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

SCM Repository

[smlnj] Diff of /sml/trunk/HISTORY
ViewVC logotype

Diff of /sml/trunk/HISTORY

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

revision 586, Thu Mar 30 05:08:07 2000 UTC revision 1021, Wed Jan 16 20:33:57 2002 UTC
# Line 8  Line 8 
8  The form of an entry should be:  The form of an entry should be:
9    
10  Name:  Name:
11  Date:  Date: yyyy/mm/dd
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    ----------------------------------------------------------------------
15    Name: Matthias Blume
16    Date: 2002/01/16
17    Tag: blume-20020116-ffifiddle
18    Description:
19    
20    More fiddling with the FFI interface:
21    
22     - Make constness 'c instead of rw wherever possible.  This eliminates
23       the need for certain explicit coercions.  (However, due to ML's
24       value polymorphism, there will still be many cases where explicit
25       coercions are necessary.  Phantom types are not the whole answer
26       to modeling a subtyping relationship in ML.)
27    
28     - ro/rw coersions for pointers added.  (Avoids the detour through */&.)
29    
30     - "printf" test example added to src/ml-nlffi-lib/Tests.  (Demonstrates
31       clumsy workaround for varargs problem.)
32    
33    ----------------------------------------------------------------------
34    Name: Lal George
35    Date: 2002/01/15
36    Tag: <none>
37    Description:
38    
39    1. Since COPY instructions are no longer native to the architecture,
40       a generic functor can be used to implement the expandCopies function.
41    
42    2. Allowed EXPORT and IMPORT pseudo-op declarations to appear inside a
43       TEXT segment.
44    
45    ----------------------------------------------------------------------
46    Name: Matthias Blume
47    Date: 2002/01/15
48    Tag: blume-20020115-ffiupdates
49    Description:
50    
51    1. Fix for bug resulting in single-precision float values being returned
52       incorrectly from FFI calls.
53    
54    2. Small modifications to C FFI API:
55    
56        - memory-allocation routines return straight objects (no options)
57          and raise an exception in out-of-memory situations
58        - unsafe extensions to cast between function pointers and pointers
59          from/to ints
60        - added structure C_Debug as an alternative to structure C where
61          pointer-dereferencing (|*| and |*!) always check for null-pointers
62        - added open_lib' to DynLinkage;  open_lib' works like open_lib
63          but also takes a (possibly empty) list of existing library handles
64          that the current library depends on
65    
66    ----------------------------------------------------------------------
67    Name: Matthias Blume
68    Date: 2002/01/10
69    Tag: blume-20020110-newffigen
70    Description:
71    
72    1. Updates to portable graph code.
73    
74    2. Major update to ml-nlffigen and ml-nlffi-lib.  Things are much
75       more scalable now so that even huge interfaces such as the one
76       for GTK compile in finite time and space. :-)
77       See src/ml-nlffigen/README for details on what's new.
78    
79    ----------------------------------------------------------------------
80    Name: Lal George
81    Date: 2001/01/09 14:31:35 EST 2002
82    Tag: george-20011206-rm-native-copy
83    Description:
84    
85            Removed the native COPY and FCOPY instructions
86            from all the architectures and replaced it with the
87            explicit COPY instruction from the previous commit.
88    
89            It is now possible to simplify many of the optimizations
90            modules that manipulate copies. This has not been
91            done in this change.
92    
93    ----------------------------------------------------------------------
94    Name: Lal George
95    Date: 2001/12/06 16:50:13 EST 2001
96    Tag: george-20011206-mlrisc-instruction
97    Description:
98    
99    Changed the representation of instructions from being fully abstract
100    to being partially concrete. That is to say:
101    
102      from
103            type instruction
104    
105      to
106            type instr                              (* machine instruction *)
107    
108            datatype instruction =
109                LIVE of {regs: C.cellset, spilled: C.cellset}
110              | KILL of {regs: C.cellset, spilled: C.cellset}
111              | COPYXXX of {k: CB.cellkind, dst: CB.cell list, src: CB.cell list}
112              | ANNOTATION of {i: instruction, a: Annotations.annotation}
113              | INSTR of instr
114    
115    This makes the handling of certain special instructions that appear on
116    all architectures easier and uniform.
117    
118    LIVE and KILL say that a list of registers are live or killed at the
119    program point where they appear. No spill code is generated when an
120    element of the 'regs' field is spilled, but the register is moved to
121    the 'spilled' (which is present, more for debugging than anything else).
122    
123    LIVE replaces the (now deprecated) DEFFREG instruction on the alpha.
124    We used to generate:
125    
126            DEFFREG f1
127            f1 := f2 + f3
128            trapb
129    
130    but now generate:
131    
132            f1 := f2 + f3
133            trapb
134            LIVE {regs=[f1,f2,f3], spilled=[]}
135    
136    Furthermore, the DEFFREG (hack) required that all floating point instruction
137    use all registers mentioned in the instruction. Therefore f1 := f2 + f3,
138    defines f1 and uses [f1,f2,f3]! This hack is no longer required resulting
139    in a cleaner alpha implementation. (Hopefully, intel will not get rid of
140    this architecture).
141    
142    COPYXXX is intended to replace the parallel COPY and FCOPY  available on
143    all the architectures. This will result in further simplification of the
144    register allocator that must be aware of them for coalescing purposes, and
145    will also simplify certain aspects of the machine description that provides
146    callbacks related to parallel copies.
147    
148    ANNOTATION should be obvious, and now INSTR represents the honest to God
149    machine instruction set!
150    
151    The <arch>/instructions/<arch>Instr.sml files define certain utility
152    functions for making porting easier -- essentially converting upper case
153    to lower case. All machine instructions (of type instr) are in upper case,
154    and the lower case form generates an MLRISC instruction. For example on
155    the alpha we have:
156    
157      datatype instr =
158         LDA of {r:cell, b:cell, d:operand}
159       | ...
160    
161      val lda : {r:cell, b:cell, d:operand} -> instruction
162        ...
163    
164    where lda is just (INSTR o LDA), etc.
165    
166    ----------------------------------------------------------------------
167    Name: Matthias Blume
168    Date: 2001/11/22 21:40:00 EST
169    Tag: Release_110_37
170    Description:
171    
172    Release 110.37.  This time for real.
173    
174    ----------------------------------------------------------------------
175    Name: Matthias Blume
176    Date: 2001/11/21 16:35:00 EST
177    Tag: blume-20011121-foot-in-mouth
178    Description:
179    
180    Removed the "Release_110_37" tag because of a serious bug.
181    This will be re-tagged once the bug is fixed.
182    
183    ----------------------------------------------------------------------
184    Name: Matthias Blume
185    Date: 2001/11/21 16:14:00 EST
186    Tag: blume-20011121-forgottenfile
187    Description:
188    
189    Forgot to add a file.  (Just a .tex-file -- part of
190    the CM manual source.)
191    
192    ----------------------------------------------------------------------
193    Name: Matthias Blume
194    Date: 2001/11/21 16:10:00 EST
195    Tag: blume-20011121-invalid_110_37
196    Description:
197    
198    Note: I removed the original tag "Release_110_37" from this commit
199          because we found a serious bug in all non-x86 backends.
200          - Matthias
201    
202    1. Modifications to the SML/NJ code generator and to the runtime system
203       so that code object name strings are directly inserted into code
204       objects at code generation time.  The only business the runtime system
205       has with this is now to read the name strings on occasions.
206       (The encoding of the name string has also changed somewhat.)
207    
208    2. CM now implements a simple "set calculus" for specifying export lists.
209       In particular, it is now possible to refer to the export lists of
210       other libraries/groups/sources and form unions as well as differences.
211       See the latest CM manual for details.
212    
213    3. An separate notion of "proxy" libraries has again be eliminated from
214       CM's model.  (Proxy libraries are now simply a special case of using
215       the export list calculus.)
216    
217    4. Some of the existing libraries now take advantage of the new set
218       calculus.
219       (Notice that not all libraries have been converted because some
220       of the existing .cm-files are supposed to be backward compatible
221       with 110.0.x.)
222    
223    5. Some cleanup in stand-alone programs.  (Don't use "exnMessage" -- use
224       "General.exnMessage"!  The former relies on a certain hook to be
225       initialized, and that often does not happen in the stand-alone case.)
226    
227    ----------------------------------------------------------------------
228    Name: Lal George
229    Date: 2001/11/21  13:56:18 EST
230    Tag: george-2001121-pseudo-ops
231    Description:
232    
233      Implemented a complete redesign of MLRISC pseudo-ops. Now there
234      ought to never be any question of incompatabilities with
235      pseudo-op syntax expected by host assemblers.
236    
237      For now, only modules supporting GAS syntax are implemented
238      but more should follow, such as MASM, and vendor assembler
239      syntax, e.g. IBM as, Sun as, etc.
240    
241    ----------------------------------------------------------------------
242    Name: Matthias Blume
243    Date: 2001/11/14 11:52:00 EST
244    Tag: blume-20011114-srcname
245    Description:
246    
247    1. Routed the name of the current source file to mlriscgen where it
248       should be directly emitted into the code object.  (This last part
249       is yet to be done.)
250    
251    2. Some cleanup of the pgraph code to make it match the proposal that
252       I put out the other day.  (The proposal notwithstanding, things are
253       still in flux here.)
254    
255    ----------------------------------------------------------------------
256    Name: Lal George
257    Date: 2001/11/14 09:44:04 EST
258    Tag:
259    Description:
260    
261      Fix for a backpatching bug reported by Allen.
262    
263      Because the boundary between short and long span-dependent
264      instructions is +/- 128, there are an astounding number of
265      span-dependent instructions whose size is over estimated.
266    
267      Allen came up with the idea of letting the size of span
268      dependent instructions be non-monotonic, for a maxIter
269      number of times, after which the size must be monotonically
270      increasing.
271    
272      This table shows the number of span-dependent instructions
273      whose size was over-estimated as a function of maxIter, for the
274      file Parse/parse/ml.grm.sml:
275    
276         maxIter            # of instructions:
277            10                      687
278            20                      438
279            30                      198
280            40                        0
281    
282      In compiling the compiler, there is no significant difference in
283      compilation speed between maxIter=10 and maxIter=40. Actually,
284      my measurements showed that maxIter=40 was a tad faster than
285      maxIter=10! Also 96% of the  files in the compiler reach a fix
286      point within 13 iterations, so fixing maxIter at 40, while high,
287      is okay.
288    
289    ----------------------------------------------------------------------
290    Name: Matthias Blume
291    Date: 2001/10/31 15:25:00 EST
292    Tag: blume-20011031-pgraph
293    Description:
294    
295    CKIT:
296    * Changed the "Function" constructor of type Ast.ctype to carry optional
297      argument identifiers.
298    * Changed the return type of TypeUtil.getFunction accordingly.
299    * Type equality ignores the argument names.
300    * TypeUtil.composite tries to preserve argument names but gives up quickly
301      if there is a mismatch.
302    
303    installation script:
304    * attempts to use "curl" if available (unless "wget" is available as well)
305    
306    CM:
307    * has an experimental implementation of "portable graphs" which I will
308      soon propose as an implementation-independent library format
309    * there are also new libraries $/pgraph.cm and $/pgraph-util.cm
310    
311    NLFFI-LIB:
312    * some cleanup (all cosmetic)
313    
314    NLFFIGEN:
315    * temporarily disabled the mechanism that suppresses ML output for
316      C definitions whose identifiers start with an underscore character
317    * generate val bindings for enum constants
318    * user can request that only one style (light or heavy) is being used;
319      default is to use both (command-line arguments: -heavy and -light)
320    * fixed bug in handling of function types involving incomplete pointers
321    * generate ML entry points that take record arguments (i.e., using
322      named arguments) for C functions that have a prototype with named
323      arguments
324      (see changes to CKIT)
325    
326    ----------------------------------------------------------------------
327    Name: Allen Leung
328    Date: 2001/10/27 20:34:00 EDT
329    Tag: leunga-20011027-x86-fast-fp-call
330    Description:
331    
332       Fixed the bug described in blume-20010920-slowfp.
333    
334       The fix involves
335          1. generating FCOPYs in FSTP in ia32-svid
336          2. marking a CALL with the appropriate annotation
337    
338    ----------------------------------------------------------------------
339    Name: Matthias Blume
340    Date: 2001/10/16 11:32:00 EDT
341    Tag: blume-20011016-netbsd
342    Description:
343    
344    Underscore patch from Chris Richards (fixing problem with compiling
345    runtime system under recent NetBSD).
346    
347    ----------------------------------------------------------------------
348    Name: Allen Leung
349    Date: 2001/10/12 17:18:32 EDT 2001
350    Tag: leung-20011012-x86-printflowgraph
351    Description:
352    
353    X86RA now uses a valid (instead of dummy) PrintFlowgraph module.
354    
355    ----------------------------------------------------------------------
356    Name: Lal George
357    Date: 2001/10/11 23:51:34 EDT
358    Tag: george-20011011-too-many-instrs
359    Description:
360    
361    The representation of a program point never expected to see more
362    than 65536 instructions in a basic block!
363    
364    ----------------------------------------------------------------------
365    Name: Lal George
366    Date: 2001/10/09 09:41:37 EDT
367    Tag: george-20011008-mlrisc-labels
368    Description:
369    
370    Changed the machine description files to support printing of
371    local and global labels in assembly code, based on host assembler
372    conventions.
373    
374    ----------------------------------------------------------------------
375    Name: Matthias Blume
376    Date: 2001/09/25 15:25:00 EDT
377    Tag: blume-20010925-exninfo
378    Description:
379    
380    I provided a non-hook implementation of exnName (at the toplevel) and
381    made the "dummy" implementation of exnMessage (at the toplevel) more
382    useful: if nothing gets "hooked in", then at least you are going to
383    see the exception name and a message indicating why you don't see more.
384    
385    [For the time being, programs that need exnMessage and want to use
386    ml-build should either use General.exnMessage (strongly recommended) or
387    refer to structure General at some other point so that CM sees a
388    static dependency.]
389    
390    [Similar remarks go for "print" and "use":  If you want to use their
391    functionality in stand-alone programs generated by ml-build, then use
392    TextIO.output and Backend.Interact.useFile (from $smlnj/compiler.cm).]
393    
394    ----------------------------------------------------------------------
395    Name: Matthias Blume
396    Date: 2001/09/20 17:28:00 EDT
397    Tag: blume-20010920-slowfp
398    Description:
399    
400    Allen says that x86-fast-fp is not safe yet, so I turned it off again...
401    
402    ----------------------------------------------------------------------
403    Name: Matthias Blume
404    Date: 2001/09/20 17:20:00 EDT
405    Tag: blume-20010920-canonicalpaths
406    Description:
407    
408    0. Updated the BOOT file (something that I forgot to do earlier).
409    
410    1. Small internal change to CM so that it avoids "/../" in filenames
411       as much as possible (but only where it is safe).
412    
413    2. Changed config/_run-sml (resulting in a changed bin/.run-sml) so
414       that arguments that contain delimiters are passed through correctly.
