pcedittmp
230		Description:
231
232		Changed install.sh to use the current working directory instead of
233		/usr/tmp for a temporary file (pcedittmp).  The previous choice
234		of /usr/tmp caused trouble with MacOS X because of file premission
235		problems.
236
237		----------------------------------------------------------------------
238		Name: Matthias Blume
239		Date: 2001/04/20 11:10:00 EDT
240		Tag: blume-20010420-inMLflag
241		Description:
242
243		- added vp_limitPtrMask to vproc-state.h
244		(for use by the raw-C-calls mechanism to implement proper interrupt
245		handling)
246		- made the ML compiler aware of various data-structure offsets so it
247		can generate code for accessing the vp_inML flag and vp_limitPtrMask
248		- tweaked mlriscGen.sml to have it emit interrupt-handling code for
249		raw C-calls
250
251		----------------------------------------------------------------------
252		Name: Lal George
253		Date: 2001/04/20 09:15:28 EDT
254		Tag: george-20010420-macosX
255		Description:
256
257		- Changes to port to Mac OS X; Darwin.
258
259		- In the process I found that sqrt was broken on the PPC, because the
260		fsqrt instruction is not implemented.
261
262		----------------------------------------------------------------------
263		Name: Matthias Blume
264		Date: 2001/04/18 12:45:00 EDT
265		Tag: blume-20010418-ccalls
266		Description:
267
268		- fixed two off-by-4 errors in the x86-specific c-calls implementation
269		(this bug prevented structure arguments containing pointers from being
270		passed correctly)
271		- changed the raw-C-call code in mlriscGen.sml in such a way that
272		structure arguments are represented as a pointer to the beginning
273		of the structure (instead of having a series of synthesized arguments,
274		one for each structure member)
275
276		- made makeml script's verbosity level configurable via environment
277		variable (MAKEML_VERBOSITY)
278
279		- eliminated placeholder implementations for f32l, w16s, i16s, and f32s
280		in rawmem-x86.sml; we are now using the real thing
281
282		----------------------------------------------------------------------
283		Name: Matthias Blume
284		Date: 2001/03/22 16:25:00 EST
285		Tag: blume-20010322-bootfiles
286		Description:
287
288		Created a new set of bootfiles (for your automatic installation convenience).
289
290		----------------------------------------------------------------------
291		Name: Matthias Blume
292		Date: 2001/03/22 15:10:00 EST
293		Tag: blume-20010322-rawmem-parcm
294		Description:
295
296		1. All "raw memory access" primitives for the new FFI are implemented now
297		(at least on the x86).
298		2. Some further cleanup of CM's parallel make mechanism.
299
300		----------------------------------------------------------------------
301		Name: Matthias Blume
302		Date: 2001/03/19 17:53:00 EST
303		Tag: blume-20010319-parallel
304		Description:
305
306		Parallel make (using compile servers) now works again.
307
308		To this end, CM.stabilize and CMB.make have been modified to work in
309		two passes when compile servers are attached:
310		1. Compile everything, do not perform stabilization; this pass
311		uses compile servers
312		2. Stabilize everything; this pass does not use compile servers
313		If there are no compile servers, the two passes are combined into one
314		(as before).  Splitting the passes increases the inherent parallelism
315		in the dependency graph because the entire graph including all
316		libraries is available at the same time.  This, in turn, improves
317		server utilization.  The downside is that the master process will
318		have to do some extra work after compilation is done (because for
319		technical reasons it must re-read all the binfiles during stabilization).
320
321		----------------------------------------------------------------------
322		Name: Matthias Blume
323		Date: 2001/03/16 12:22:00 EST
324		Tag: blume-20010316-bootfiles
325		Description:
326
327		Created a new set of bootfiles (for your automatic installation convenience).
328
329		----------------------------------------------------------------------
330		Name: Matthias Blume
331		Date: 2001/03/16 11:00:00 EST
332		Tag: blume-20010316-MLTREE-fixup
333		Description:
334
335		This is a minor fixup for an (untagged) earlier commit by Allen.
336		(A file was missing).
337
338		----------------------------------------------------------------------
339		Name: Allen Leung
340		Date: Mon Mar  5 18:54:57 EST 2001
341		Tag: leunga-20010305-cut-support
342
343		1. New support for alternative control-flow in MLTREE.
344		Currently we support
345
346		FLOW_TO(CALL ...., [k1,...,kn])
347
348		This is needed for 'cuts to' in C-- and try/handle-like constructs
349		in Moby
350
351		New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
352
353		2. Register Allocator
354
355		Changes in interface [from Fermin, John]
356
357		3. Alpha 8-bit SLL support [Fermin]
358
359		4. All architectures
360
361		A new module (ClusterExpandCopies) for expanding parallel copies.
362
363		----------------------------------------------------------------------
364		Name: Allen Leung
365		Date: 2001/02/27 23:07:00 EST
366		Tag: leunga-20010227-minor-stuff
367
368		1. Alpha bug fix for CMOVNE
369		2. Handle mltree COND(..,FCMP ...,...)
370		3. Bug fix in simplifier
371
372		----------------------------------------------------------------------
373		Name: Matthias Blume
374		Date: 2001/01/30 17:50:00 EST
375		Tag: blume-20010130-sync
376		Description:
377
378		This is just a minor update to sync my devel branch with the main brach.
379		The only visible change is the addition of some README files.
380
381		----------------------------------------------------------------------
382		Name: Matthias Blume
383		Date: 2001/01/12 23:30:00 JST
384		Tag: blume-20010112-bootfiles
385		Description:
386
387		Made a new set of bootfiles that goes with the current state of the
388		repository.
389
390		----------------------------------------------------------------------
391		Name: Matthias Blume
392		Date: 2001/01/12 21:20:00 JST
393		Tag: blume-20010112-sync
394		Description:
395
396		I am just flushing out some minor changes that had accumulated in
397		my private branch in order to sync with the main tree.  (This is
398		mainly because I had CVS trouble when trying to merge _into_ my
399		private branch.)
400
401		Most people should be completely unaffected by this.
402
403		----------------------------------------------------------------------
404		Name: Allen Leung
405		Date: Thu Jan 11 21:03:00 EST 2001
406		Tag: leunga-20010111-labexp=mltree
407		Description:
408
409		1.  Removed the type LabelExp and replace it by MLTree.
410		2.  Rewritten mltree-simplify with the pattern matcher tool.
411		3.  There were some bugs in alpha code generator which would break
412		64-bit code generation.
413		4.  Redo the tools to generate code with the
414		5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
415		are now generated by perl scripts.
416
417		----------------------------------------------------------------------
418		Name: Matthias Blume
419		Date: 2001/01/10 21:55:00 JST
420		Tag: blume-20010110-rcc
421		Description:
422
423		The RCC stuff now seems to work (but only on the x86).
424		This required hacking of the c-calls interface (and -implementation) in
425		MLRISC.
426
427		Normal compiler users should be unaffected.
428
429		----------------------------------------------------------------------
430		Name: Matthias Blume
431		Date: 2001/01/09 01:20:00 JST
432		Tag: blume-20010109-rcc
433		Description:
434
435		This is a fairly big patch, flushing out a large number of pending
436		changes that I made to my development copy over the last couple of days.
437
438		Of practical relevance at this moment is a workaround for a pickling
439		bug that Allen ran into the other day.  The cause of the bug itself is
440		still unknown and it might be hard to fix it properly, but the
441		workaround has some merits of its own (namely somewhat reducing pickling
442		overhead for certain libraries).  Therefore, I think this solution should
443		be satisfactory at this time.
444
445		The rest of the changes (i.e., the vast majority) has to do with my
446		ongoing efforts of providing direct support for C function calls from
447		ML.  At the moment there is a new primop "RAW_CCALL", typing magic
448		in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
449		case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
450		translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
451		handle RCC.
452
453		The last part (the changes to mlriscGen.sml) are still known to be
454		wrong on the x86 and not implemented on all other architectures.  But
455		the infrastructure is in place. I had to change a few functor
456		signatures in the backend to be able to route the CCalls interface
457		from MLRISC there, and I had to specialize the mltree type (on the
458		x86) to include the necessary extensions. (The extensions themselves
459		were already there and redy to go in MLRISC/x86).
460
461		Everything should be very happy as soon as someone helps me with
462		mlriscGen.sml...
463
464		In any case, nothing of this should matter to anyone as long as the
465		new primop is not being used (which is going to be the case unless you
466		find it where I hid it :). The rest of the compiler is completely
467		unaffected.
468
469		----------------------------------------------------------------------
470		Name: Matthias Blume
471		Date: 2001/01/05 00:30:00 JST
472		Tag: blume-20010105-primops
473		Description:
474
475		Added some experimental support for work that I am doing right now.
476		These changes mostly concern added primops, but there is also a new
477		experimental C library in the runtime system (but currently not enabled
478		anywhere except on Linux/X86).
479
480		In the course of adding primops (and playing with them), I discovered that
481		Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
482		broken.  (Zhong was very right he labeled this stuff as "major gross hack".)
483		To recover, I made type information in INL_PRIM mandatory and changed
484		prim.sml as well as built-in.sml accordingly.  The InLine structure now
485		has complete, correct type information (i.e., no bottom types).
486
487		Since all these changes mean that we need new binfiles, I also bumped the
488		version number to 110.32.1.
489
490		----------------------------------------------------------------------
491		Name: Matthias Blume
492		Date: 2000/12/30 22:10:00 JST
493		Tag: blume-20001230-various
494		Description:
495
496		Added proxy libraries for MLRISC and let MLRISC libraries refer
497		to each other using path anchors.  (See CM manual for explanation.)
498
499		Updated CM documentation.
500
501		Fixed some bugs in CM.
502
503		Implemented "proxy" libraries (= syntactic sugar for CM).
504
505		Added "-quiet" option to makeml and changed runtime system accordingly.
506
507		Added cleanup handler for exportML to reset timers and compiler stats.
508
509		----------------------------------------------------------------------
510		Name: Lal George
511		Date: 2000/12/22 22:22:58 EST 2000
512		Tag: Release_110_32
513		Description:
514
515		Infinite precision used throughout MLRISC.
516		see MLRISC/mltree/machine-int.sig
517
518		----------------------------------------------------------------------
519		Name: Matthias Blume
520		Date: 2000/12/22 23:16:00 JST
521		Tag: blume-20001222-warn
522		Description:
523
524		Corrected wording and formatting of some CM warning message which I
525		broke in my previous patch.
526
527		----------------------------------------------------------------------
528		Name: Matthias Blume
529		Date: 2000/12/22 21:20:00 JST
530		Tag: blume-20001222-anchorenv
531		Description:
532
533		Fixed CM's handling of anchor environments in connection with CMB.make.
534
535		----------------------------------------------------------------------
536		Name: Matthias Blume
537		Date: 2000/12/22 13:15:00 JST
538		Tag: blume-20001222-cleanup
539		Description:
540
541		Removed src/cm/ffi which does not (and did not) belong here.
542
543		----------------------------------------------------------------------
544		Name: Matthias Blume
545		Date: 2000/12/21 23:55:00 JST
546		Tag: blume-20001221-exn
547		Description:
548
549		Probably most important: CM no longer silently swallows all exceptions
550		in the compiler.
551		Plus: some other minor CM changes.  For example, CM now reports some
552		sizes for generated binfiles (code, data, envpickle, lambdapickle).
553
554		----------------------------------------------------------------------
555		Name: Matthias Blume
556		Date: 2000/12/15 00:01:05 JST
557		Tag: blume-20001215-dirtool
558		Description:
559
560		- "dir" tool added.
561		- improvements and cleanup to Tools structure
562		- documentation updates
563
564		----------------------------------------------------------------------
565		Name: Allen Leung
566		Date: Thu Dec 14 03:45:24 EST 2000
567		Description:
568		Tag:  leunga-20001214-int-inf
569		Description:
570
571		In IntInf, added these standard functions, which are missing from our
572		implementation:
573
574		andb : int * int -> int
575		xorb : int * int -> int
576		orb  : int * int -> int
577		notb : int -> int
578		<<   : int * word -> int
579		~>>  : int * word -> int
580
581		Not tested, I hope they are correct.
582
583		----------------------------------------------------------------------
584		Name: Allen Leung
585		Date: Fri Dec  8 19:23:26 EST 2000
586		Description:
587		Tag:  leunga-20001208-nowhere
588		Description:
589
590		Slight improvements to the 'nowhere' tool to handle OR-patterns,
591		to generate better error messages etc.  Plus a brief manual.
592
593		----------------------------------------------------------------------
594		Name: Lal George
595		Date: 2000/12/08 09:54:02 EST 2000
596		Tag: Release_110_31
597		Description:
598
599		- Version 110.31
600		----------------------------------------------------------------------
601		Name: Allen Leung
602		Date: Thu Dec  7 22:01:04 EST 2000
603		Tag:  leunga-20001207-cell-monster-hack
604		Description:
605
606		Major MLRISC internal changes.  Affect all clients.
607		Summary:
608
609		1.  Type CELLS.cell = int is now replaced by a datatype.
610		As a result, the old regmap is now gone.  Almost all interfaces
611		in MLRISC change as a consequence.
612
613		2.  A new brand version of machine description tool (v3.0) that generates
614		modules expecting the new interface.  The old version is removed.
615
616		3.  The RA interface has been further abstracted into two new functors.
617		RISC_RA and X86RA.  These functors have much simpler interfaces.
618		[See also directory MLRISC/demo.]
619
620		4.  Some other new source->source code generation tools are available:
621
622		a. MLRISC/Tools/RewriteGen -- generate rewriters from rules.