415       This change also means that all "special" arguments of the form
416       @SMLxxx... must come first.
417    
418    3. Changed install script to put relative anchor names for tool commands
419       into pathconfig.
420    
421    ----------------------------------------------------------------------
422    Name: Matthias Blume
423    >>>>>>> 1.169
424    Date: 2001/09/18 15:35:00 EDT
425    Tag: blume-20010918-readme11036
426    Description:
427    
428    Added README files.
429    
430    ----------------------------------------------------------------------
431    Name: Matthias Blume
432    Date: 2001/09/18 11:45:00 EDT
433    Tag: Release_110_36 (retag)
434    Description:
435    
436    Fixed mistake in config/preloads. Retagged as 110.36.
437    
438    ----------------------------------------------------------------------
439    Name: Matthias Blume
440    Date: 2001/09/18 09:40:00 EDT
441    Tag: Release_110_36_orig (tag changed)
442    Description:
443    
444    New version (110.36).  New bootfiles.
445    
446    ----------------------------------------------------------------------
447    Name: Matthias Blume
448    Date: 2001/09/14 16:15:00 EDT
449    Tag: blume-20010914-x86fastfp
450    Description:
451    
452    John committed some changes that Allen made, in particular a (hopefully)
453    correctly working version of the x86-fp module.
454    
455    I changed the default setting of the Control.MLRISC.getFlag "x86-fast-fp"
456    flag to "true".  Everything seems to compile to a fixpoint ok, and
457    "mandelbrot" speeds up by about 15%.
458    
459    ----------------------------------------------------------------------
460    Name: Matthias Blume
461    Date: 2001/09/13 11:20:00 EDT
462    Tag: blume-20010913-minimal
463    Description:
464    
465    1. Stefan Monnier's patch to fix a miscompilation problem that
466       was brought to light by John Reppy's work on Moby.
467    
468    2. Implemented a minimal "structure Compiler" that contains just
469       "version" and "architecture".  The minimal version will be
470       available when the full version is not.  This is for backward-
471       compatibility with code that wants to test Compiler.version.
472    
473    ----------------------------------------------------------------------
474    Name: Matthias Blume
475    Date: 2001/08/28 14:03:00 EDT
476    Tag: blume-20010828-ml-lex
477    Description:
478    
479    Fix for bug 1581, received from Neophytos Michael.
480    
481    ----------------------------------------------------------------------
482    Name: Matthias Blume
483    Date: 2001/08/27 11:20:00 EDT
484    Tag: blume-20010827-readme11035
485    Description:
486    
487    Fleshed out the README file for 110.35.
488    
489    ----------------------------------------------------------------------
490    Name: Matthias Blume
491    Date: 2001/08/24 17:10:00 EDT
492    Tag: Release_110_35
493    Description:
494    
495    New version number (110.35).  New bootfiles.
496    
497    ----------------------------------------------------------------------
498    Name: Lal George
499    Date: 2001/08/24 13:47:18 EDT 2001
500    Tag: george-20010824-MLRISC-graphs
501    Description:
502    
503     removed clusters from MLRISC completely and replaced with graphs.
504    
505    ----------------------------------------------------------------------
506    Name: Matthias Blume
507    Date: 2001/08/23 17:50:00 EDT
508    Tag: blume-20010823-toplevel
509    Description:
510    
511    - some reorganization of the code that implements various kinds of
512      environments in the compiler (static, dynamic, symbolic, combined)
513    - re-implemented the EnvRef module so that evalStream works properly
514      (if the stream contains references to "use", "CM.make", etc.)
515    - cleaned up evalloop.sml and interact.sml (but they need more cleaning)
516    
517    ----------------------------------------------------------------------
518    Name: Matthias Blume
519    Date: 2001/08/20 15:50 EDT
520    Tag: blume20010820-slipup
521    Description:
522    
523    I forgot to commit a few files.  Here they are...
524    
525    ----------------------------------------------------------------------
526    Name: Matthias Blume
527    Date: 2001/08/20 15:35:00 EDT
528    Tag: blume-20010820-debugprof
529    Description:
530    
531    !!!! NEW BOOTFILES !!!!
532    
533    This is another round of reorganizing the compiler sources.  This
534    time the main goal was to factor out all the "instrumentation"
535    passes (for profiling and backtracing) into their own library.
536    The difficulty was to do it in such a way that it does not depend
537    on elaborate.cm but only on elabdata.cm.
538    
539    Therefore there have been further changes to both elaborate.cm and
540    elabdata.cm -- more "generic" things have been moved from the former
541    to the latter.  As a result, I was forced to split the assignment
542    of numbers indicating "primtyc"s into two portions: SML-generic and
543    SML/NJ-specific.  Since it would have been awkward to maintain,
544    I bit the bullet and actually _changed_ the mapping between these
545    numbers and primtycs.  The bottom line of this is that you need
546    a new set of bin- and bootfiles.
547    
548    I have built new bootfiles for all architectures, so doing a fresh
549    checkout and config/install.sh should be all you need.
550    
551    The newly created library's name is
552    
553        $smlnj/viscomp/debugprof.cm
554    
555    and its sources live under
556    
557        src/compiler/DebugProf
558    
559    ----------------------------------------------------------------------
560    Name: Matthias Blume
561    Date: 2001/08/15 17:15:00 EDT
562    Tag: blume-20010815-compreorg
563    Description:
564    
565    This is a first cut at reorganizing the CM libraries that make up the
566    core of the compiler.  The idea is to separate out pieces that could
567    be used independently by tools, e.g., the parser, the typechecker, etc.
568    
569    The current status is a step in this direction, but it is not quite
570    satisfactory yet.  Expect more changes in the future.
571    
572    Here is the current (new) organization...
573    
574        What used to be $smlnj/viscomp/core.cm is now divided into
575        six CM libraries:
576    
577             $smlnj/viscomp/basics.cm
578                           /parser.cm
579                           /elabdata.cm
580                           /elaborate.cm
581                           /execute.cm
582                           /core.cm
583    
584        The CM files for these libraries live under src/system/smlnj/viscomp.
585        All these libraries are proxy libraries that contain precisely
586        one CM library component.  Here are the locations of the components
587        (all within the src/compiler tree):
588    
589             Basics/basics.cm
590             Parse/parser.cm
591             ElabData/elabdata.cm
592             Elaborator/elaborate.cm
593             Execution/execute.cm
594             core.cm
595    
596         [This organization is the same that has been used already
597         for a while for the architecture-specific parts of the visible
598         compiler and for the old version of core.cm.]
599    
600         As you will notice, many source files have been moved from their
601         respective original locations to a new home in one of the above
602         subtrees.
603    
604         The division of labor between the new libraries is the following:
605    
606             basics.cm:
607                - Simple, basic definitions that pertain to many (or all) of
608                  the other libraries.
609             parser.cm:
610                - The SML parser, producing output of type Ast.dec.
611                - The type family for Ast is also defined and exported here.
612             elabdata.cm:
613                - The datatypes that describe input and output of the elaborator.
614                  This includes types, absyn, and static environments.
615             elaborator.cm:
616                - The SML/NJ type checker and elaborator.
617                  This maps an Ast.dec (with a given static environment) to
618                  an Absyn.dec (with a new static environment).
619                - This libraries implements certain modules that used to be
620                  structures as functors (to remove dependencies on FLINT).
621             execute.cm:
622                - Everything having to do with executing binary code objects.
623                - Dynamic environments.
624             core.cm:
625                - SML/NJ-specific instantiations of the elaborator and MLRISC.
626                - Top-level modules.
627                - FLINT (this should eventually become its own library)
628    
629    Notes:
630    
631    I am not 100% happy with the way I separated the elaborator (and its
632    data structures) from FLINT.  Two instances of the same problem:
633    
634        1. Data structures contain certain fields that carry FLINT-specific
635           information.  I hacked around this using exn and the property list
636           module from smlnj-lib.  But the fact that there are middle-end
637           specific fields around at all is a bit annoying.
638    
639        2. The elaborator calculates certain FLINT-related information.  I tried
640           to make this as abstract as I could using functorization, but, again,
641           the fact that the elaborator has to perform calculations on behalf
642           of the middle-end at all is not nice.
643    
644        3. Having to used exn and property lists is unfortunate because it
645           weakens type checking.  The other alternative (parameterizing
646           nearly *everything*) is not appealing, though.
647    
648    I removed the "rebinding =" warning hack because due to the new organization
649    it was awkward to maintain it.  As a result, the compiler now issues some of
650    these warnings when compiling init.cmi during bootstrap compilation. On
651    the plus side, you also get a warning when you do, for example:
652       val op = = Int32.+
653    which was not the case up to now.
654    
655    I placed "assign" and "deref" into the _Core structure so that the
656    code that deals with the "lazy" keyword can find them there.  This
657    removes the need for having access to the primitive environment
658    during elaboration.
659    
660    ----------------------------------------------------------------------
661    Name: Matthias Blume
662    Date: 2001/08/13
663    Tag: blume-20010813-closures
664    Description:
665    
666    This fix was sent to us by Zhong Shao.  It is supposed to improve the
667    performance of certain loops by avoiding needless closure allocation.
668    
669    ----------------------------------------------------------------------
670    Name: Lal George
671    Date: 2001/07/31 10:03:23 EDT 2001
672    Tag: george-20010731-x86-fmalloc
673    Description: Fixed bug in x86 calls
674    
675        There was a bug where call instructions would mysteriously
676        vanish. The call instruction had to be one that returned
677        a floating point value.
678    
679    ----------------------------------------------------------------------
680    Name: Lal George
681    Date: 2001/07/19 16:36:29 EDT 2001
682    Tag: george-20010719-simple-cells
683    Description:
684    
685    I have dramatically simplified the interface for CELLS in MLRISC.
686    
687    In summary, the cells interface is broken up into three parts:
688    
689      1. CellsBasis : CELLS_BASIS
690    
691            CellsBasis is a top level structure and common for all
692            architectures.  it contains the definitions of basic datatypes
693            and utility  functions over these types.
694    
695      2. functor Cells() : CELLS
696    
697            Cells generates an interface for CELLS that incorporates the
698            specific resources on the target architecture, such as the
699            presence of special register classes, their number and size,
700            and various useful substructures.
701    
702      3. <ARCH>CELLS
703    
704            e.g. SparcCells: SPARCCELLS
705    
706            <ARCH>CELLS usually contains additional bindings for special
707            registers  on the architecture, such as:
708    
709                    val r0 : cell           (* register zero *)
710                    val y : cell            (* Y register *)
711                    val psr : cell          (* processor status register *)
712                    ...
713    
714            The structure returned by applying the Cells functor is opened
715            in this interface.
716    
717    The main implication of all this is that the datatypes for cells is
718    split between CellsBasis and CELLS -- a fairly simple change for user
719    code.
720    
721    In the old scheme the CELLS interface had a definitional binding of
722    the form:
723    
724            signature CELLS = sig
725    
726               structure CellsBasis = CellsBasis
727    
728               ...
729    
730            end
731    
732    With all the sharing constraints that goes on in MLRISC, this old
733    design  quickly leads to errors such as:
734    
735            "structure definition spec inside of sharing ... "
736    
737    
738    and appears to require an unacceptable amount of sharing and where
739    constraint hackery.
740    
741    I think this error message (the interaction of definitional specs and
742    sharing) requires more explanation on our web page.
743    
744    ----------------------------------------------------------------------
745    Name: Matthias Blume
746    Date: 2001/07/19 15:00:00 EDT
747    Tag: blume-20010719-libreorg
748    Description:
749    
750    This update puts together a fairly extensive but straightforward change
751    to the way the libraries that implement the interactive system are
752    organized:
753    
754       The biggest change is the elimination of structure Compiler.  As a
755       replacement for this structure, there is now a CM library
756       (known as $smlnj/compiler.cm or $smlnj/compiler/current.cm)
757       that exports all the substructures of the original structure Compiler
758       directly.  So instead of saying Compiler.Foo.bar one now simply
759       says Foo.bar.  (The CM libraries actually export a collection of
760       structures that is richer than the collection of substructures of
761       structure Compiler.)
762    
763       To make the transition smooth, there is a separate library called
764       $smlnj/compiler/compiler.cm which puts together and exports the
765       original structure Compiler (or at least something very close to it).
766    
767       There are five members of the original structure Compiler
768       that are not exported directly but which instead became members
769       of a new structure Backend (described by signature BACKEND).  These are:
770       structure Profile (: PROFILE), structure Compile (: COMPILE), structure
771       Interact (: INTERACT), structure Machine (: MACHINE), and val
772       architecture (: string).
773    
774       Structure Compiler.Version has become structure CompilerVersion.
775    
776       Cross-compilers for alpha32, hppa, ppc, sparc, and x86 are provided
777       by $smlnj/compiler/<arch>.cm where <arch> is alpha32, hppa, ppc, sparc,
778       or x86, respectively.
779       Each of these exports the same frontend structures that
780       $smlnj/compiler.cm exports.  But they do not have a structure Backend
781       and instead export some structure <Arch>Backend where <Arch> is Alpha32,
782       Hppa, PPC, Sparc, or X86, respectively.
783    
784       Library $smlnj/compiler/all.cm exports the union of the exports of
785       $smlnj/compiler/<arch>.cm
786    
787       There are no structures <Arch>Compiler anymore, use
788       $smlnj/compiler/<arch>.cm instead.
789    
790       Library host-compiler-0.cm is gone.  Instead, the internal library
791       that instantiates CM is now called cm0.cm.  Selection of the host
792       compiler (backend) is no longer done here but. (Responsibility for it
793       now lies with $smlnj/compiler/current.cm.  This seems to be more
794       logical.)
795    
796       Many individual files have been moved or renamed.  Some files have
797       been split into multiple files, and some "dead" files have been deleted.
798    
799    Aside from these changes to library organization, there are also changes
800    to the way the code itself is organized:
801    
802       Structure Binfile has been re-implemented in such a way that it no
803       longer needs any knowledge of the compiler.  It exclusively deals
804       with the details of binfile layout.  It no longer invokes the
805       compiler (for the purpose of creating new prospective binfile
806       content), and it no longer has any knowledge of how to interpret
807       pickles.
808    
809       Structure Compile (: COMPILE) has been stripped down to the bare
810       essentials of compilation.  It no longer deals with linking/execution.
811       The interface has been cleaned up considerably.
812    
813       Utility routines for dealing with linking and execution have been
814       moved into their own substructures.
815    
816       (The ultimate goal of these changes is to provide a light-weight
817       binfile loader/linker (at least for, e.g., stable libraries) that
818       does not require CM or the compiler to be present.)
819    
820    CM documentation has been updated to reflect the changes to library
821    organization.