623		b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules.
624		I use this tool to generate the peephole optimizers---with the new
625		cell type changes, peephole rules are becoming difficult to write
626		without conditional pattern matching.
627
628		5.  More Intmap -> IntHashTable change.  Previous changes by Matthias didn't
629		cover the entire MLRISC source tree so many things broke.
630
631		6.  CM files have been moved to the subdirectory MLRISC/cm.
632		They are moved because there are a lot of them and they clutter up the
633		root dir.
634
635		7.  More detailed documentation to come...
636
637		NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do
638		a makeml -rebuild first.  This is because of other other
639		changes that Matthias has made (see below).
640
641
642		----------------------------------------------------------------------
643		Name: Matthias Blume
644		Date: 2000/11/30 23:12:00 JST
645		Tag: blume-20001130-filereorg
646		Description:
647
648		Some manual updates and some file reorganizations in CM.
649
650		----------------------------------------------------------------------
651		Name: Matthias Blume
652		Date: 2000/11/24 17:45:00 JST
653		Tag: blume-20001124-link
654		Description:
655
656		Drastically improved link traversal code for the case that the dynamic
657		value was already loaded at bootstrap time.  As a result, CM and CMB
658		now both load blazingly fast -- even on a very slow machine.  Also,
659		memory consumption has been further reduced by this.
660
661		Warning: The format of the PIDMAP file has changed.  THerefore, to
662		bootstrap you have to do this:
663
664		1. Run CMB.make
665		2. Make a symbolic link for the boot directory:
666		ln -s sml.boot.ARCH-OS xxx
667		3. "Rebuild" the boot directory:
668		./makeml -boot xxx -rebuild sml ; rm xxx
669		4. Boot normally:
670		./makeml
671
672		----------------------------------------------------------------------
673		Name: Matthias Blume
674		Date: 2000/11/21 21:20:00 JST
675		Tag: blume-20001121-tools
676		Description:
677
678		Continued hacking on autoloading problem -- with success this time.
679		Also changed tool-plugin mechanism.  See new CM manual.
680
681		----------------------------------------------------------------------
682		Name: Matthias Blume
683		Date: 2000/11/19 14:30:00 JST
684		Tag:  blume-20001119-autoload
685		Description:
686
687		Some hacking to make autoloading faster.  Success for CMB, no success
688		so far for CM.  There is a reduced structure CM' that autoloads faster.
689		(This is a temporary, non-documented hack to be eliminated again when
690		the general problem is solved.)
691
692		----------------------------------------------------------------------
693		Name: Matthias Blume
694		Date: 2000/11/17 14:10:00 JST
695		Tag: blume-20001117-pickle-lib
696		Description:
697
698		1. Eliminated comp-lib.cm
699		2. Made pickle-lib.cm
700		3. Eliminated all uses of intset.sml (from comp-lib.cm)
701		4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
702		equivalent constructs from smlnj-lib.cm (INtHashTable).
703		5. Point 4. also goes for those uses of intmap.* in MLRISC.
704		Duplicated intmap modules thrown out.
705		6. Hunted down all duplicated SCC code and replaced it with
706		equivalent stuff (GraphSCCFn from smlnj-lib.cm).
707		7. Rewrote Feedback module.
708		8. Moved sortedlist.sml into viscomp-lib.cm.  Eventually it
709		should be thrown out and equivalent modules from smlnj-lib.cm
710		should be used (IntRedBlackSet, IntListSet, ...).
711
712		Confirmed that compiler compiles to fixpoint.
713
714		----------------------------------------------------------------------
715		Name: Allen Leung
716		Date: 2000/11/10 18:00:00
717		Tag: leunga-20001110-new-x86-fp
718
719		A new x86 floating point code generator has been added.
720		By default this is turned off.  To turn this on, do:
721
722		CM.autoload "$smlnj/compiler.cm";
723		Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
724
725		Changes:
726
727		1.  Changed FTAN to FPTAN so that the assembly output is correct.
728		2.  Changed the extension callback for FTANGENT to generate:
729
730		fptan
731		fstp  %st(0)
732		instead of
733		fptan
734		fstpl ftempmem
735
736		3.  Numerous assembly fixes for x86.
737
738		5.  Cleaned up the machine code output module x86/x86MC.sml and added
739		support for a whole bunch of instructions and addressing modes:
740
741		fadd/fsub/fsubr/fmul/fdiv/fdivr  %st, %st(n)
742		faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp  %st, %st(n)
743		fadd/fsub/fsubr/fmul/fdiv/fdivr  %st(n), %st
744		fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
745		fxch %st(n)
746		fld %st(n)
747		fst %st(n)
748		fst mem
749		fstp %st(n)
750		fucom %st(n)
751		fucomp %st(n)
752
753		All these are now generated when the fast fp mode is turned on.
754
755		6.  Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
756
757		----------------------------------------------------------------------
758		Name: Matthias Blume
759		Date: 2000/11/09 11:20:00 JST
760		Tag: blume-20001109-scc
761		Description:
762
763		Eliminated some code duplication:
764
765		1. Added "where" clause to GraphSCCFn in SML/NJ Library.
766		(Otherwise the functor is useless.)
767		2. Used GraphSCCFn where SCCUtilFun was used previously.
768		3. Got rid of SCCUtilFun (in comp-lib.cm).
769
770		----------------------------------------------------------------------
771		Name: Lal George
772		Date: 2000/11/06 09:02:21 EST 2000
773		Tag: Release_110_30
774		Description:
775
776		- Version 110.30
777		----------------------------------------------------------------------
778		Name: Matthias Blume
779		Date: 2000/11/04 14:45:00
780		Tag: blume-20001104-mlbuild
781		Description:
782
783		- Made ml-build faster on startup.
784		- Documentation fixes.
785
786		----------------------------------------------------------------------
787		Name: Matthias Blume
788		Date: 2000/11/02 17:00:00 JST
789		Tag: blume-20001102-condcomp
790		Description:
791
792		- Small tweaks to pickler -- new BOOTFILES!
793		- Version bumped to 110.29.2.
794		- Added conditional compilation facility to init.cmi (see comment there).
795		----------------------------------------------------------------------
796		Name: Allen Leung
797		Date: 2000/10/23 19:31:00
798		Tag: leunga-20001023-demo-ra
799
800		1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
801		2. Test programs for the graph library updated
802		3. Some new MLRISC demo programs added
803
804		----------------------------------------------------------------------
805		Name: Matthias Blume
806		Date: 2000/08/31 22:15:00 JST
807		Tag: blume-20001017-errmsg
808		Description:
809
810		More error message grief: Where there used to be no messages, there
811		now were some that had bogus error regions.  Fixed.
812
813		----------------------------------------------------------------------
814		Name: Matthias Blume
815		Date: 2000/08/31 17:30:00 JST
816		Tag: blume-20001017-v110p29p1
817		Description:
818
819		I made a version 110.29.1 with new bootfiles.
820
821		Changes:  Modified pickler/unpickler for faster and leaner unpickling.
822		CM documentation changes and a small bugfix in CM's error reporting.
823
824		----------------------------------------------------------------------
825		Name: Lal George
826		Date: 2000/09/27 14:42:35 EDT
827		Tag: george-20000927-nodestatus
828		Description:
829
830		Changed the type of the nodestatus, so that:
831
832		SPILLED(~1)             is now SPILLED
833		SPILLED(m) where m>=0   is now MEMREG(m)
834		SPILLED(s) where s<~1   is now SPILL_LOC(~s)
835
836		----------------------------------------------------------------------
837		Name: Matthias Blume
838		Date: 2000/09/07 14:45:00 JST
839		Tag: blume-20000907-cmerrmsg
840		Description:
841
842		Small tweak to CM to avoid getting ML syntax error messages twice.
843
844		----------------------------------------------------------------------
845		Name: Matthias Blume
846		Date: 2000/08/31 18:00:00 JST
847		Tag: blume-20000831-cvsbootfiles
848		Description:
849
850		New URL for boot files (because the 110.29 files on the BL server do
851		now work correctly with my updated install scripts for yacc and lex).
852
853		----------------------------------------------------------------------
854		Name: Matthias Blume
855		Date: 2000/08/08 12:33:00 JST
856		Tag: blume-20000808-manual
857		Description:
858
859		Tiny update to CM manual.
860
861		----------------------------------------------------------------------
862		Name: Allen Leung
863		Date: 2000/08/7 19:31:00
864		Tag: leunga-20000807-a-whole-bunch-of-stuff
865
866		Moby, C--, SSA, x86, machine descriptions etc.  Should only affect C--
867		and Mobdy.
868
869		1.  x86
870
871		a.  Fixes to peephole module by John and Dan.
872		b.  Assembly fix to SETcc by Allen.
873		c.  Fix to c-call by John.
874		d.  Fix to spilling by John.  (This one deals with the missing FSTPT case)
875		e.  Instruction selection optimization to SETcc as suggested by John.
876
877		For example,
878
879		MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
880
881		should generate:
882
883		MOVL a, x
884		SUBL b, x
885		SHRL 31, x
886
887		2.  IR stuff
888
889		A bunch of new DJ-graph related algorithms added.  These
890		speed up SSA construction.
891
892		3.  SSA + Scheduling
893
894		Added code for SSA and scheduling to the repository
895
896		----------------------------------------------------------------------
897		Name: Lal George
898		Date: 2000/07/27 11:53:14 EDT
899
900		Tag: lal-20000727-linux-ppc
901		Description:
902
903		Made changes to support Linux PPC.
904		p.s. I have confirmation that the 110.29 boot files work fine.
905
906		----------------------------------------------------------------------
907		Name: Matthias Blume
908		Date: 2000/07/27 17:40:00 JST
909		Tag: blume-20000727-scripts
910		Description:
911
912		!!!! WARNING !!!!
913		You must recompile the runtime system!
914		!!!! WARNING !!!!
915
916		This is basically another round of script-enhancements:
917
918		1. sml, ml-build, and ml-makedepend accept options -D and -U to define
919		and undefine CM preprocessor symbols.
920
921		2. ml-build avoids generating a new heap image if it finds that the
922		existing one is still ok.  (The condition is that no ML file had to
923		be recompiled and all ML files are found to be older that the heap
924		file.)
925
926		To make this work smoothly, I also hacked the runtime system as
927		well as SMLofNJ.SysInfo to get access to the heap image suffix
928		(.sparc-solaris, ...) that is currently being used.
929
930		Moreover, the signature of CM.mk_standalone has changed.  See the
931		CM manual.
932
933		3. ml-makedepend accepts additional options -n, -a, and -o.  (See the
934		CM manual for details.)
935
936		4. More CM manual updates:
937		- all of the above has been documented.
938		- there is now a section describing the (CM-related) command line
939		arguments that are accepted by the "sml" command
940
941		----------------------------------------------------------------------
942		Name: Matthias Blume
943		Date: 2000/07/25 16:20:00 JST
944		Tag: blume-20000725-makedepend
945		Description:
946
947		Added a script called ml-makedepend.  This can be used in makefiles
948		for Unix' make in a way very similar to the "makedepend" command for
949		C.
950
951		The script internally uses function CM.sources.
952
953		Synopsis:
954
955		ml-makedepend [-f makefile] cmfile targetname
956
957		The default for the makefile is "makefile" (or "Makefile" should
958		"makefile" not exist).
959
960		ml-makedepend adds a cmfile/targetname-specific section to this
961		makefile (after removing the previous version of this section).  The
962		section contains a single dependency specification with targetname on
963		the LHS (targetname is an arbitrary name), and a list of files derived
964		from the cmfile on the RHS.  Some of the files on the RHS are
965		ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
966		"make" variables $(ARCH) and $(OPSYS) in place of the corresponding
967		path names.  The makefile writer is responsible for making sure that
968		these variables have correct at the time "make" is invoked.
969
970		----------------------------------------------------------------------
971		Name: Matthias Blume
972		Date: 2000/07/22 23:30:00 JST
973		Tag: blume-20000722-urlupdate
974		Description:
975
976		Changed BOOT and config/srcarchiveurl to point to BL server:
977
978		ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
979
980		----------------------------------------------------------------------
981		Name: Matthias Blume
982		Date: 2000/07/18 18:00:00 JST
983		Tag: blume-20000718-Version_110_29
984		Description:
985
986		1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
987
988		2. Updated config/version to 110.29
989
990		3. Updated config/srcarchiveurl
991
992		3. New boot files!
993		ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
994
995		----------------------------------------------------------------------
996		Name: Matthias Blume
997		Date: 2000/07/11 13:58:00 JST
998		Tag: blume-20000711-doctypo
999		Description:
1000
1001		Fixed a few typos in CM manual.
1002
1003		----------------------------------------------------------------------
1004		Name: Allen Leung
1005		Date: 2000/06/15 00:38:00
1006		Tag: leunga-20000704-sparc-x86
1007
1008		1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
1009		2. fix to x86 RET bug [found by Dan Grossman]
1010		3. sparc assembly bug fix for ticc instructions [found by Fermin]
1011
1012		Affects c-- and moby only
1013
1014		----------------------------------------------------------------------
1015		Name: Matthias Blume
1016		Date: 2000/07/04 15:26:00
1017		Tag: blume-20000704-trigger
1018		Description:
1019
1020		1. Improvements to CM manual.
1021		2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
1022		of getting a back-trace.  The function, when called, raises an
1023		internal exception which explicitly carries the full back-trace history,
1024		so it is unaffected by any intervening handle-raise pairs ("trivial"
1025		or not).  The interactive loop will print that history once it arrives
1026		at top level.