822    
823    ----------------------------------------------------------------------
824    Name: Matthias Blume
825    Date: 2001/07/10 17:30:00 EDT
826    Tag: Release_110_34
827    Description:
828    
829    Minor tweak to 110.34 (re-tagged):
830    
831      - README.html file added to CVS repository
832      - runtime compiles properly under FreeBSD 3.X and 4.X
833    
834    ----------------------------------------------------------------------
835    Name: Matthias Blume
836    Date: 2001/07/10 17:30:00 EDT
837    Tag: Release_110_34
838    Description:
839    
840    New version number (110.34). New bootfiles.
841    
842    ----------------------------------------------------------------------
843    Name: Matthias Blume
844    Date: 2001/07/09 16:00:00 EDT
845    Tag: blume-20010709-more-varargs
846    Description:
847    
848    I changed the handling of varargs in ml-nlffigen again:
849    The ellipsis ... will now simply be ignored (with an accompanying warning).
850    
851    The immediate effect is that you can actually call a varargs function
852    from ML -- but you can't actually supply any arguments beyond the ones
853    specified explicitly.  (For example, you can call printf with its format
854    string, but you cannot pass additional arguments.)
855    
856    This behavior is only marginally more useful than the one before, but
857    it has the advantage that a function or, more importantly, a function
858    type never gets dropped on the floor, thus avoiding follow-up problems with
859    other types that refer to the offending one.
860    
861    ----------------------------------------------------------------------
862    Name: Matthias Blume
863    Date: 2001/07/09 11:25:00 EDT
864    Tag: blume-20010709-varargs
865    Description:
866    
867    1. ckit-lib.cm now exports structure Error
868    2. ml-nlffigen reports occurences of "..." (i.e., varargs function types)
869       with a warning accompanied by a source location.  Moreover, it
870       merely skips the offending function or type and proceeds with the
871       rest of its work.u  As a result, one can safely feed C code containing
872       "..." to ml-nlffigen.
873    3. There are some internal improvements to CM, providing slightly
874       more general string substitutions in the tools subsystem.
875    
876    ----------------------------------------------------------------------
877    Name: Matthias Blume
878    Date: 2001/06/27 15:10:00 EDT
879    Tag: blume-20010627-concur
880    Description:
881    
882    Fixed a small bug in CM's handling of parallel compilation.
883    (You could observe the bug by Control-C-interrupting an ordinary
884    CMB.make or CM.stabilize and then attaching some compile servers.
885    The result was that all of a sudden the previously interrupted
886    compilation would continue on its own.  This was because of
887    an over-optimization: CM did not bother to clean out certain queues
888    when no servers were attached "anyway", resulting in the contents
889    of these queues to grab control when new servers did get attached.)
890    
891    There is also another minor update to the CM manual.
892    
893    ----------------------------------------------------------------------
894    Name: Matthias Blume
895    Date: 2001/06/26 16:15:00 EDT
896    Tag: blume-20010626-cmdoc
897    Description:
898    
899    Minor typo fixed in CM manual (syntax diagram for libraries).
900    
901    ----------------------------------------------------------------------
902    Name: Matthias Blume
903    Date: 2001/06/25 22:55:00 EDT
904    Tag: blume-20010625-x86pc
905    Description:
906    
907    Fixed a nasty bug in the X86 assembly code that caused signal
908    handlers to fail (crash) randomly.
909    
910    ----------------------------------------------------------------------
911    Name: Matthias Blume
912    Date: 2001/06/25 12:05:00 EDT
913    Tag: blume-20010625-nlffigen
914    Description:
915    
916    This update fixes a number of minor bugs in ml-nlffigen as reported by
917    Nick Carter <nbc@andrew.cmu.edu>.
918    
919      1. Silly but ok typedefs of the form "typedef void myvoid;" are now accepted.
920      2. Default names for generated files are now derived from the name of
921         the C file *without its directory*.  In particular, this causes generated
922         files to be placed locally even if the C file is in some system directory.
923      3. Default names for generated signatures and structures are also derived
924         from the C file name without its directory.  This avoids silly things
925         like "structure GL/GL".
926         (Other silly names are still possible because ml-nlffigen does not do
927          a thorough check of whether generated names are legal ML identifiers.
928          When in doubt, use command line arguments to force particular names.)
929    
930    ----------------------------------------------------------------------
931    Name: Matthias Blume
932    Date: 2001/06/21 12:25:00 EDT
933    Tag: blume-20010621-eXene
934    Description:
935    
936    eXene now compiles and (sort of) works again.
937    
938    The library name (for version > 110.33) is $/eXene.cm.
939    
940    I also added an new example in src/eXene/examples/nbody.  See the
941    README file there for details.
942    
943    ----------------------------------------------------------------------
944    Name: Matthias Blume
945    Date: 2001/06/20 16:40:00 EDT
946    Tag: blume-20010620-cml
947    Description:
948    
949    CML now compiles and works again.
950    
951    Libraries (for version > 110.33):
952    
953      $cml/cml.cm            Main CML library.
954      $cml/basis.cm          CML's version of $/basis.cm.
955      $cml/cml-internal.cm   Internal helper library.
956      $cml/core-cml.cm       Internal helper library.
957      $cml-lib/trace-cml.cm  Tracing facility.
958      $cml-lib/smlnj-lib.cm  CML's version of $/smlnj-lib.cm
959    
960    The installer (config/install.sh) has been taught how to properly
961    install this stuff.
962    
963    ----------------------------------------------------------------------
964    Name: Matthias Blume
965    Date: 2001/06/19 17:55:00 EDT
966    Tag: blume-20010619-instantiate
967    Description:
968    
969    This un-breaks the fix for bug 1432.
970    (The bug was originally fixed in 110.9 but I broke it again some
971    time after that.)
972    
973    ----------------------------------------------------------------------
974    Name: Matthias Blume
975    Date: 2001/06/19 17:25:00 EDT
976    Tag: blume-20010619-signals
977    Description:
978    
979    This should (hopefully) fix the long-standing signal handling bug.
980    (The runtime system was constructing a continuation record with an
981    incorrect descriptor which would cause the GC to drop data on the floor...)
982    
983    ----------------------------------------------------------------------
984    Name: Matthias Blume
985    Date: 2001/06/15 15:05:00 EDT
986    Tag: blume-20010615-moresparc
987    Description:
988    
989    Here is a short late-hour update related to Sparc c-calls:
990    
991     -- made handling of double-word arguments a bit smarter
992    
993     -- instruction selection phase tries to collapse certain clumsily
994        constructed ML-Trees; typical example:
995    
996            ADD(ty,ADD(_,e,LI d1),LI d2)  ->  ADD(ty,e,LI(d1+d2))
997    
998        This currently has no further impact on SML/NJ since mlriscGen does
999        not seem to generate such patterns in the first place, and c-calls
1000        (which did generate them in the beginning) has meanwhile been fixed
1001        so as to avoid them as well.
1002    
1003    ----------------------------------------------------------------------
1004    Name: Matthias Blume
1005    Date: 2001/06/15 15:05:00 EDT
1006    Tag: blume-20010615-sparc
1007    Description:
1008    
1009    The purpose of this update is to provide an implementation of NLFFI
1010    on Sparc machines.
1011    
1012    Here are the changes in detail:
1013    
1014       * src/MLRISC/sparc/c-calls/sparc-c-calls.sml is a new file containing
1015       the Sparc implementation of the c-calls API.
1016       * The Sparc backend of SML/NJ has been modified to uniformely use %fp
1017       for accessing the ML frame.  Thus, we have a real frame pointer and
1018       can freely modify %sp without need for an omit-frame-ptr phase.
1019       The vfp logic in src/compiler/CodeGen/* has been changed to accomodate
1020       this case.
1021       * ml-nlffigen has been taught to produce code for different architectures
1022       and calling conventions.
1023       * In a way similar to what was done in the x86 case, the Sparc
1024       backend uses its own specific extension to mltree.  (For example,
1025       it needs to be able to generate UNIMP instructions which are part
1026       of the calling convention.)
1027       * ml-nlffi-lib was reorganized to make it more modular (in particular,
1028       to make it easier to plug in new machine- and os-dependent parts).
1029    
1030    There are some other fairly unrelated bug fixes and cleanups as well:
1031    
1032       * I further hacked the .cm files for MLRISC tools (like MDLGen) so
1033       that they properly share their libraries with existing SML/NJ libraries.
1034       * I fixed a minor cosmetic bug in CM, supressing certain spurious
1035       follow-up error messages.
1036       * Updates to CM/CMB documentation.
1037    
1038    TODO items:
1039    
1040       * MLRISC should use a different register as its asmTemp on the Sparc.
1041         (The current %o2 is a really bad choice because it is part of the
1042         calling conventions, so things might interfere in unexpected ways.)
1043    
1044    ----------------------------------------------------------------------
1045    Name: Matthias Blume
1046    Date: 2001/06/07
1047    Tag: blume-20010607-calls
1048    Description:
1049    
1050    A number of internal changes related to C calls and calling conventions:
1051    
1052    1. ML-Tree CALL statements now carry a "pops" field.  It indicates the
1053       number of bytes popped implicitly (by the callee).  In most cases
1054       this field is 0 but on x86/win32 it is some non-zero value.  This
1055       is information provided for the benefit of the "omit-frameptr" pass.
1056    2. The CALL instruction on the x86 carries a similar "pops" field.
1057       The instruction selection phase copies its value from the ML-Tree
1058       CALL statement.
1059    3. On all other architectures, the instruction selection phase checks
1060       whether "pops=0" and complains if not.
1061    4. The c-calls implementation for x86 now accepts two calling conventions:
1062       "ccall" and "stdcall".  When "ccall" is selected, the caller cleans
1063       up after the call and pops is set to 0.  For "stdcall", the caller
1064       does nothing, leaving the cleanup to the callee; pops is set to
1065       the number of bytes that were pushed onto the stack.
1066    5. The cproto decoder (compiler/Semant/types/cproto.sml) now can
1067       distinguish between "ccall" and "stdcall".
1068    6. The UNIMP instruction has been added to the supported Sparc instruction
1069       set. (This is needed for implementing the official C calling convention
1070       on this architecture.)
1071    7. I fixed some of the .cm files under src/MLRISC/Tools to make them
1072       work with the latest CM.
1073    
1074    ----------------------------------------------------------------------
1075    Name: Matthias Blume
1076    Date: 2001/06/05 15:10:00 EDT
1077    Tag: blume-20010605-cm-index
1078    Description:
1079    
1080    0. The "lambdasplit" parameter for class "sml" in CM has been documented.
1081    
1082    1. CM can now generate "index files".  These are human-readable files
1083       that list on a per-.cm-file basis each toplevel symbol defined or
1084       imported.  The location of the index file for
1085       <p>/<d>.cm is <p>/CM/INDEX/<d>.cm.
1086       To enable index-file generation, set CM.Control.generate_index to true
1087       or export an environment-symbol: export CM_GENERATE_INDEX=true.
1088    
1089       The CM manual has been updated accordingly.
1090    
1091    2. I made some slight modifications to the c-calls API in MLRISC.
1092    
1093         a) There is now a callback to support saving/restoring of
1094            dedicated but caller-save registers around the actual call
1095            instruction.
1096         b) One can optionally specify a comment-annotation for the
1097            call instruction.
1098    
1099    3. SML/NJ (mlriscGen.sml) uses this new API for the rawccall primop.
1100       (For example, the comment annotation shows the C prototype of
1101        the function being called.)
1102    
1103    ----------------------------------------------------------------------
1104    Name: Matthias Blume
1105    Date: 2001/06/01 13:30:00 EDT
1106    Tag: blume-20010601-nlffi-cleanup
1107    Description:
1108    
1109    This is mostly a cleanup of MLFFI stuff:
1110    
1111       - some signature files have been put into a more exposed place
1112       - the ugly 'f type parameter is gone (simplifies types tremendously!)
1113       - ml-nlffigen changed accordingly
1114       - tutorial updated
1115    
1116    Other changes:
1117    
1118       - author's affiliation in CM manual(s) updated
1119       - some more recognized keywords added to Allen's sml.sty
1120    
1121    ----------------------------------------------------------------------
1122    Name: Matthias Blume
1123    Date: 2001/05/25 15:30:00 EDT
1124    Tag: blume-20010525-iptr
1125    Description:
1126    
1127      - put the official 110.33-README (as it appears on the ftp server) under
1128        CVS
1129      - fixed a small bug related to incomplete pointer types in
1130        ml-nlffigen
1131      - small cosmetic change to the ml-nlffi-lib's "arr" type constructor
1132        (it does not need the 'f type parameter)
1133    
1134    ----------------------------------------------------------------------
1135    Name: Matthias Blume
1136    Date: 2001/05/23 14:30:00 EDT
1137    Tag: Release_110_33
1138    Description:
1139    
1140    New version number (110.33).  New bootfiles.
1141    
1142    ----------------------------------------------------------------------
1143    Name: Matthias Blume
1144    Date: 2001/05/22 18:06:00 EDT
1145    Tag: blume-20010522-targets
1146    Description:
1147    
1148    Made install.sh use file config/targets.customized if it exists, falling
1149    back to config/targets if it doesn't.  This way one can have a customized
1150    version of the targets file without touching the "real thing", thus
1151    eliminating the constant fear of accidentally checking something bogus
1152    back into the CVS repository...  (File config/targets.customized must
1153    not be added to the repository!)
1154    
1155    ----------------------------------------------------------------------
1156    Name: Matthias Blume
1157    Date: 2001/05/22 16:30:00 EDT
1158    Tag: blume-20010522-minitut
1159    Description:
1160    
1161    1. Bug fix in ml-nlffigen; now (hopefully) correctly handling
1162       struct returns.
1163    2. Added src/ml-nlffi-lib/Doc/mini-tutorial.txt.  This is some very
1164       incomplete, preliminary documentation for NLFFI.
1165    
1166    ----------------------------------------------------------------------
1167    Name: Matthias Blume
1168    Date: 2001/05/14 11:30:00 EDT
1169    Tag: blume-20010514-script
1170    Description:
1171    
1172    Some bugs in install script fixed.
1173    
1174    In addition to that I also made a slight change to the NLFFI API:
1175    Functors generated by ml-nlffigen now take the dynamic library as a
1176    straight functor argument, not as a suspended one.  (The original
1177    functor code used to force the suspension right away anyway, so there
1178    was nothing gained by this complication of the interface.)
1179    
1180    ----------------------------------------------------------------------
1181    Name: Matthias Blume
1182    Date: 2001/05/11 14:35:00 EDT
1183    Tag: blume-20010511-ml-nlffi
1184    Description:
1185    
1186    I finally took the plunge and added my new FFI code to the main
1187    repository.  For x86-linux it is now ready for prime-time.
1188    
1189    There are two new subdirectories of "src":
1190    
1191      - ml-nlffi-lib:
1192           The utility library for programs using the FFI interface.
1193           Here is the implementation of $/c.cm and its associated low-level
1194           partners $/c-int.cm and $/memory.cm.
1195      - ml-nlffigen:
1196           A stand-alone program for generating ML glue code from C source
1197           code.
1198    
1199    Building ml-nlffigen requires $/ckit-lib.cm.
1200    
1201    The config/install.sh script has been updates to do the Right Thing
1202    (hopefully).
1203    
1204    Notice that the source tree for the C-Kit will not be put under "src"
1205    but directly under the installation root directory.  (This is the
1206    structure that currently exists on the CVS server when you check out
1207    module "sml".)  Fortunately, config/install.sh knows about this oddity.