1027		Short of having all exceptions implicitly carry the full history, the
1028		recommended way of using this facility is:
1029		- compile your program with instrumentation "on"
1030		- run it, when it raises an exception, look at the history
1031		- if the history is "cut off" because of some handler, go and modify
1032		your program so that it explicitly calls BTrace.trigger
1033		- recompile (still instrumented), and rerun; look at the full history
1034
1035		----------------------------------------------------------------------
1036		Name: Matthias Blume
1037		Date: 2000/07/03 15:36:00 JST
1038		Tag: blume-20000702-manual
1039		Description:
1040
1041		Small corrections and updates to CM manual.
1042
1043		----------------------------------------------------------------------
1044		Name: Matthias Blume
1045		Date: 2000/06/29 16:04:00 JST
1046		Tag: blume-20000629-yacctool
1047		Description:
1048
1049		Changes:
1050
1051		1. Class "mlyacc" now takes separate arguments to pass options to
1052		generated .sml- and .sig-files independently.
1053		2. Corresponding CM manual updates.
1054		3. BTrace module now also reports call sites.  (However, for loop clusters
1055		it only shows from where the cluster was entered.)  There are associated
1056		modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
1057
1058		----------------------------------------------------------------------
1059		Name: Matthias Blume
1060		Date: 2000/06/27 16:51:00 JST
1061		Tag: blume-20000627-noweb
1062		Description:
1063
1064		Changes:
1065
1066		1. Implemented "subdir" and "witness" options for noweb tool.
1067		This caused some slight internal changes in CM's tool implementation.
1068		2. Fixed bug in "tool plugin" mechanism.  This is essentially cleaning
1069		some remaining issues from earlier path anchor changes.
1070		3. Updated CM manual accordingly.
1071
1072		4. Changed implementation of back-tracing so that I now consider it
1073		ready for prime-time.
1074
1075		In particular, you don't have to explicitly trigger the back-trace
1076		anymore.  Instead, if you are running BTrace-instrumented code and
1077		there is an uncaught exception (regardless of whether or not it was
1078		raised in instrumented code), the top-level evalloop will print
1079		the back-trace.
1080
1081		Features:
1082
1083		- Instrumented and uninstrumented code work together seemlessly.
1084		(Of course, uninstrumented code is never mentioned in actual
1085		back-traces.)
1086
1087		- Asymptotic time- and space-complexity of instrumented code is
1088		equal to that of uninstrumented code.  (This means that
1089		tail-recursion is preserved by the instrumentation phase.)
1090
1091		- Modules whose code has been instrumented in different sessions
1092		work together without problem.
1093
1094		- There is no penalty whatsoever on uninstrumented code.
1095
1096		- There is no penalty on "raise" expressions, even in
1097		instrumented code.
1098
1099		A potential bug (or perhaps it is a feature, too):
1100
1101		A back-trace reaches no further than the outermost instrumented
1102		non-trivial "raise".  Here, a "trivial" raise is one that is the
1103		sole RHS of a "handle" rule.  Thus, back-traces reach trough
1104
1105		<exp> handle e => raise e
1106
1107		and even
1108
1109		<exp> handle Foo => raise Bar
1110
1111		and, of course, through
1112
1113		<exp> handle Foo => ...
1114
1115		if the exception was not Foo.
1116
1117		Back-traces always reach right through any un-instrumented code
1118		including any of its "handle" expressions, trivial or not.
1119
1120		To try this out, do the following:
1121
1122		- Erase all existing binfiles for your program.
1123		(You may keep binfiles for those modules where you think you
1124		definitely don't need back-tracing.)
1125		- Turn on back-trace instrumentation:
1126		SMLofNJ.Internals.BTrace.mode (SOME true);
1127		- Recompile your program.  (I.e., run "CM.make" or "use".)
1128		- You may now turn instrumentation off again (if you want):
1129		SMLofNJ.Internals.BTrace.mode (SOME false);
1130		- Run your program as usual.  If it raises an exception that
1131		reaches the interactive toplevel, then a back-trace will
1132		automatically be printed.  After that, the toplevel loop
1133		will print the exception history as usual.
1134
1135		----------------------------------------------------------------------
1136		Name: Matthias Blume
1137		Date: 2000/06/26 09:56:46 JST
1138		Tag: blume-20000626-setup
1139		Description:
1140
1141		CM: - setup-parameter to "sml" added; this can be used to run arbitrary
1142		ML code before and after compiling a file (e.g., to set compiler
1143		flags)
1144
1145		Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
1146		- associated changes to btrace.sml (BTrace instrumentation pass)
1147		- cleaner implementation of btimp.sml (BTrace tracing and report
1148		module)
1149
1150		CM manual: * new path encoding documented
1151		* description of setup-parameter to "sml" added
1152
1153		The biggest user-visible change to back-tracing is that it is no
1154		longer necessary to compile all traced modules within the same
1155		session.  (This was a real limitation.)
1156
1157		----------------------------------------------------------------------
1158		Name: Matthias Blume
1159		Date: 2000/06/24 12:40:00 JST
1160		Tag: blume-20000624-startup
1161		Description:
1162
1163		Fixes startup slowdown problem.  (I was calling SrcPath.sync a _tad_
1164		bit too often -- to put it mildly. :)
1165
1166		----------------------------------------------------------------------
1167		Name: Matthias Blume
1168		Date: 2000/06/23 18:20:00 JST
1169		Tag: blume-20000623-btrace
1170		Description:
1171
1172		This updates adds a backtrace facility to aid programmers in debugging
1173		their programs.  This involves the following changes:
1174
1175		1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
1176		keeping track of the current call stack.  When programs are compiled
1177		in a special mode, the compiler will insert calls to these hooks
1178		into the user program.
1179		"Hook" means that it is possible for different implementations of
1180		back-tracing to register themselves (at different times).
1181
1182		2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
1183		as an Absyn.dec->Absyn.dec rewrite.  Normally this phase is turned off.
1184		It can be turned on using this call:
1185		SMLofNJ.Internals.BTrace.mode (SOME true);
1186		Turning it off again:
1187		SMLofNJ.Internals.BTrace.mode (SOME false);
1188		Querying the current status:
1189		SMLofNJ.Internals.BTrace.mode NONE;
1190		Annotated programs are about twice as big as normal ones, and they
1191		run a factor of 2 to 4 slower with a dummy back-trace plugin (one
1192		where all hooks do nothing).  The slowdown with a plugin that is
1193		actually useful (such as the one supplied by default) is even greater,
1194		but in the case of the default plugin it is still only an constant
1195		factor (amortized).
1196
1197		3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
1198		with a sub-structure BTrace for controlling back-tracing.  In particular,
1199		the above-mentioned function "mode" controls whether the annotation
1200		phase is invoked by the compiler.  Another important function is
1201		"trigger": when called it aborts the current execution and causes
1202		the top-level loop to print a full back-trace.
1203
1204		4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
1205		for back-tracing.  It keeps track of the dynamic call stack and in
1206		addition to that it keeps a partial history at each "level" of that
1207		stack.  For example, if a tail-calls b, b tail-calls c, and c tail-calls
1208		d and b (at separate times, dynamically), then the report will show:
1209
1210		GOTO   d
1211		/c
1212		GOTO  \b
1213		CALL   a
1214
1215		This shows that there was an initial non-tail call of a, then a
1216		tail-call to b or c, looping behavior in a cluster of functions that
1217		consist of b and c, and then a goto from that cluster (i.e., either from
1218		b or from c) to d.
1219
1220		Note that (depending on the user program) the amount of information
1221		that the back-trace module has to keep track of at each level is bounded
1222		by a constant.  Thus, the whole implementation has the same asymptotical
1223		complexity as the original program (both in space and in time).
1224
1225		5. compiler/TopLevel/interact/evalloop.sml has been modified to
1226		handle the special exception SMLofNJ.Internals.BTrace.BTrace
1227		which is raised by the "trigger" function mentioned above.
1228
1229		Notes on usage:
1230
1231		- Annotated code works well together with unannotated code:
1232		Unannotated calls simply do not show up at all in the backtrace.
1233
1234		- It is not a good idea to let modules that were annotated during
1235		different sessions run at the same time.  This is because the compiler
1236		chooses small integers to identify individual functions, and there
1237		will be clashes if different modules were compiled in separate sessions.
1238		(Nothing will crash, and you will even be told about the clashes, but
1239		back-trace information will in general not be useful.)
1240
1241		- Back-tracing can be confused by callcc and capture.
1242
1243		- The only way of getting a back-trace right now is to explicitly
1244		invoke the "trigger" function from your user program.  Eventually, we
1245		should make every exception carry back-trace information (if
1246		available).  But since this creates more overhead at "raise"-time
1247		(similar to the current exnHistory overhead), I have not yet
1248		implemented this.  (The implementation will be rather easy.)  With
1249		exceptions carrying back-trace information, this facility will be even
1250		more useful because users don't need to modify their programs...
1251
1252		- While it is possible to compile the compiler with back-trace
1253		annotations turned on (I did it to get some confidence in
1254		correctness), you must make absolutely sure that core.sml and
1255		btimp.sml are compiled WITHOUT annotation!  (core.sml cannot actually
1256		be compiled with annotation because there is no core access yet, but
1257		if you compile btimp.sml with annotation, then the system will go into
1258		an infinite recursion and crash.)
1259		Since CM currently does not know about BTrace, the only way to turn
1260		annotations on and off for different modules of the compiler is to
1261		interrupt CMB.make, change the settings, and re-invoke it.  Of course,
1262		this is awkward and clumsy.
1263
1264		Sample sessions:
1265
1266		Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
1267		- SMLofNJ.Internals.BTrace.mode (SOME true);
1268		[autoloading]
1269		[autoloading done]
1270		val it = false : bool
1271		- structure X = struct
1272		-     fun main n = let
1273		-         fun a (x, 0) = d x
1274		-           | a (x, n) = b (x, n - 1)
1275		-         and b (x, n) = c (x, n)
1276		-         and c (x, n) = a (x, n)
1277		-         and d x = e (x, 3)
1278		-         and e (x, 0) = f x
1279		-           | e (x, n) = e (x, n - 1)
1280		-         and f 0 = SMLofNJ.Internals.BTrace.trigger ()
1281		-           | f n = n * g (n - 1)
1282		-         and g n = a (n, 3)
1283		-     in
1284		-         f n
1285		-     end
1286		- end;
1287		structure X : sig val main : int -> int end
1288		- X.main 3;
1289		*** BACK-TRACE ***
1290		GOTO   stdIn:4.2-13.20: X.main[2].f
1291		GOTO-( stdIn:4.2-13.20: X.main[2].e
1292		GOTO   stdIn:4.2-13.20: X.main[2].d
1293		/ stdIn:4.2-13.20: X.main[2].a
1294		| stdIn:4.2-13.20: X.main[2].b
1295		GOTO-\ stdIn:4.2-13.20: X.main[2].c
1296		CALL   stdIn:4.2-13.20: X.main[2].g
1297		GOTO   stdIn:4.2-13.20: X.main[2].f
1298		GOTO-( stdIn:4.2-13.20: X.main[2].e
1299		GOTO   stdIn:4.2-13.20: X.main[2].d
1300		/ stdIn:4.2-13.20: X.main[2].a
1301		| stdIn:4.2-13.20: X.main[2].b
1302		GOTO-\ stdIn:4.2-13.20: X.main[2].c
1303		CALL   stdIn:4.2-13.20: X.main[2].g
1304		GOTO   stdIn:4.2-13.20: X.main[2].f
1305		GOTO-( stdIn:4.2-13.20: X.main[2].e
1306		GOTO   stdIn:4.2-13.20: X.main[2].d
1307		/ stdIn:4.2-13.20: X.main[2].a
1308		| stdIn:4.2-13.20: X.main[2].b
1309		GOTO-\ stdIn:4.2-13.20: X.main[2].c
1310		CALL   stdIn:4.2-13.20: X.main[2].g
1311		GOTO   stdIn:4.2-13.20: X.main[2].f
1312		CALL   stdIn:2.15-17.4: X.main[2]
1313		-
1314
1315		(Note that because of a FLINt bug the above code currently does not
1316		compile without BTrace turned on.)
1317
1318		Here is another example, using my modified Tiger compiler:
1319
1320		Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
1321		- SMLofNJ.Internals.BTrace.mode (SOME true);
1322		[autoloading]
1323		[autoloading done]
1324		val it = false : bool
1325		- CM.make "sources.cm";
1326		[autoloading]
1327		...
1328		[autoloading done]
1329		[scanning sources.cm]
1330		[parsing (sources.cm):parse.sml]
1331		[creating directory CM/SKEL ...]
1332		[parsing (sources.cm):tiger.lex.sml]
1333		...
1334		[wrote CM/sparc-unix/semant.sml]
1335		[compiling (sources.cm):main.sml]
1336		[wrote CM/sparc-unix/main.sml]
1337		[New bindings added.]
1338		val it = true : bool
1339		- Main.compile ("../testcases/merge.tig", "foo.out");
1340		*** BACK-TRACE ***
1341		CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
1342		CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1343		CALL   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1344		GOTO   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1345		CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1346		CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1347		CALL   lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
1348		/ lib/semant.sml:411.65-543.8: SemantFun[2].transDec
1349		CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
1350		CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1351		CALL   lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
1352		CALL   main.sml:1.18-118.4: Main.compile[2]
1353		-
1354
1355		----------------------------------------------------------------------
1356		Name: Matthias Blumen
1357		Date: 2000/06/21 18:00:00 JST
1358		Tag: blume-20000621-manual
1359		Description:
1360
1361		CM manual update: Path environments documented.
1362
1363		----------------------------------------------------------------------
1364		Name: Matthias Blume
1365		Date: 2000/06/19 13:40:00
1366		Tag: blume-20000619-manual
1367		Description:
1368
1369		CM manual and system/README update.  This only covers the fact that
1370		there are no more implicit anchors.  (Path environments and the "bind"
1371		option to "cm" have yet to be documented.)