1208    
1209    Bugs: No documentation yet.
1210    
1211    ----------------------------------------------------------------------
1212    Name: Matthias Blume
1213    Date: 2001/05/09 16:35:00 EDT
1214    Tag: blume-20010509-cpscontract
1215    Description:
1216    
1217    Fixed a bug in the accounting code in cpsopt/contract.sml.  (The
1218    wrapper/unwrapper elimination did not decrement usage counts and some
1219    dead variables got overlooked by the dead-up logic.)
1220    
1221    ----------------------------------------------------------------------
1222    Name: Lal George
1223    Date: 2001/05/08  17:26:09 EDT
1224    Tag: george-20010508-omit-frameptr
1225    Description:
1226    
1227    Changes to implement the omit-frame-pointer optimization to support
1228    raw C calls. For now, there is only support on the Intel x86, but
1229    other architectures will follow as more experience is gained with this.
1230    
1231    
1232    ----------------------------------------------------------------------
1233    Name: Matthias Blume
1234    Date: 2001/05/07 14:40:00 EDT
1235    Tag: blume-20010507-proxies
1236    Description:
1237    
1238    I made into "proxy libraries" all libraries that qualify for such a
1239    change.  (A qualifying library is a library that has another library or
1240    groups as its sole member and repeats that member's export list
1241    verbatim.  A proxy library avoids this repetition by omitting its export
1242    list, effectively inheriting the list that its (only) member exports.
1243    See the CM manual for more explanation.)
1244    The main effect is that explicit export lists for these libraries
1245    do not have to be kepts in sync, making maintenance a bit easier.
1246    
1247    I also added copyright notices to many .cm-files.
1248    
1249    Last but not least, I made a new set of bootfiles.
1250    
1251    ----------------------------------------------------------------------
1252    Name: Matthias Blume
1253    Date: 2001/05/04 17:00:00 EDT
1254    Tag: blume-20010504-cm-lsplit
1255    Description:
1256    
1257    0. John merged pending changes to $/smlnj-lib.cm
1258    
1259    1. Allen's previous change accidentally backed out of one of Lal's
1260       earlier changes.  I undid this mistake (re-introducing Lal's change).
1261    
1262    2. I used the new topOrder' function from graph-scc.sml (from $/smlnj-lib.cm)
1263       within the compiler where applicable.  There is some code simplification
1264       because of that.
1265    
1266    3. The "split" phase (in FLINT) is now part of the default list of phases.
1267       Compiler.Control.LambdaSplitting.* can be used to globally control the
1268       lambda-splitting (cross-module-inlining) engine.  In addition to that,
1269       it can now also be controlled on a per-source basis: CM has been taught
1270       a new tool parameter applicable to ML source files.
1271    
1272       - To turn lambda-splitting off completely:
1273            local open Compiler.Control.LambdaSplitting in
1274                val _ = set Off
1275            end
1276       - To make "no lambda-splitting" the global default (but allow per-source
1277         overriding); this is the initial setting:
1278            local open Compiler.Control.LambdaSplitting in
1279                val _ = set (Default NONE)
1280            end
1281       - To make "lambda-splitting with aggressiveness a" the global default
1282         (and allow per-source overriding):
1283            local open Compiler.Control.LambdaSplitting in
1284                val _ = set (Default (SOME a))
1285            end
1286    
1287       - To turn lambda-splitting off for a given ML souce file (say: a.sml)
1288         write (in the respective .cm-file):
1289            a.sml (lambdasplitting:off)
1290       - To turn lambda-splitting for a.sml on with minimal aggressiveness:
1291            a.sml (lambdasplitting:on)
1292       - To turn lambda-splitting for a.sml on with aggressiveness <a> (where
1293         <a> is a decimal non-negative integer):
1294            a.sml (lambdasplitting:<a>)
1295       - To turn lambda-splitting for a.sml on with maximal aggressiveness:
1296            a.sml (lambdasplitting:infinity)
1297       - To use the global default for a.sml:
1298            a.sml (lambdasplitting:default)
1299         or simply
1300            a.sml
1301    
1302    ----------------------------------------------------------------------
1303    Name: Allen Leung
1304    Date: 2001/05/04 01:57:00 EDT
1305    Tag: leunga-20010504-sync
1306    Description:
1307    
1308      MLRISC features.
1309    
1310      1. Fix to CMPXCHG instructions.
1311      2. Changed RA interface to allow annotations in callbacks.
1312      3. Added a new method to the stream interface to allow annotations updates.
1313    
1314    ----------------------------------------------------------------------
1315    Name: Matthias Blume
1316    Date: 2001/05/01 11:45:00 EDT
1317    Tag: blume-20010501-pcedittmp
1318    Description:
1319    
1320    Changed install.sh to use the current working directory instead of
1321    /usr/tmp for a temporary file (pcedittmp).  The previous choice
1322    of /usr/tmp caused trouble with MacOS X because of file premission
1323    problems.
1324    
1325    ----------------------------------------------------------------------
1326    Name: Matthias Blume
1327    Date: 2001/04/20 11:10:00 EDT
1328    Tag: blume-20010420-inMLflag
1329    Description:
1330    
1331     - added vp_limitPtrMask to vproc-state.h
1332       (for use by the raw-C-calls mechanism to implement proper interrupt
1333        handling)
1334     - made the ML compiler aware of various data-structure offsets so it
1335       can generate code for accessing the vp_inML flag and vp_limitPtrMask
1336     - tweaked mlriscGen.sml to have it emit interrupt-handling code for
1337       raw C-calls
1338    
1339    ----------------------------------------------------------------------
1340    Name: Lal George
1341    Date: 2001/04/20 09:15:28 EDT
1342    Tag: george-20010420-macosX
1343    Description:
1344    
1345     - Changes to port to Mac OS X; Darwin.
1346    
1347     - In the process I found that sqrt was broken on the PPC, because the
1348       fsqrt instruction is not implemented.
1349    
1350    ----------------------------------------------------------------------
1351    Name: Matthias Blume
1352    Date: 2001/04/18 12:45:00 EDT
1353    Tag: blume-20010418-ccalls
1354    Description:
1355    
1356     - fixed two off-by-4 errors in the x86-specific c-calls implementation
1357       (this bug prevented structure arguments containing pointers from being
1358        passed correctly)
1359     - changed the raw-C-call code in mlriscGen.sml in such a way that
1360       structure arguments are represented as a pointer to the beginning
1361       of the structure (instead of having a series of synthesized arguments,
1362       one for each structure member)
1363    
1364     - made makeml script's verbosity level configurable via environment
1365       variable (MAKEML_VERBOSITY)
1366    
1367     - eliminated placeholder implementations for f32l, w16s, i16s, and f32s
1368       in rawmem-x86.sml; we are now using the real thing
1369    
1370    ----------------------------------------------------------------------
1371    Name: Matthias Blume
1372    Date: 2001/03/22 16:25:00 EST
1373    Tag: blume-20010322-bootfiles
1374    Description:
1375    
1376    Created a new set of bootfiles (for your automatic installation convenience).
1377    
1378    ----------------------------------------------------------------------
1379    Name: Matthias Blume
1380    Date: 2001/03/22 15:10:00 EST
1381    Tag: blume-20010322-rawmem-parcm
1382    Description:
1383    
1384    1. All "raw memory access" primitives for the new FFI are implemented now
1385       (at least on the x86).
1386    2. Some further cleanup of CM's parallel make mechanism.
1387    
1388    ----------------------------------------------------------------------
1389    Name: Matthias Blume
1390    Date: 2001/03/19 17:53:00 EST
1391    Tag: blume-20010319-parallel
1392    Description:
1393    
1394    Parallel make (using compile servers) now works again.
1395    
1396    To this end, CM.stabilize and CMB.make have been modified to work in
1397    two passes when compile servers are attached:
1398       1. Compile everything, do not perform stabilization; this pass
1399          uses compile servers
1400       2. Stabilize everything; this pass does not use compile servers
1401    If there are no compile servers, the two passes are combined into one
1402    (as before).  Splitting the passes increases the inherent parallelism
1403    in the dependency graph because the entire graph including all
1404    libraries is available at the same time.  This, in turn, improves
1405    server utilization.  The downside is that the master process will
1406    have to do some extra work after compilation is done (because for
1407    technical reasons it must re-read all the binfiles during stabilization).
1408    
1409    ----------------------------------------------------------------------
1410    Name: Matthias Blume
1411    Date: 2001/03/16 12:22:00 EST
1412    Tag: blume-20010316-bootfiles
1413    Description:
1414    
1415    Created a new set of bootfiles (for your automatic installation convenience).
1416    
1417    ----------------------------------------------------------------------
1418    Name: Matthias Blume
1419    Date: 2001/03/16 11:00:00 EST
1420    Tag: blume-20010316-MLTREE-fixup
1421    Description:
1422    
1423    This is a minor fixup for an (untagged) earlier commit by Allen.
1424    (A file was missing).
1425    
1426    ----------------------------------------------------------------------
1427    Name: Allen Leung
1428    Date: Mon Mar  5 18:54:57 EST 2001
1429    Tag: leunga-20010305-cut-support
1430    
1431    1. New support for alternative control-flow in MLTREE.
1432       Currently we support
1433    
1434          FLOW_TO(CALL ...., [k1,...,kn])
1435    
1436       This is needed for 'cuts to' in C-- and try/handle-like constructs
1437       in Moby
1438    
1439       New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
1440    
1441    2. Register Allocator
1442    
1443       Changes in interface [from Fermin, John]
1444    
1445    3. Alpha 8-bit SLL support [Fermin]
1446    
1447    4. All architectures
1448    
1449       A new module (ClusterExpandCopies) for expanding parallel copies.
1450    
1451    ----------------------------------------------------------------------
1452    Name: Allen Leung
1453    Date: 2001/02/27 23:07:00 EST
1454    Tag: leunga-20010227-minor-stuff
1455    
1456    1. Alpha bug fix for CMOVNE
1457    2. Handle mltree COND(..,FCMP ...,...)
1458    3. Bug fix in simplifier
1459    
1460    ----------------------------------------------------------------------
1461    Name: Matthias Blume
1462    Date: 2001/01/30 17:50:00 EST
1463    Tag: blume-20010130-sync
1464    Description:
1465    
1466    This is just a minor update to sync my devel branch with the main brach.
1467    The only visible change is the addition of some README files.
1468    
1469    ----------------------------------------------------------------------
1470    Name: Matthias Blume
1471    Date: 2001/01/12 23:30:00 JST
1472    Tag: blume-20010112-bootfiles
1473    Description:
1474    
1475    Made a new set of bootfiles that goes with the current state of the
1476    repository.
1477    
1478    ----------------------------------------------------------------------
1479    Name: Matthias Blume
1480    Date: 2001/01/12 21:20:00 JST
1481    Tag: blume-20010112-sync
1482    Description:
1483    
1484    I am just flushing out some minor changes that had accumulated in
1485    my private branch in order to sync with the main tree.  (This is
1486    mainly because I had CVS trouble when trying to merge _into_ my
1487    private branch.)
1488    
1489    Most people should be completely unaffected by this.
1490    
1491    ----------------------------------------------------------------------
1492    Name: Allen Leung
1493    Date: Thu Jan 11 21:03:00 EST 2001
1494    Tag: leunga-20010111-labexp=mltree
1495    Description:
1496    
1497    1.  Removed the type LabelExp and replace it by MLTree.
1498    2.  Rewritten mltree-simplify with the pattern matcher tool.
1499    3.  There were some bugs in alpha code generator which would break
1500        64-bit code generation.
1501    4.  Redo the tools to generate code with the
1502    5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
1503        are now generated by perl scripts.
1504    
1505    ----------------------------------------------------------------------
1506    Name: Matthias Blume
1507    Date: 2001/01/10 21:55:00 JST
1508    Tag: blume-20010110-rcc
1509    Description:
1510    
1511    The RCC stuff now seems to work (but only on the x86).
1512    This required hacking of the c-calls interface (and -implementation) in
1513    MLRISC.
1514    
1515    Normal compiler users should be unaffected.
1516    
1517    ----------------------------------------------------------------------
1518    Name: Matthias Blume
1519    Date: 2001/01/09 01:20:00 JST
1520    Tag: blume-20010109-rcc
1521    Description:
1522    
1523    This is a fairly big patch, flushing out a large number of pending
1524    changes that I made to my development copy over the last couple of days.
1525    
1526    Of practical relevance at this moment is a workaround for a pickling
1527    bug that Allen ran into the other day.  The cause of the bug itself is
1528    still unknown and it might be hard to fix it properly, but the
1529    workaround has some merits of its own (namely somewhat reducing pickling
1530    overhead for certain libraries).  Therefore, I think this solution should
1531    be satisfactory at this time.
1532    
1533    The rest of the changes (i.e., the vast majority) has to do with my
1534    ongoing efforts of providing direct support for C function calls from
1535    ML.  At the moment there is a new primop "RAW_CCALL", typing magic
1536    in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
1537    case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
1538    translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
1539    handle RCC.
1540    
1541    The last part (the changes to mlriscGen.sml) are still known to be
1542    wrong on the x86 and not implemented on all other architectures.  But
1543    the infrastructure is in place. I had to change a few functor
1544    signatures in the backend to be able to route the CCalls interface
1545    from MLRISC there, and I had to specialize the mltree type (on the
1546    x86) to include the necessary extensions. (The extensions themselves
1547    were already there and redy to go in MLRISC/x86).
1548    
1549    Everything should be very happy as soon as someone helps me with
1550    mlriscGen.sml...
1551    
1552    In any case, nothing of this should matter to anyone as long as the
1553    new primop is not being used (which is going to be the case unless you
1554    find it where I hid it :). The rest of the compiler is completely
1555    unaffected.
1556    
1557    ----------------------------------------------------------------------
1558    Name: Matthias Blume
1559    Date: 2001/01/05 00:30:00 JST
1560    Tag: blume-20010105-primops
1561    Description:
1562    
1563    Added some experimental support for work that I am doing right now.
1564    These changes mostly concern added primops, but there is also a new
1565    experimental C library in the runtime system (but currently not enabled
1566    anywhere except on Linux/X86).
1567    
1568    In the course of adding primops (and playing with them), I discovered that
1569    Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
1570    broken.  (Zhong was very right he labeled this stuff as "major gross hack".)
1571    To recover, I made type information in INL_PRIM mandatory and changed
1572    prim.sml as well as built-in.sml accordingly.  The InLine structure now
1573    has complete, correct type information (i.e., no bottom types).
1574    
1575    Since all these changes mean that we need new binfiles, I also bumped the
1576    version number to 110.32.1.
1577    
1578    ----------------------------------------------------------------------
1579    Name: Matthias Blume
1580    Date: 2000/12/30 22:10:00 JST
1581    Tag: blume-20001230-various
1582    Description:
1583    
1584    Added proxy libraries for MLRISC and let MLRISC libraries refer
1585    to each other using path anchors.  (See CM manual for explanation.)
1586    
1587    Updated CM documentation.
1588    
1589    Fixed some bugs in CM.
1590    
1591    Implemented "proxy" libraries (= syntactic sugar for CM).
1592    
1593    Added "-quiet" option to makeml and changed runtime system accordingly.
1594    
1595    Added cleanup handler for exportML to reset timers and compiler stats.
1596    
1597    ----------------------------------------------------------------------
1598    Name: Lal George
1599    Date: 2000/12/22 22:22:58 EST 2000
1600    Tag: Release_110_32
1601    Description:
1602    
1603            Infinite precision used throughout MLRISC.
1604            see MLRISC/mltree/machine-int.sig
1605    
1606    ----------------------------------------------------------------------
1607    Name: Matthias Blume
1608    Date: 2000/12/22 23:16:00 JST
1609    Tag: blume-20001222-warn
1610    Description:
1611    
1612    Corrected wording and formatting of some CM warning message which I
1613    broke in my previous patch.
1614    
1615    ----------------------------------------------------------------------
1616    Name: Matthias Blume
1617    Date: 2000/12/22 21:20:00 JST
1618    Tag: blume-20001222-anchorenv
1619    Description:
1620    
1621    Fixed CM's handling of anchor environments in connection with CMB.make.
1622    
1623    ----------------------------------------------------------------------
1624    Name: Matthias Blume
1625    Date: 2000/12/22 13:15:00 JST
1626    Tag: blume-20001222-cleanup
1627    Description:
1628    
1629    Removed src/cm/ffi which does not (and did not) belong here.
1630    
1631    ----------------------------------------------------------------------
1632    Name: Matthias Blume
1633    Date: 2000/12/21 23:55:00 JST
1634    Tag: blume-20001221-exn
1635    Description:
1636    
1637    Probably most important: CM no longer silently swallows all exceptions
1638    in the compiler.
1639    Plus: some other minor CM changes.  For example, CM now reports some
1640    sizes for generated binfiles (code, data, envpickle, lambdapickle).
1641    
1642    ----------------------------------------------------------------------
1643    Name: Matthias Blume
1644    Date: 2000/12/15 00:01:05 JST
1645    Tag: blume-20001215-dirtool
1646    Description:
1647    
1648    - "dir" tool added.
1649    - improvements and cleanup to Tools structure
1650    - documentation updates
1651    
1652    ----------------------------------------------------------------------
1653    Name: Allen Leung
1654    Date: Thu Dec 14 03:45:24 EST 2000
1655    Description:
1656    Tag:  leunga-20001214-int-inf
1657    Description:
1658    
1659       In IntInf, added these standard functions, which are missing from our
1660    implementation:
1661    
1662        andb : int * int -> int
1663        xorb : int * int -> int
1664        orb  : int * int -> int
1665        notb : int -> int
1666         <<   : int * word -> int
1667        ~>>  : int * word -> int
1668    
1669       Not tested, I hope they are correct.
1670    
1671    ----------------------------------------------------------------------
1672    Name: Allen Leung
1673    Date: Fri Dec  8 19:23:26 EST 2000
1674    Description:
1675    Tag:  leunga-20001208-nowhere
1676    Description:
1677    
1678      Slight improvements to the 'nowhere' tool to handle OR-patterns,
1679    to generate better error messages etc.  Plus a brief manual.
1680    
1681    ----------------------------------------------------------------------
1682    Name: Lal George
1683    Date: 2000/12/08 09:54:02 EST 2000
1684    Tag: Release_110_31
1685    Description:
1686    
1687    - Version 110.31
1688    ----------------------------------------------------------------------
1689    Name: Allen Leung
1690    Date: Thu Dec  7 22:01:04 EST 2000
1691    Tag:  leunga-20001207-cell-monster-hack
1692    Description:
1693    
1694    Major MLRISC internal changes.  Affect all clients.
1695    Summary:
1696    
1697    1.  Type CELLS.cell = int is now replaced by a datatype.
1698        As a result, the old regmap is now gone.  Almost all interfaces
1699        in MLRISC change as a consequence.
1700    
1701    2.  A new brand version of machine description tool (v3.0) that generates
1702        modules expecting the new interface.  The old version is removed.
1703    
1704    3.  The RA interface has been further abstracted into two new functors.
1705        RISC_RA and X86RA.  These functors have much simpler interfaces.
1706        [See also directory MLRISC/demo.]
1707    
1708    4.  Some other new source->source code generation tools are available:
1709    
1710        a. MLRISC/Tools/RewriteGen -- generate rewriters from rules.
1711        b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules.
1712           I use this tool to generate the peephole optimizers---with the new
1713           cell type changes, peephole rules are becoming difficult to write
1714           without conditional pattern matching.
1715    
1716    5.  More Intmap -> IntHashTable change.  Previous changes by Matthias didn't
1717        cover the entire MLRISC source tree so many things broke.
1718    
1719    6.  CM files have been moved to the subdirectory MLRISC/cm.
1720        They are moved because there are a lot of them and they clutter up the
1721        root dir.
1722    
1723    7.  More detailed documentation to come...
1724    
1725        NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do
1726        a makeml -rebuild first.  This is because of other other
1727        changes that Matthias has made (see below).
1728    
1729    
1730    ----------------------------------------------------------------------
1731    Name: Matthias Blume
1732    Date: 2000/11/30 23:12:00 JST
1733    Tag: blume-20001130-filereorg
1734    Description:
1735    
1736    Some manual updates and some file reorganizations in CM.
1737    
1738    ----------------------------------------------------------------------
1739    Name: Matthias Blume
1740    Date: 2000/11/24 17:45:00 JST
1741    Tag: blume-20001124-link
1742    Description:
1743    
1744    Drastically improved link traversal code for the case that the dynamic
1745    value was already loaded at bootstrap time.  As a result, CM and CMB
1746    now both load blazingly fast -- even on a very slow machine.  Also,
1747    memory consumption has been further reduced by this.
1748    
1749    Warning: The format of the PIDMAP file has changed.  THerefore, to
1750    bootstrap you have to do this:
1751    
1752    1. Run CMB.make
1753    2. Make a symbolic link for the boot directory:
1754         ln -s sml.boot.ARCH-OS xxx
1755    3. "Rebuild" the boot directory:
1756         ./makeml -boot xxx -rebuild sml ; rm xxx
1757    4. Boot normally:
1758          ./makeml
1759    
1760    ----------------------------------------------------------------------
1761    Name: Matthias Blume
1762    Date: 2000/11/21 21:20:00 JST
1763    Tag: blume-20001121-tools
1764    Description:
1765    
1766    Continued hacking on autoloading problem -- with success this time.
1767    Also changed tool-plugin mechanism.  See new CM manual.
1768    
1769    ----------------------------------------------------------------------
1770    Name: Matthias Blume
1771    Date: 2000/11/19 14:30:00 JST
1772    Tag:  blume-20001119-autoload
1773    Description:
1774    
1775    Some hacking to make autoloading faster.  Success for CMB, no success
1776    so far for CM.  There is a reduced structure CM' that autoloads faster.
1777    (This is a temporary, non-documented hack to be eliminated again when
1778    the general problem is solved.)
1779    
1780    ----------------------------------------------------------------------
1781    Name: Matthias Blume
1782    Date: 2000/11/17 14:10:00 JST
1783    Tag: blume-20001117-pickle-lib
1784    Description:
1785    
1786    1. Eliminated comp-lib.cm
1787    2. Made pickle-lib.cm
1788    3. Eliminated all uses of intset.sml (from comp-lib.cm)
1789    4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
1790       equivalent constructs from smlnj-lib.cm (INtHashTable).
1791    5. Point 4. also goes for those uses of intmap.* in MLRISC.
1792       Duplicated intmap modules thrown out.
1793    6. Hunted down all duplicated SCC code and replaced it with
1794       equivalent stuff (GraphSCCFn from smlnj-lib.cm).
1795    7. Rewrote Feedback module.
1796    8. Moved sortedlist.sml into viscomp-lib.cm.  Eventually it
1797       should be thrown out and equivalent modules from smlnj-lib.cm
1798       should be used (IntRedBlackSet, IntListSet, ...).
1799    
1800    Confirmed that compiler compiles to fixpoint.
1801    
1802    ----------------------------------------------------------------------
1803    Name: Allen Leung
1804    Date: 2000/11/10 18:00:00
1805    Tag: leunga-20001110-new-x86-fp
1806    
1807    A new x86 floating point code generator has been added.
1808    By default this is turned off.  To turn this on, do:
1809    
1810        CM.autoload "$smlnj/compiler.cm";
1811        Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
1812    
1813    Changes:
1814    
1815    1.  Changed FTAN to FPTAN so that the assembly output is correct.
1816    2.  Changed the extension callback for FTANGENT to generate:
1817    
1818              fptan
1819              fstp  %st(0)
1820        instead of
1821              fptan
1822              fstpl ftempmem
1823    
1824    3.  Numerous assembly fixes for x86.
1825    
1826    5.  Cleaned up the machine code output module x86/x86MC.sml and added
1827        support for a whole bunch of instructions and addressing modes:
1828    
1829          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st, %st(n)
1830          faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp  %st, %st(n)
1831          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st(n), %st
1832          fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
1833          fxch %st(n)
1834          fld %st(n)
1835          fst %st(n)
1836          fst mem
1837          fstp %st(n)
1838          fucom %st(n)
1839          fucomp %st(n)
1840    
1841        All these are now generated when the fast fp mode is turned on.
1842    
1843    6.  Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
1844    
1845    ----------------------------------------------------------------------
1846    Name: Matthias Blume
1847    Date: 2000/11/09 11:20:00 JST
1848    Tag: blume-20001109-scc
1849    Description:
1850    
1851    Eliminated some code duplication:
1852    
1853    1. Added "where" clause to GraphSCCFn in SML/NJ Library.
1854       (Otherwise the functor is useless.)
1855    2. Used GraphSCCFn where SCCUtilFun was used previously.
1856    3. Got rid of SCCUtilFun (in comp-lib.cm).
1857    
1858    ----------------------------------------------------------------------
1859    Name: Lal George
1860    Date: 2000/11/06 09:02:21 EST 2000
1861    Tag: Release_110_30
1862    Description:
1863    
1864    - Version 110.30
1865    ----------------------------------------------------------------------
1866    Name: Matthias Blume
1867    Date: 2000/11/04 14:45:00
1868    Tag: blume-20001104-mlbuild
1869    Description:
1870    
1871    - Made ml-build faster on startup.
1872    - Documentation fixes.
1873    
1874    ----------------------------------------------------------------------
1875    Name: Matthias Blume
1876    Date: 2000/11/02 17:00:00 JST
1877    Tag: blume-20001102-condcomp
1878    Description:
1879    
1880    - Small tweaks to pickler -- new BOOTFILES!
1881    - Version bumped to 110.29.2.
1882    - Added conditional compilation facility to init.cmi (see comment there).
1883    ----------------------------------------------------------------------
1884    Name: Allen Leung
1885    Date: 2000/10/23 19:31:00
1886    Tag: leunga-20001023-demo-ra
1887    
1888    1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
1889    2. Test programs for the graph library updated
1890    3. Some new MLRISC demo programs added
1891    
1892    ----------------------------------------------------------------------
1893    Name: Matthias Blume
1894    Date: 2000/08/31 22:15:00 JST
1895    Tag: blume-20001017-errmsg
1896    Description:
1897    
1898    More error message grief: Where there used to be no messages, there
1899    now were some that had bogus error regions.  Fixed.
1900    
1901    ----------------------------------------------------------------------
1902    Name: Matthias Blume
1903    Date: 2000/08/31 17:30:00 JST
1904    Tag: blume-20001017-v110p29p1
1905    Description:
1906    
1907    I made a version 110.29.1 with new bootfiles.
1908    
1909    Changes:  Modified pickler/unpickler for faster and leaner unpickling.
1910              CM documentation changes and a small bugfix in CM's error reporting.
1911    
1912    ----------------------------------------------------------------------
1913    Name: Lal George
1914    Date: 2000/09/27 14:42:35 EDT
1915    Tag: george-20000927-nodestatus
1916    Description:
1917    
1918    Changed the type of the nodestatus, so that:
1919    
1920            SPILLED(~1)             is now SPILLED
1921            SPILLED(m) where m>=0   is now MEMREG(m)
1922            SPILLED(s) where s<~1   is now SPILL_LOC(~s)
1923    
1924    ----------------------------------------------------------------------
1925    Name: Matthias Blume
1926    Date: 2000/09/07 14:45:00 JST
1927    Tag: blume-20000907-cmerrmsg
1928    Description:
1929    
1930    Small tweak to CM to avoid getting ML syntax error messages twice.
1931    
1932    ----------------------------------------------------------------------
1933    Name: Matthias Blume
1934    Date: 2000/08/31 18:00:00 JST
1935    Tag: blume-20000831-cvsbootfiles
1936    Description:
1937    
1938    New URL for boot files (because the 110.29 files on the BL server do
1939    now work correctly with my updated install scripts for yacc and lex).
1940    
1941    ----------------------------------------------------------------------
1942    Name: Matthias Blume
1943    Date: 2000/08/08 12:33:00 JST
1944    Tag: blume-20000808-manual
1945    Description:
1946    
1947    Tiny update to CM manual.
1948    
1949    ----------------------------------------------------------------------
1950    Name: Allen Leung
1951    Date: 2000/08/7 19:31:00
1952    Tag: leunga-20000807-a-whole-bunch-of-stuff
1953    
1954      Moby, C--, SSA, x86, machine descriptions etc.  Should only affect C--
1955    and Mobdy.
1956    
1957    1.  x86
1958    
1959       a.  Fixes to peephole module by John and Dan.
1960       b.  Assembly fix to SETcc by Allen.
1961       c.  Fix to c-call by John.
1962       d.  Fix to spilling by John.  (This one deals with the missing FSTPT case)
1963       e.  Instruction selection optimization to SETcc as suggested by John.
1964    
1965           For example,
1966    
1967            MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
1968    
1969           should generate:
1970    
1971            MOVL a, x
1972            SUBL b, x
1973            SHRL 31, x
1974    
1975    2.  IR stuff
1976    
1977         A bunch of new DJ-graph related algorithms added.  These
1978         speed up SSA construction.
1979    
1980    3.  SSA + Scheduling
1981    
1982         Added code for SSA and scheduling to the repository
1983    
1984    ----------------------------------------------------------------------
1985    Name: Lal George
1986    Date: 2000/07/27 11:53:14 EDT
1987    
1988    Tag: lal-20000727-linux-ppc
1989    Description:
1990    
1991     Made changes to support Linux PPC.
1992     p.s. I have confirmation that the 110.29 boot files work fine.
1993    
1994    ----------------------------------------------------------------------
1995    Name: Matthias Blume
1996    Date: 2000/07/27 17:40:00 JST
1997    Tag: blume-20000727-scripts
1998    Description:
1999    
2000    !!!! WARNING !!!!
2001    You must recompile the runtime system!
2002    !!!! WARNING !!!!