1372
1373		----------------------------------------------------------------------
1374		Name: Matthias Blume
1375		Date: 2000/06/19 11:05:00 JST
1376		Tag: blume-20000619-chdir-bugfix
1377		Description:
1378
1379		Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
1380
1381		----------------------------------------------------------------------
1382		Name: Matthias Blume
1383		Date: 2000/06/18 22:00:10 JST
1384		Tag: blume-20000618-implicit-anchors-really-gone
1385		Description:
1386
1387		I updates the previous HISTORY entry where I forgot to mention that
1388		implicit anchors are no longer with us.
1389
1390		The current update also gets rid of the (now useless) controller
1391		CM.Control.implicit_anchors.
1392
1393		----------------------------------------------------------------------
1394		Name: Matthias Blume
1395		Date: 2000/06/16 17:30:00 JST
1396		Tag: blume-20000616-anchorenv
1397		Description:
1398
1399		This patch implements the long anticipated (just kidding :) "anchor
1400		environment" mechanism.  In the course of doing this, I also
1401		re-implemented CM's internal "SrcPath" module from scratch.  The new
1402		one should be more robust in certain boundary cases.  In any case, it
1403		is a lot cleaner than its predecessor (IMHO).
1404
1405		This time, although there is yet another boot file format change, I
1406		kept the unpickler backward-compatible.  As a result, no new bootfiles
1407		are necessary and bootstrapping is straightforward.  (You cannot read
1408		new bootfiles into an old system, but the other way around is no
1409		problem.)
1410
1411		Visible changes:
1412
1413		** 0. Implicit path anchors (without the leading $-symbol) are no
1414		longer recognized at all. This means that such path names are not
1415		illegal either.  For example, the name basis.cm simply refers to a
1416		local file called "basis.cm" (i.e, the name is an ordinary path
1417		relative to .cm-files directory).  Or, to put it differently, only
1418		names that start with $ are anchored paths.
1419
1420		** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
1421		vanished.
1422
1423		John (Reppy) had critizised this as soon as I originally proposed and
1424		implemented it, but at that time I did not really deeply believe
1425		him. :) Now I came full-circle because I need the $<singlearc> syntax
1426		in another place where it cannot be seen as an abbreviation for
1427		$/<singlearc>.  To avoid the confusion, $<singlearc> now means what it
1428		seems to mean (i.e., it "expands" into the corresponding anchor
1429		value).
1430
1431		However, when paths are used as members in CM description files, it
1432		continues to be true that there must be at least another arc after the
1433		anchor.  This is now enforced separately during semantic analysis
1434		(i.e., from a lexical/syntactical point of view, the notation is ok.)
1435
1436		** 2. The "cm" class now accepts an option "bind".  The option's value
1437		is a sub-option list of precisely two items -- one labeled "anchor"
1438		and the other one labeled "value".  As you might expect, "anchor" is
1439		used to specify an anchor name to be bound, and "value" specifies what
1440		the anchor is being bound to.
1441
1442		The value must be a directory name and can be given in either standard
1443		syntax (including the possibility that it is itself an anchored path)
1444		or native syntax.
1445
1446		Examples:
1447
1448		foo.cm (bind:(anchor:bar value:$mystuff/bar))
1449		lib.cm (bind:(anchor:a value:"H:\\x\\y\\z"))  (* only works under windows *)
1450
1451		and so on.
1452
1453		The meaning of this is that the .cm-file will be processed with an
1454		augmented anchor environment where the given anchor(s) is/are bound to
1455		the given values(s).
1456
1457		The rationale for having this feature is this: Suppose you are trying
1458		to use two different (already stable) libraries a.cm and b.cm (that
1459		you perhaps didn't write yourself).  Further, suppose each of these
1460		two libraries internally uses its own auxiliary library $aux/lib.cm.
1461		Normally you would now have a problem because the anchor "lib" can not
1462		be bound to more than one value globally.  Therefore, the project that
1463		uses both a.cm and b.cm must locally redirect the anchor to some other
1464		place:
1465
1466		a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
1467		b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
1468
1469		This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
1470		/usr/lib/smlnj/b-stuff/aux.cm, respectively.
1471
1472		Hard-wiring path names is a bit inflexible (and CM will verbosely warn
1473		you when you do so at the time of CM.stabilize).  Therefore, you can
1474		also use an anchored path as the value:
1475
1476		a.cm (bind:(anchor:lib value:$a-lib))
1477		b.cm (bind:(anchor:lib value:$b-lib))
1478
1479		Now you can globally configure (using the usual CM.Anchor.anchor or
1480		pathconfig machinery) bindings for "a-lib" and "b-lib".  Since "lib"
1481		itself is always locally bound, setting it globally is no longer
1482		meaningful or necessary (but it does not hurt either).  In fact, "lib"
1483		can still be used as a global anchor for separate purposes.  As a
1484		matter of fact, one can locally define "lib" in terms of a global
1485		"lib":
1486
1487		a.cm (bind:(anchor:lib value:$lib/a))
1488		b.cm (bind:(anchor:lib value:$lib/b))
1489
1490		** 3: The encoding of path names has changed.  This affects the way
1491		path names are shown in CM's progress report and also the internal
1492		protocol encoding used for parallel make.
1493
1494		The encoding now uses one or more ':'-separated segments.  Each
1495		segments corresponds to a file that has been specified relative to the
1496		file given by its preceding segment.  The first segment is either
1497		relative to the CWD, absolute, or anchored.  Each segment itself is
1498		basically a Unix pathname; all segments but the first are relative.
1499
1500		Example:
1501
1502		$foo/bar/baz.cm:a/b/c.sml
1503
1504		This path denotes the file bar/a/b/c.sml relative to the directory
1505		denoted by anchor "foo".  Notice that the encoding also includes
1506		baz.cm which is the .cm-file that listed a/b/c.sml.  As usual, such
1507		paths are resolved relative to the .cm-files directory, so baz.cm must
1508		be ignored to get the "real" pathname.
1509
1510		To make this fact more obvious, CM puts the names of such "virtual
1511		arcs" into parentheses when they appear in progress reports. (No
1512		parentheses will appear in the internal protocol encoding.)  Thus,
1513		what you really see is:
1514
1515		$foo/bar/(baz.cm):a/b/c.sml
1516
1517		I find this notation to be much more informative than before.
1518
1519		Another new feature of the encoding is that special characters
1520		including parentheses, colons, (back)slashes, and white space are
1521		written as \ddd (where ddd is the decimal encoding of the character).
1522
1523		*** The CM manual still needs to be updated.