2003    
2004    This is basically another round of script-enhancements:
2005    
2006    1. sml, ml-build, and ml-makedepend accept options -D and -U to define
2007       and undefine CM preprocessor symbols.
2008    
2009    2. ml-build avoids generating a new heap image if it finds that the
2010       existing one is still ok.  (The condition is that no ML file had to
2011       be recompiled and all ML files are found to be older that the heap
2012       file.)
2013    
2014       To make this work smoothly, I also hacked the runtime system as
2015       well as SMLofNJ.SysInfo to get access to the heap image suffix
2016       (.sparc-solaris, ...) that is currently being used.
2017    
2018       Moreover, the signature of CM.mk_standalone has changed.  See the
2019       CM manual.
2020    
2021    3. ml-makedepend accepts additional options -n, -a, and -o.  (See the
2022       CM manual for details.)
2023    
2024    4. More CM manual updates:
2025        - all of the above has been documented.
2026        - there is now a section describing the (CM-related) command line
2027          arguments that are accepted by the "sml" command
2028    
2029    ----------------------------------------------------------------------
2030    Name: Matthias Blume
2031    Date: 2000/07/25 16:20:00 JST
2032    Tag: blume-20000725-makedepend
2033    Description:
2034    
2035    Added a script called ml-makedepend.  This can be used in makefiles
2036    for Unix' make in a way very similar to the "makedepend" command for
2037    C.
2038    
2039    The script internally uses function CM.sources.
2040    
2041    Synopsis:
2042    
2043        ml-makedepend [-f makefile] cmfile targetname
2044    
2045    The default for the makefile is "makefile" (or "Makefile" should
2046    "makefile" not exist).
2047    
2048    ml-makedepend adds a cmfile/targetname-specific section to this
2049    makefile (after removing the previous version of this section).  The
2050    section contains a single dependency specification with targetname on
2051    the LHS (targetname is an arbitrary name), and a list of files derived
2052    from the cmfile on the RHS.  Some of the files on the RHS are
2053    ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
2054    "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
2055    path names.  The makefile writer is responsible for making sure that
2056    these variables have correct at the time "make" is invoked.
2057    
2058    ----------------------------------------------------------------------
2059    Name: Matthias Blume
2060    Date: 2000/07/22 23:30:00 JST
2061    Tag: blume-20000722-urlupdate
2062    Description:
2063    
2064    Changed BOOT and config/srcarchiveurl to point to BL server:
2065    
2066        ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
2067    
2068    ----------------------------------------------------------------------
2069    Name: Matthias Blume
2070    Date: 2000/07/18 18:00:00 JST
2071    Tag: blume-20000718-Version_110_29
2072    Description:
2073    
2074    1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
2075    
2076    2. Updated config/version to 110.29
2077    
2078    3. Updated config/srcarchiveurl
2079    
2080    3. New boot files!
2081       ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
2082    
2083    ----------------------------------------------------------------------
2084    Name: Matthias Blume
2085    Date: 2000/07/11 13:58:00 JST
2086    Tag: blume-20000711-doctypo
2087    Description:
2088    
2089    Fixed a few typos in CM manual.
2090    
2091    ----------------------------------------------------------------------
2092    Name: Allen Leung
2093    Date: 2000/06/15 00:38:00
2094    Tag: leunga-20000704-sparc-x86
2095    
2096    1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
2097    2. fix to x86 RET bug [found by Dan Grossman]
2098    3. sparc assembly bug fix for ticc instructions [found by Fermin]
2099    
2100       Affects c-- and moby only
2101    
2102    ----------------------------------------------------------------------
2103    Name: Matthias Blume
2104    Date: 2000/07/04 15:26:00
2105    Tag: blume-20000704-trigger
2106    Description:
2107    
2108    1. Improvements to CM manual.
2109    2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
2110       of getting a back-trace.  The function, when called, raises an
2111       internal exception which explicitly carries the full back-trace history,
2112       so it is unaffected by any intervening handle-raise pairs ("trivial"
2113       or not).  The interactive loop will print that history once it arrives
2114       at top level.
2115       Short of having all exceptions implicitly carry the full history, the
2116       recommended way of using this facility is:
2117         - compile your program with instrumentation "on"
2118         - run it, when it raises an exception, look at the history
2119         - if the history is "cut off" because of some handler, go and modify
2120           your program so that it explicitly calls BTrace.trigger
2121         - recompile (still instrumented), and rerun; look at the full history
2122    
2123    ----------------------------------------------------------------------
2124    Name: Matthias Blume
2125    Date: 2000/07/03 15:36:00 JST
2126    Tag: blume-20000702-manual
2127    Description:
2128    
2129    Small corrections and updates to CM manual.
2130    
2131    ----------------------------------------------------------------------
2132    Name: Matthias Blume
2133    Date: 2000/06/29 16:04:00 JST
2134    Tag: blume-20000629-yacctool
2135    Description:
2136    
2137    Changes:
2138    
2139    1. Class "mlyacc" now takes separate arguments to pass options to
2140       generated .sml- and .sig-files independently.
2141    2. Corresponding CM manual updates.
2142    3. BTrace module now also reports call sites.  (However, for loop clusters
2143       it only shows from where the cluster was entered.)  There are associated
2144       modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
2145    
2146    ----------------------------------------------------------------------
2147    Name: Matthias Blume
2148    Date: 2000/06/27 16:51:00 JST
2149    Tag: blume-20000627-noweb
2150    Description:
2151    
2152    Changes:
2153    
2154     1. Implemented "subdir" and "witness" options for noweb tool.
2155        This caused some slight internal changes in CM's tool implementation.
2156     2. Fixed bug in "tool plugin" mechanism.  This is essentially cleaning
2157        some remaining issues from earlier path anchor changes.
2158     3. Updated CM manual accordingly.
2159    
2160     4. Changed implementation of back-tracing so that I now consider it
2161        ready for prime-time.
2162    
2163        In particular, you don't have to explicitly trigger the back-trace
2164        anymore.  Instead, if you are running BTrace-instrumented code and
2165        there is an uncaught exception (regardless of whether or not it was
2166        raised in instrumented code), the top-level evalloop will print
2167        the back-trace.
2168    
2169        Features:
2170    
2171          - Instrumented and uninstrumented code work together seemlessly.
2172            (Of course, uninstrumented code is never mentioned in actual
2173             back-traces.)
2174    
2175          - Asymptotic time- and space-complexity of instrumented code is
2176            equal to that of uninstrumented code.  (This means that
2177            tail-recursion is preserved by the instrumentation phase.)
2178    
2179          - Modules whose code has been instrumented in different sessions
2180            work together without problem.
2181    
2182          - There is no penalty whatsoever on uninstrumented code.
2183    
2184          - There is no penalty on "raise" expressions, even in
2185            instrumented code.
2186    
2187        A potential bug (or perhaps it is a feature, too):
2188    
2189          A back-trace reaches no further than the outermost instrumented
2190          non-trivial "raise".  Here, a "trivial" raise is one that is the
2191          sole RHS of a "handle" rule.  Thus, back-traces reach trough
2192    
2193               <exp> handle e => raise e
2194    
2195          and even
2196    
2197               <exp> handle Foo => raise Bar
2198    
2199          and, of course, through
2200    
2201               <exp> handle Foo => ...
2202    
2203         if the exception was not Foo.
2204    
2205         Back-traces always reach right through any un-instrumented code
2206         including any of its "handle" expressions, trivial or not.
2207    
2208       To try this out, do the following:
2209    
2210         - Erase all existing binfiles for your program.
2211           (You may keep binfiles for those modules where you think you
2212            definitely don't need back-tracing.)
2213         - Turn on back-trace instrumentation:
2214              SMLofNJ.Internals.BTrace.mode (SOME true);
2215         - Recompile your program.  (I.e., run "CM.make" or "use".)
2216         - You may now turn instrumentation off again (if you want):
2217              SMLofNJ.Internals.BTrace.mode (SOME false);
2218         - Run your program as usual.  If it raises an exception that
2219           reaches the interactive toplevel, then a back-trace will
2220           automatically be printed.  After that, the toplevel loop
2221           will print the exception history as usual.
2222    
2223    ----------------------------------------------------------------------
2224    Name: Matthias Blume
2225    Date: 2000/06/26 09:56:46 JST
2226    Tag: blume-20000626-setup
2227    Description:
2228    
2229    CM: - setup-parameter to "sml" added; this can be used to run arbitrary
2230          ML code before and after compiling a file (e.g., to set compiler
2231          flags)
2232    
2233    Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
2234              - associated changes to btrace.sml (BTrace instrumentation pass)
2235              - cleaner implementation of btimp.sml (BTrace tracing and report
2236                module)
2237    
2238    CM manual: * new path encoding documented
2239               * description of setup-parameter to "sml" added
2240    
2241    The biggest user-visible change to back-tracing is that it is no
2242    longer necessary to compile all traced modules within the same
2243    session.  (This was a real limitation.)
2244    
2245    ----------------------------------------------------------------------
2246    Name: Matthias Blume
2247    Date: 2000/06/24 12:40:00 JST
2248    Tag: blume-20000624-startup
2249    Description:
2250    
2251    Fixes startup slowdown problem.  (I was calling SrcPath.sync a _tad_
2252    bit too often -- to put it mildly. :)
2253    
2254    ----------------------------------------------------------------------
2255    Name: Matthias Blume
2256    Date: 2000/06/23 18:20:00 JST
2257    Tag: blume-20000623-btrace
2258    Description:
2259    
2260    This updates adds a backtrace facility to aid programmers in debugging
2261    their programs.  This involves the following changes:
2262    
2263    1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
2264       keeping track of the current call stack.  When programs are compiled
2265       in a special mode, the compiler will insert calls to these hooks
2266       into the user program.
2267       "Hook" means that it is possible for different implementations of
2268       back-tracing to register themselves (at different times).
2269    
2270    2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
2271       as an Absyn.dec->Absyn.dec rewrite.  Normally this phase is turned off.
2272       It can be turned on using this call:
2273         SMLofNJ.Internals.BTrace.mode (SOME true);
2274       Turning it off again:
2275         SMLofNJ.Internals.BTrace.mode (SOME false);
2276       Querying the current status:
2277         SMLofNJ.Internals.BTrace.mode NONE;
2278       Annotated programs are about twice as big as normal ones, and they
2279       run a factor of 2 to 4 slower with a dummy back-trace plugin (one
2280       where all hooks do nothing).  The slowdown with a plugin that is
2281       actually useful (such as the one supplied by default) is even greater,
2282       but in the case of the default plugin it is still only an constant
2283       factor (amortized).
2284    
2285    3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
2286       with a sub-structure BTrace for controlling back-tracing.  In particular,
2287       the above-mentioned function "mode" controls whether the annotation
2288       phase is invoked by the compiler.  Another important function is
2289       "trigger": when called it aborts the current execution and causes
2290       the top-level loop to print a full back-trace.
2291    
2292    4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
2293       for back-tracing.  It keeps track of the dynamic call stack and in
2294       addition to that it keeps a partial history at each "level" of that
2295       stack.  For example, if a tail-calls b, b tail-calls c, and c tail-calls
2296       d and b (at separate times, dynamically), then the report will show:
2297    
2298       GOTO   d
2299             /c
2300       GOTO  \b
2301       CALL   a
2302    
2303       This shows that there was an initial non-tail call of a, then a
2304       tail-call to b or c, looping behavior in a cluster of functions that
2305       consist of b and c, and then a goto from that cluster (i.e., either from
2306       b or from c) to d.
2307    
2308       Note that (depending on the user program) the amount of information
2309       that the back-trace module has to keep track of at each level is bounded
2310       by a constant.  Thus, the whole implementation has the same asymptotical
2311       complexity as the original program (both in space and in time).
2312    
2313    5. compiler/TopLevel/interact/evalloop.sml has been modified to
2314       handle the special exception SMLofNJ.Internals.BTrace.BTrace
2315       which is raised by the "trigger" function mentioned above.
2316    
2317    Notes on usage:
2318    
2319    - Annotated code works well together with unannotated code:
2320    Unannotated calls simply do not show up at all in the backtrace.
2321    
2322    - It is not a good idea to let modules that were annotated during
2323    different sessions run at the same time.  This is because the compiler
2324    chooses small integers to identify individual functions, and there
2325    will be clashes if different modules were compiled in separate sessions.
2326    (Nothing will crash, and you will even be told about the clashes, but
2327    back-trace information will in general not be useful.)
2328    
2329    - Back-tracing can be confused by callcc and capture.
2330    
2331    - The only way of getting a back-trace right now is to explicitly
2332    invoke the "trigger" function from your user program.  Eventually, we
2333    should make every exception carry back-trace information (if
2334    available).  But since this creates more overhead at "raise"-time
2335    (similar to the current exnHistory overhead), I have not yet
2336    implemented this.  (The implementation will be rather easy.)  With
2337    exceptions carrying back-trace information, this facility will be even
2338    more useful because users don't need to modify their programs...
2339    
2340    - While it is possible to compile the compiler with back-trace
2341    annotations turned on (I did it to get some confidence in
2342    correctness), you must make absolutely sure that core.sml and
2343    btimp.sml are compiled WITHOUT annotation!  (core.sml cannot actually
2344    be compiled with annotation because there is no core access yet, but
2345    if you compile btimp.sml with annotation, then the system will go into
2346    an infinite recursion and crash.)
2347    Since CM currently does not know about BTrace, the only way to turn
2348    annotations on and off for different modules of the compiler is to
2349    interrupt CMB.make, change the settings, and re-invoke it.  Of course,
2350    this is awkward and clumsy.
2351    
2352    Sample sessions:
2353    
2354    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
2355    - SMLofNJ.Internals.BTrace.mode (SOME true);
2356    [autoloading]
2357    [autoloading done]
2358    val it = false : bool
2359    - structure X = struct
2360    -     fun main n = let
2361    -         fun a (x, 0) = d x
2362    -           | a (x, n) = b (x, n - 1)
2363    -         and b (x, n) = c (x, n)
2364    -         and c (x, n) = a (x, n)
2365    -         and d x = e (x, 3)
2366    -         and e (x, 0) = f x
2367    -           | e (x, n) = e (x, n - 1)
2368    -         and f 0 = SMLofNJ.Internals.BTrace.trigger ()
2369    -           | f n = n * g (n - 1)
2370    -         and g n = a (n, 3)
2371    -     in
2372    -         f n
2373    -     end
2374    - end;
2375    structure X : sig val main : int -> int end
2376    - X.main 3;
2377    *** BACK-TRACE ***
2378    GOTO   stdIn:4.2-13.20: X.main[2].f
2379    GOTO-( stdIn:4.2-13.20: X.main[2].e
2380    GOTO   stdIn:4.2-13.20: X.main[2].d
2381         / stdIn:4.2-13.20: X.main[2].a
2382         | stdIn:4.2-13.20: X.main[2].b
2383    GOTO-\ stdIn:4.2-13.20: X.main[2].c
2384    CALL   stdIn:4.2-13.20: X.main[2].g
2385    GOTO   stdIn:4.2-13.20: X.main[2].f
2386    GOTO-( stdIn:4.2-13.20: X.main[2].e
2387    GOTO   stdIn:4.2-13.20: X.main[2].d
2388         / stdIn:4.2-13.20: X.main[2].a
2389         | stdIn:4.2-13.20: X.main[2].b
2390    GOTO-\ stdIn:4.2-13.20: X.main[2].c
2391    CALL   stdIn:4.2-13.20: X.main[2].g
2392    GOTO   stdIn:4.2-13.20: X.main[2].f
2393    GOTO-( stdIn:4.2-13.20: X.main[2].e
2394    GOTO   stdIn:4.2-13.20: X.main[2].d
2395         / stdIn:4.2-13.20: X.main[2].a
2396         | stdIn:4.2-13.20: X.main[2].b
2397    GOTO-\ stdIn:4.2-13.20: X.main[2].c
2398    CALL   stdIn:4.2-13.20: X.main[2].g
2399    GOTO   stdIn:4.2-13.20: X.main[2].f
2400    CALL   stdIn:2.15-17.4: X.main[2]
2401    -
2402    
2403    (Note that because of a FLINt bug the above code currently does not
2404    compile without BTrace turned on.)
2405    
2406    Here is another example, using my modified Tiger compiler:
2407    
2408    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
2409    - SMLofNJ.Internals.BTrace.mode (SOME true);
2410    [autoloading]
2411    [autoloading done]
2412    val it = false : bool
2413    - CM.make "sources.cm";
2414    [autoloading]
2415    ...
2416    [autoloading done]
2417    [scanning sources.cm]
2418    [parsing (sources.cm):parse.sml]
2419    [creating directory CM/SKEL ...]
2420    [parsing (sources.cm):tiger.lex.sml]
2421    ...
2422    [wrote CM/sparc-unix/semant.sml]
2423    [compiling (sources.cm):main.sml]
2424    [wrote CM/sparc-unix/main.sml]
2425    [New bindings added.]
2426    val it = true : bool
2427    - Main.compile ("../testcases/merge.tig", "foo.out");
2428    *** BACK-TRACE ***
2429    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
2430    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
2431    CALL   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
2432    GOTO   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
2433    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
2434    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
2435    CALL   lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
2436         / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
2437    CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
2438    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
2439    CALL   lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
2440    CALL   main.sml:1.18-118.4: Main.compile[2]
2441    -
2442    
2443    ----------------------------------------------------------------------
2444    Name: Matthias Blumen
2445    Date: 2000/06/21 18:00:00 JST
2446    Tag: blume-20000621-manual
2447    Description:
2448    
2449    CM manual update: Path environments documented.
2450    
2451    ----------------------------------------------------------------------
2452    Name: Matthias Blume
2453    Date: 2000/06/19 13:40:00
2454    Tag: blume-20000619-manual
2455    Description:
2456    
2457    CM manual and system/README update.  This only covers the fact that
2458    there are no more implicit anchors.  (Path environments and the "bind"
2459    option to "cm" have yet to be documented.)
2460    
2461    ----------------------------------------------------------------------
2462    Name: Matthias Blume
2463    Date: 2000/06/19 11:05:00 JST
2464    Tag: blume-20000619-chdir-bugfix
2465    Description:
2466    
2467    Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
2468    
2469    ----------------------------------------------------------------------
2470    Name: Matthias Blume
2471    Date: 2000/06/18 22:00:10 JST
2472    Tag: blume-20000618-implicit-anchors-really-gone
2473    Description:
2474    
2475    I updates the previous HISTORY entry where I forgot to mention that
2476    implicit anchors are no longer with us.
2477    
2478    The current update also gets rid of the (now useless) controller
2479    CM.Control.implicit_anchors.
2480    
2481    ----------------------------------------------------------------------
2482    Name: Matthias Blume
2483    Date: 2000/06/16 17:30:00 JST
2484    Tag: blume-20000616-anchorenv
2485    Description:
2486    
2487    This patch implements the long anticipated (just kidding :) "anchor
2488    environment" mechanism.  In the course of doing this, I also
2489    re-implemented CM's internal "SrcPath" module from scratch.  The new
2490    one should be more robust in certain boundary cases.  In any case, it
2491    is a lot cleaner than its predecessor (IMHO).
2492    
2493    This time, although there is yet another boot file format change, I
2494    kept the unpickler backward-compatible.  As a result, no new bootfiles
2495    are necessary and bootstrapping is straightforward.  (You cannot read
2496    new bootfiles into an old system, but the other way around is no
2497    problem.)
2498    
2499    Visible changes:
2500    
2501    ** 0. Implicit path anchors (without the leading $-symbol) are no
2502    longer recognized at all. This means that such path names are not
2503    illegal either.  For example, the name basis.cm simply refers to a
2504    local file called "basis.cm" (i.e, the name is an ordinary path
2505    relative to .cm-files directory).  Or, to put it differently, only
2506    names that start with $ are anchored paths.
2507    
2508    ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
2509    vanished.
2510    
2511    John (Reppy) had critizised this as soon as I originally proposed and
2512    implemented it, but at that time I did not really deeply believe
2513    him. :) Now I came full-circle because I need the $<singlearc> syntax
2514    in another place where it cannot be seen as an abbreviation for
2515    $/<singlearc>.  To avoid the confusion, $<singlearc> now means what it
2516    seems to mean (i.e., it "expands" into the corresponding anchor
2517    value).
2518    
2519    However, when paths are used as members in CM description files, it
2520    continues to be true that there must be at least another arc after the
2521    anchor.  This is now enforced separately during semantic analysis
2522    (i.e., from a lexical/syntactical point of view, the notation is ok.)
2523    
2524    ** 2. The "cm" class now accepts an option "bind".  The option's value
2525    is a sub-option list of precisely two items -- one labeled "anchor"
2526    and the other one labeled "value".  As you might expect, "anchor" is
2527    used to specify an anchor name to be bound, and "value" specifies what
2528    the anchor is being bound to.
2529    
2530    The value must be a directory name and can be given in either standard
2531    syntax (including the possibility that it is itself an anchored path)
2532    or native syntax.
2533    
2534    Examples:
2535    
2536       foo.cm (bind:(anchor:bar value:$mystuff/bar))
2537       lib.cm (bind:(anchor:a value:"H:\\x\\y\\z"))  (* only works under windows *)
2538    
2539    and so on.
2540    
2541    The meaning of this is that the .cm-file will be processed with an
2542    augmented anchor environment where the given anchor(s) is/are bound to
2543    the given values(s).
2544    
2545    The rationale for having this feature is this: Suppose you are trying
2546    to use two different (already stable) libraries a.cm and b.cm (that
2547    you perhaps didn't write yourself).  Further, suppose each of these
2548    two libraries internally uses its own auxiliary library $aux/lib.cm.
2549    Normally you would now have a problem because the anchor "lib" can not
2550    be bound to more than one value globally.  Therefore, the project that
2551    uses both a.cm and b.cm must locally redirect the anchor to some other
2552    place:
2553    
2554       a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
2555       b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
2556    
2557    This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
2558    /usr/lib/smlnj/b-stuff/aux.cm, respectively.
2559    
2560    Hard-wiring path names is a bit inflexible (and CM will verbosely warn
2561    you when you do so at the time of CM.stabilize).  Therefore, you can
2562    also use an anchored path as the value:
2563    
2564      a.cm (bind:(anchor:lib value:$a-lib))
2565      b.cm (bind:(anchor:lib value:$b-lib))
2566    
2567    Now you can globally configure (using the usual CM.Anchor.anchor or
2568    pathconfig machinery) bindings for "a-lib" and "b-lib".  Since "lib"
2569    itself is always locally bound, setting it globally is no longer
2570    meaningful or necessary (but it does not hurt either).  In fact, "lib"
2571    can still be used as a global anchor for separate purposes.  As a
2572    matter of fact, one can locally define "lib" in terms of a global
2573    "lib":
2574    
2575      a.cm (bind:(anchor:lib value:$lib/a))
2576      b.cm (bind:(anchor:lib value:$lib/b))
2577    
2578    ** 3: The encoding of path names has changed.  This affects the way
2579    path names are shown in CM's progress report and also the internal
2580    protocol encoding used for parallel make.
2581    
2582    The encoding now uses one or more ':'-separated segments.  Each
2583    segments corresponds to a file that has been specified relative to the
2584    file given by its preceding segment.  The first segment is either
2585    relative to the CWD, absolute, or anchored.  Each segment itself is
2586    basically a Unix pathname; all segments but the first are relative.
2587    
2588    Example:
2589    
2590       $foo/bar/baz.cm:a/b/c.sml
2591    
2592    This path denotes the file bar/a/b/c.sml relative to the directory
2593    denoted by anchor "foo".  Notice that the encoding also includes
2594    baz.cm which is the .cm-file that listed a/b/c.sml.  As usual, such
2595    paths are resolved relative to the .cm-files directory, so baz.cm must
2596    be ignored to get the "real" pathname.
2597    
2598    To make this fact more obvious, CM puts the names of such "virtual
2599    arcs" into parentheses when they appear in progress reports. (No
2600    parentheses will appear in the internal protocol encoding.)  Thus,
2601    what you really see is:
2602    
2603      $foo/bar/(baz.cm):a/b/c.sml
2604    
2605    I find this notation to be much more informative than before.
2606    
2607    Another new feature of the encoding is that special characters
2608    including parentheses, colons, (back)slashes, and white space are
2609    written as \ddd (where ddd is the decimal encoding of the character).
2610    
2611    *** The CM manual still needs to be updated.
2612    
2613    ----------------------------------------------------------------------
2614    Name: Allen Leung
2615    Date: 2000/06/15 00:38:00
2616    Tag: leunga-20000615-x86-peephole
2617    
2618    x86 Peephole fix by Fermin.  Affects c-- and moby only.
2619    
2620    ----------------------------------------------------------------------
2621    Name: Matthias Blume
2622    Date: 2000/06/12 11:40:00
2623    Tag: blume-20000612-parmakefix
2624    Description:
2625    
2626    More cleanup after changing the file naming scheme: This time I
2627    repaired the parallel make mechanism for CMB.make which I broke earlier.
2628    
2629    ----------------------------------------------------------------------
2630    Name: Allen Leung
2631    Date: 2000/06/09 01:25:00
2632    Tag: leunga-20000609-various
2633    
2634    None of these things should affect normal SML/NJ operations
2635    
2636    1. Peephole improvements provided by Fermin (c--)
2637    2. New annotation DEFUSE for adding extra dependence (moby)
2638    3. New X86 LOCK instructions (moby)
2639    4. New machine description language for reservation tables (scheduling)
2640    5. Fixes to various optimization/analysis modules (branch chaining, dominator
2641       trees etc.)
2642    6. I've changed the CM files so that they can work with versions
2643       110.0.6, 110.25 and 110.28
2644    
2645    ----------------------------------------------------------------------
2646    Name: Matthias Blume
2647    Date: 2000/06/09 12:40:00
2648    Tag: blume-20000609-log
2649    Description:
2650    
2651    - Removed all(?) remaining RCS Log entries from sources.
2652    
2653    - Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
2654      anchored paths).
2655    
2656    ----------------------------------------------------------------------
2657    Name: Matthias Blume
2658    Date: 2000/06/07 17:00:00 JST
2659    Tag: blume-20000607-no-implicit-anchors
2660    Description:
2661    
2662    1. This update changes the default setting for
2663    CM.Control.implicit_anchors from true to false.  This means that
2664    implicit anchors are no longer permitted by default.  I also tried to
2665    make sure that nothing else still relies on implicit anchors.
2666    (This is the next step on the schedule towards a CM that does not even
2667    have the notion of implicit anchors anymore.)
2668    
2669    2. More CM manual updates.
2670    
2671    3. I managed to track down and fix the pickling bug I mentioned last
2672    time.  Because of the previously existing workaround, this entails no
2673    immediate practical changes.
2674    
2675    ----------------------------------------------------------------------
2676    Name: Matthias Blume
2677    Date: 2000/06/06 11:15:00 JST
2678    Tag: blume-20000606-lazierpickle
2679    Description:
2680    
2681    !!!! NEW BOOT FILES !!!!
2682    
2683    * The main purpose of this update is to make library pickles lazier in
2684    order to reduce the initial space penalty for autoloading a library.
2685    As a result, it is now possible to have $smlnj/compiler.cm
2686    pre-registered.  This should take care of the many complaints or
2687    inquiries about missing structure Compiler.  This required changes to
2688    CM's internal data structures and small tweaks to some algorithms.
2689    
2690    As a neat additional effect, it is no longer necessary (for the sake
2691    of lean heap image files) to distinguish between a "minimal" CM and a
2692    "full" CM.  Now, there is only one CM (i.e., the "full" version:
2693    $smlnj/cm.cm aka $smlnj/cm/full.cm), and it is always available at the
2694    interactive top level. ($smlnj/cm/minimal.cm is gone.)
2695    
2696    To make the life of compiler-hackers easier, "makeml" now also
2697    pre-registers $smlnj/cmb.cm (aka $smlnj/cmb/current.cm).  In other
2698    words, after you bootstrap a new sml for the first time, you will not
2699    have to autoload $smlnj/cmb.cm again afterwards.  (The first time
2700    around you will still have to do it, though.)
2701    
2702    * A second change consists of major updates to the CM manual.  There
2703    are now several appendices with summary information and also a full
2704    specification of the CM description file syntax.
2705    
2706    * In directory src/system I added the script "allcross".  This script
2707    invokes sml and cross-compiles the compiler for all supported
2708    architectures.  (Useful when providing a new set of boot files.)
2709    
2710    * There seems to be a latent bug in my "lazy pickles" mechanism.  I
2711    added a small tweak to pickle-util.sml to work around this problem,
2712    but it is not a proper fix yet.  I will investigate further.  (The
2713    effect of the bug was an inflation of library pickle size.)
2714    
2715    * Version number increased to 110.28.1 (to avoid compatibility problems).
2716    
2717    ----------------------------------------------------------------------
2718    Name: Allen Leung
2719    Date: 2000/05/25 17:28 EDT
2720    Tag: leunga-20000525-ra
2721    Description:
2722    
2723      Fixed a bug in freezing phase of the register allocator.
2724    
2725    ----------------------------------------------------------------------
2726    Name: Allen Leung
2727    Date: 2000/05/15 22:53 EDT
2728    Tag: leunga-20000515-alpha-x86-ra
2729    Description:
2730    
2731      1. Alpha
2732    
2733          Slight cleanup.  Removed the instruction SGNXL
2734    
2735      2. X86
2736    
2737          Added the following instructions to the instruction set:
2738    
2739            ROLx, RORx,
2740            BTx, BTSx, BTLx, BTRx,
2741            XCHGx, and variants with the LOCK prefix
2742    
2743      3. Register Allocation
2744    
2745          The module ra-rewrite-with-renaming has been improved.
2746    
2747      These have no effect on SML/NJ.
2748    
2749    ----------------------------------------------------------------------
2750    Name: Matthias Blume
2751    Date: 2000/05/15 16:20:00 JST
2752    Tag: blume-20000515-lightrebuild
2753    Description:
2754    
2755    1. I added an alternative to "-rebuild" to "makeml".  The difference is
2756       that prior to calling CMB.make' the CM-variable "LIGHT" will be
2757       defined.  In effect, the command will not build any cross-compiler
2758       backends and therefore finish more quickly.