1524
1525		----------------------------------------------------------------------
1526		Name: Allen Leung
1527		Date: 2000/06/15 00:38:00
1528		Tag: leunga-20000615-x86-peephole
1529
1530		x86 Peephole fix by Fermin.  Affects c-- and moby only.
1531
1532		----------------------------------------------------------------------
1533		Name: Matthias Blume
1534		Date: 2000/06/12 11:40:00
1535		Tag: blume-20000612-parmakefix
1536		Description:
1537
1538		More cleanup after changing the file naming scheme: This time I
1539		repaired the parallel make mechanism for CMB.make which I broke earlier.
1540
1541		----------------------------------------------------------------------
1542		Name: Allen Leung
1543		Date: 2000/06/09 01:25:00
1544		Tag: leunga-20000609-various
1545
1546		None of these things should affect normal SML/NJ operations
1547
1548		1. Peephole improvements provided by Fermin (c--)
1549		2. New annotation DEFUSE for adding extra dependence (moby)
1550		3. New X86 LOCK instructions (moby)
1551		4. New machine description language for reservation tables (scheduling)
1552		5. Fixes to various optimization/analysis modules (branch chaining, dominator
1553		trees etc.)
1554		6. I've changed the CM files so that they can work with versions
1555		110.0.6, 110.25 and 110.28
1556
1557	----------------------------------------------------------------------	----------------------------------------------------------------------
1558	Name: Matthias Blume	Name: Matthias Blume
1559	Date: 2000/06/09 12:40:00	Date: 2000/06/09 12:40:00
#	Line 574	Line 2117
2117
2118	3.  Assembly	3.  Assembly
2119
2120	When generating assemby, resolve the value of client defined constants,	When generating assembly, resolve the value of client defined constants,
2121	instead of generating symbolic values.  This is controlled by the	instead of generating symbolic values.  This is controlled by the
2122	new flag "asm-resolve-constants", which is default to true.	new flag "asm-resolve-constants", which is default to true.
2123
#	Line 597	Line 2140
2140
2141	To this end, I arranged that instead of "structure Core" as "structure	To this end, I arranged that instead of "structure Core" as "structure
2142	_Core" is bound in the pervasive environment.  Core access is done via	_Core" is bound in the pervasive environment.  Core access is done via
2143	_Core (which can never be accidentially rebound because _Core is not a	_Core (which can never be accidentally rebound because _Core is not a
2144	legal surface-syntax symbol).	legal surface-syntax symbol).
2145
2146	The current solution is much cleaner because the core environment is	The current solution is much cleaner because the core environment is
#	Line 607	Line 2150
2150	with dynamic and symbolic parts of the core environment.	with dynamic and symbolic parts of the core environment.
2151
2152	Remaining hackery (to bind the "magic" symbol _Core) is localized in the	Remaining hackery (to bind the "magic" symbol _Core) is localized in the
2153	compilation mananger's bootstrap compiler (actually: in the "init group"	compilation manager's bootstrap compiler (actually: in the "init group"
2154	handling).  See the comments in src/system/smlnj/init/init.cmi for	handling).  See the comments in src/system/smlnj/init/init.cmi for
2155	more details.	more details.
2156
#	Line 722	Line 2265
2265	(specified in the .cm file at each instance where the tool's class is	(specified in the .cm file at each instance where the tool's class is
2266	used).	used).
2267
2268	This was done to accomodate the new "make" and "shell" tools which	This was done to accommodate the new "make" and "shell" tools which
2269	facilitate fairly seemless hookup to portions of code managed using	facilitate fairly seamless hookup to portions of code managed using
2270	Makefiles or Shell scripts.	Makefiles or Shell scripts.
2271
2272	There are no classes "shared" or "private" anymore.  Instead, the	There are no classes "shared" or "private" anymore.  Instead, the
#	Line 735	Line 2278
2278
2279	All existing tools are described in the CM manual.	All existing tools are described in the CM manual.
2280
2281	- Slightly better error handling.  (CM now surpresses many followup	- Slightly better error handling.  (CM now suppresses many followup
2282	error messages that tended to be more annoying than helpful.)	error messages that tended to be more annoying than helpful.)
2283
2284	2. Major changes to the compiler's static environment data structures.	2. Major changes to the compiler's static environment data structures.
#	Line 869	Line 2412
2412
2413	I've changed andl to testl in the floating point test sequence	I've changed andl to testl in the floating point test sequence
2414	whenever appropriate.  The Intel optimization guide states that	whenever appropriate.  The Intel optimization guide states that
2415	testl is perferable to andl.	testl is preferable to andl.
2416
2417	7. RA (x86 only)	7. RA (x86 only)
2418
#	Line 1051	Line 2594
2594
2595	1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an	1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
2596	additional argument called "template" which is an optional string that	additional argument called "template" which is an optional string that
2597	specifiel the layout of the tool command line.  See the CM manual for	specifies the layout of the tool command line.  See the CM manual for
2598	explanation.	explanation.
2599
2600	2. A special-purpose tool can be "regisitered" by simply dropping the	2. A special-purpose tool can be "registered" by simply dropping the
2601	corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same	corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
2602	directory where the .cm file lives that uses this tool.  (The	directory where the .cm file lives that uses this tool.  (The
2603	behavior/misfeature until now was to look for the tool description	behavior/misfeature until now was to look for the tool description
#	Line 1098	Line 2641
2641	rebuild.  Having sets of the form <base><k>.{bin,boot}.<arch>-unix for	rebuild.  Having sets of the form <base><k>.{bin,boot}.<arch>-unix for
2642	<k>=1,2,... is normally not a good idea when invoking fixpt.  However,	<k>=1,2,... is normally not a good idea when invoking fixpt.  However,
2643	they might be the result of an earlier partial run of fixpt (which	they might be the result of an earlier partial run of fixpt (which
2644	perhaps got accidentially killed).  In this case, fixpt will quickly	perhaps got accidentally killed).  In this case, fixpt will quickly
2645	move through what exists before continuing where it left off earlier,	move through what exists before continuing where it left off earlier,
2646	and, thus, saves a lot of time.	and, thus, saves a lot of time.
2647
#	Line 1148	Line 2691
2691	it from that remote directory.	it from that remote directory.
2692	This should simplify installation further:  For machines that have	This should simplify installation further:  For machines that have
2693	access to the internet, just fetch <version>-config.tgz, unpack it,	access to the internet, just fetch <version>-config.tgz, unpack it,
2694	edit config/targets, and go (run config/install.sh).  The scipt will	edit config/targets, and go (run config/install.sh).  The script will
2695	fetch everything else that it might need all by itself.	fetch everything else that it might need all by itself.
2696
2697	For CVS users, this mechanism is not relevant for source archives, but	For CVS users, this mechanism is not relevant for source archives, but

---

Colored Diff Long Colored Diff Unidiff Context Diff Side by Side

Legend: Removed from v.656   changed lines   Added in v.837

---

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