2759    
2760       The "fixpt" script also takes a "-light" switch to be able to use
2761       this new facility while compiling for a fixpoint.
2762    
2763    2. I replaced all mentions of anchored paths in group owner specifications
2764       with simple relative paths (usually starting with "..").
2765       The rationale is that a library's internal workings should not be
2766       compromised by the lack of some anchor.  (An anchor is necessary
2767       for someone who wants to refer to the library by an anchored path,
2768       but it should not be necessary to build the same library in the first
2769       place.)
2770    
2771    3. I changed the way CM's tool mechanism determines the shell command
2772       string used for things like ml-yacc etc. so that it does not break
2773       when CM.Control.implicit_anchors is turned off.
2774    
2775    ----------------------------------------------------------------------
2776    Name: Matthias Blume
2777    Date: 2000/05/12 18:20:00 JST
2778    Tag: blume-20000512-ml-build
2779    Description:
2780    
2781    Fixed a bug in config/_ml-build that prevented ml-yacc and ml-lex from
2782    getting installed properly (by config/install.sh).
2783    
2784    ----------------------------------------------------------------------
2785    Name: Matthias Blume
2786    Date: 2000/05/12 17:30:00 JST
2787    Tag: blume-20000512-anchors
2788    Description:
2789    
2790    !!! NEW BOOT FILES !!!
2791    
2792    This change is in preparation of fading out support for "implicitly
2793    anchored path names".  I went through all sources and used the
2794    explicit (and relatively new) $-notation.  See system/README and the
2795    CM manual for more info on this.
2796    
2797    I also modified the anchoring scheme for some things such as "smlnj",
2798    "MLRISC", "cm", etc. to take advantage of the fact that explicit
2799    anchors are more expressive: anchor name and first arc do not have to
2800    coincide.  This entails the following user-visible change:
2801    
2802    You have to write $smlnj/foo/bar instead of smlnj/foo/bar.  In
2803    particular, when you fire up sml with a command-line argument, say,
2804    e.g.:
2805    
2806       sml '$smlnj/cmb.cm'
2807    
2808    At the ML toplevel prompt:
2809    
2810       CM.autoload "$smlnj/cmb.cm";
2811    
2812    There is also a new controller in CM.Control that can be used to turn
2813    off all remaining support for implicit anchors by saying:
2814    
2815        CM.autoload "$smlnj/
2816        #set CM.Control.implicit_anchors false;
2817    
2818    This causes CM to reject implicitly anchored paths.  This is (for the
2819    time being) less permissive than the "final" version where there will
2820    be no more such implicit anchors and relative paths will be just that:
2821    relative.
2822    
2823    The next step (version after next version?) will be to make the
2824    default for CM.Control.implicit_anchors false.  After the dust has
2825    settled, I can then produce the "final" version of this...
2826    
2827    Note: Since bootstrapping is a bit tricky, I provided new boot files.
2828    
2829    ----------------------------------------------------------------------
2830    Name: Matthias Blume
2831    Date: 2000/05/11 16:30:00 JST
2832    Tag: blume-20000511-sources
2833    Description:
2834    
2835    The main change is that I added function CM.sources as a generalized
2836    version of the earlier CM.makedepend.  This entails the following
2837    additional changes:
2838    
2839      - CM.makedepend has been dropped.
2840    
2841      - CM manual has been updated.
2842    
2843      - TOOLS signature and API have been changed.
2844    
2845    ----------------------------------------------------------------------
2846    Name: Allen Leung
2847    Date: 2000/05/10 21:17 EDT
2848    Tag: leunga-20000510-moby-c--ssa
2849    Description:
2850    
2851      Various bug fixes and new features for C--, Moby and MLRISC optimizations.
2852    None of these affect SML/NJ.
2853    
2854    1. Register Allocation
2855    
2856        a. A new ra spilling module (ra/ra-spill-with-renaming) is implemented.
2857           This module tries to remove local (i.e. basic block level) redundancies
2858           during spilling.
2859    
2860        b. A new framework for performing region based register allocation.
2861           Not yet entirely functional.
2862    
2863    2. X86
2864    
2865       a. DefUse for POP was missing the stack pointer [found by Lal]
2866       b. Reload for CALL was incorrect in X86Spill [found by John]
2867       c. Various fixes in X86Spill so that it can be used correctly for
2868          the new spilling module.
2869    
2870    3. SSA/IR
2871    
2872       a. New module ir/dj-dataflow.sml implements elimination based
2873          data flow analysis.
2874    
2875    4. MLRiscGen
2876    
2877       a. Fix for gc type annotation
2878    
2879    5. MDGen
2880    
2881       Various fixes for machine description -> ml code translation.  For ssa
2882       only.
2883    
2884    ----------------------------------------------------------------------
2885    Name: Allen Leung
2886    Date: 2000/05/08 22:17 EDT
2887    Tag: leunga-20000508-labexp
2888    Description:
2889    
2890      Fermin has found a few assembly problems with constant expressions
2891      generated in LabelExp.  Mostly, the problems involve extra parentheses,
2892      which choke on dumb assemblers.  This is his fix.
2893    
2894    ----------------------------------------------------------------------
2895    Name: Dave MacQueen
2896    Date: 2000/04/09 14:00 EDT
2897    Tag: dbm-20000502-Version_110_28
2898    Description:
2899    
2900    1. Updated src/compiler/TopLevel/main/version.sml to version 110.28
2901    
2902    2. Updated config/version to 110.28
2903    
2904    3. Updated config/srcarchiveurl
2905    
2906    3. New boot files!
2907       ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.28/
2908    
2909    ----------------------------------------------------------------------
2910    Name: Matthias Blume
2911    Date: 2000/05/01 19:05:00 JST
2912    Tag: blume-20000501-noweb
2913    Description:
2914    
2915    A new noweb tool has been added.  The existing system is entirely
2916    unaffected by this, but some CM users have asked for renewed noweb
2917    support.  Everything is documented in the CM manual.
2918    
2919    New (plugin) libraries:
2920    
2921       noweb-tool.cm
2922       nw-ext.cm
2923    
2924    ----------------------------------------------------------------------
2925    Name: Dave MacQueen
2926    Date: 2000/04/30 12:40PM EDT
2927    Tag: dbm-20000430-bug_fixes
2928    Description:
2929    
2930    1. Fix for bug 1498
2931       smlnj/src/system/Basis/Implementation/Unsafe/object.sig
2932       smlnj/src/system/Basis/Implementation/Unsafe/object.sml
2933         added toRealArray function
2934       smlnj/src/compiler/MiscUtil/print/ppobj.sml
2935         added check for tag Obj.RealArray to array printing case in ppObj
2936    
2937    2. Fix for bug 1510
2938       smlnj/src/compiler/Semant/types/typesutil.sml
2939         fixed definition of dummyargs (used by equalTycon) so that
2940         dummy args are distinct types
2941    
2942    ----------------------------------------------------------------------
2943    Name: Matthias Blume
2944    Date: 2000/04/30 01:00:00 JST
2945    Tag: blume-20000430-versions
2946    Description:
2947    
2948    1. CM version numbering added.  This is an implementation of Lal's
2949       proposal for adding version numbers and version checking to .cm
2950       files.  Lal said that his proposal was just that -- a proposal.
2951       For the time being I went ahead and implemented it so that people
2952       can comment on it.  Everything is completely backward-compatible
2953       (except for the stable library format, i.e., new bootfiles!).
2954    
2955       As usual, see the CM manual for details.
2956    
2957    2. An alternative syntax for anchored paths has been implemented.
2958       Dave has recently voiced the same concerns that I had when I did
2959       this, so there should be some support.  My take is that eventually
2960       I will let support for the current syntax (where anchors are
2961       "implicit") fade out in favor of the new, explicit syntax.
2962       In order to be backward-compatible, both old and new syntax are
2963       currently supported.
2964    
2965       Again, see the CM manual for details.
2966    
2967    3. Parallel make is trying to be slightly smarter:  When the master
2968       process finds a "bottleneck", i.e., when there is only one
2969       compilation unit that can be compiled and everybody else is
2970       waiting on it, then it will simply compile it directly instead
2971       of clumsily telling one of the slaves to do it.
2972    
2973    4. Support for "unsharing" added.  This is necessary in order to be
2974       able to have two different versions of the same library running
2975       at the same time (e.g., for trying out a new MLRISC while still
2976       having the old MLRISC linked into the current compiler, etc.)
2977       See the CM manual.
2978    
2979    5. Simple "makedepend" functionality added for generating Makefile
2980       dependency information.  (This is rather crude at the moment.
2981       Expect some changes here in the future.)
2982    
2983    6. ".fun" added as a recognized suffix for ML files. Also documented
2984       explicitly in the manual that the fallback behavior (unknown suffix
2985       -> ML file) is not an official feature!
2986    
2987    7. Small changes to the pickler for stable libraries.
2988    
2989    8. Several internal changes to CM (for cleanup/improvement).
2990    
2991    
2992    !!!! NEW BINFILES !!!!
2993    
2994    ----------------------------------------------------------------------
2995    Name: Matthias Blume
2996    Date: 2000/04/28 17:30:00 JST
2997    Tag: blume-20000428-pathconfig
2998    Description:
2999    
3000    1. I changed config/install.sh to remove duplicate entries from the
3001       lib/pathconfig file at the end.  Moreover, the final version of
3002       lib/pathconfig is sorted alphabetically.  The same (sorting) is done
3003       in src/system/installml.
3004    
3005    2. The config/install.sh script now consistently uses relative
3006       pathnames in lib/pathconfig whenever the anchor is in the lib
3007       directory.  (So far this was true for the libraries that come
3008       pre-compiled and bundled as part of the bootfiles but not for
3009       libraries that are compiled by the script itself.)
3010    
3011    ----------------------------------------------------------------------
3012    Name: Matthias Blume
3013    Date: 2000/04/26 13:10:00 JST
3014    Tag: blume-20000426-fun_suffix
3015    Description:
3016    
3017    Added ".fun" as a recognized file name suffix (for ML code).
3018    
3019    ----------------------------------------------------------------------
3020    Name: Allen Leung
3021    Date: 2000/04/25 17:00:00 EST
3022    Tag: leunga-20000425-alpha-ra
3023    Description:
3024    
3025    1. Alpha
3026    
3027        PSEUDOARITH was missing in AlphaRewrite.  This causes an endless loop
3028    in C--.
3029    
3030    2. RA
3031    
3032       Added a flag "ra-dump-size" to print out the size of the flowgraph
3033       and the interference graph.
3034    
3035    ----------------------------------------------------------------------
3036    Name: Dave MacQueen
3037    Date: 2000/04/25/
3038    Tag: dbm-20000425-mlyacc_doc_examples
3039    Description:
3040      Updated mlyacc.tex sections 5 and 7 for SML '97 and CM.
3041      Updated all three examples in src/ml-yacc/examples to run
3042      under 110.* using CM.make.
3043    
3044    ----------------------------------------------------------------------
3045    Name: Allen Leung
3046    Date: 2000/04/20 23:04:00 EST
3047    Tag: leunga-20000420-ssa-c---stuff
3048    Description:
3049    
3050      This update synchronizes my repository with Yale's.  Most of these
3051    changes, however, do not affect SML/NJ at all (the RA is an exception).
3052    
3053    1. Register Allocator
3054    
3055       a. An improvement in the interference graph construction:
3056          Given a copy
3057    
3058                s <- t
3059    
3060          no interference edge between s and t is added for this definition of s.
3061    
3062       b. I've added two new spill heuristic modules that Fermin and I developed
3063          (in the new library RA.cm). These are unused in SML/NJ but maybe
3064          useful for others (Moby?)
3065    
3066    2. X86
3067    
3068       a. Various fixes in the backend provided by Fermin [C--] and Lal.
3069    
3070    3. Alpha
3071    
3072       a. Added the BSR instruction and code generation that goes with it [C--]
3073       b. Other fixes too numerous to recount provided by Fermin [C--]
3074    
3075    4. Regmaps
3076    
3077       a. The regmaps are not initialized with the identity physical bindings
3078          at creation time.  This is unneeded.
3079    
3080    5. MLRISC Optimizations
3081    
3082       a. The DJ-Graph module can now compute the iterated dominance frontiers
3083          intersects with liveness incrementally in linear time! Woohoo!
3084          This is now used in my new SSA construction algorithm.
3085    
3086       b. THe branch reorganization module is now smarter about linear chains of
3087          basic blocks.
3088    
3089    
3090    ----------------------------------------------------------------------
3091    Name: Matthias Blume
3092    Date: 2000/04/12 13:52:00 JST
3093    Tag: blume_main_v110p27_1
3094    Description:
3095    
3096    Changed install.sh script to handle archive files without version number
3097    and to use "boot.<arch>-<os>" instead of "sml.boot.<arch>-<os>" for the
3098    name of the boot file archive.
3099    
3100    ----------------------------------------------------------------------
3101    Name: Dave MacQueen
3102    Date: 2000/04/09 14:00 EDT
3103    Tag: dbm-20000410-Version_110_27
3104    Description:
3105    
3106    1. Updated src/compiler/TopLevel/main/version.sml to version 110.27
3107    
3108    2. Updated src/config/version to 110.27
3109    
3110    3. New boot files!
3111    
3112    ----------------------------------------------------------------------
3113    Name: Allen Leung
3114    Date: 2000/04/09 19:09:00 EST
3115    Tag: leunga-20000409-misc
3116    Description:
3117    
3118    1.  Yet another fix for x86 assembly for idivl, imull, mull and friends.
3119    
3120    2.  Miscellaneous improvements to MLRISC (unused in sml/nj)
3121    
3122    ----------------------------------------------------------------------
3123    Name: Stefan
3124    Date: 2000/04/07 10:00:00 EDT
3125    Tag: monnier-20000406-branch-handling
3126    Description:
3127    
3128    Improved handling of branches (mostly those generated from
3129    polymorphic equality), removed switchoff and changed the
3130    default optimization settings (more cpsopt and less flintopt).
3131    
3132    ----------------------------------------------------------------------
3133    Name: Allen Leung
3134    Date: 2000/04/06 01:30:00 EST
3135    Tag: leunga-20000406-peephole-x86-SSA-2
3136    Description:
3137    
3138       Forgot a few files.
3139    
3140    ----------------------------------------------------------------------
3141    Name: Allen Leung
3142    Date: 2000/04/06 00:36:00 EST
3143    Tag: leunga-20000406-peephole-x86-SSA
3144    Description:
3145    
3146    1.  New Peephole code
3147    
3148    2.  Minor improvement to X86 instruction selection
3149    
3150    3.  Various fixes to SSA and machine description -> code translator
3151    
3152    ----------------------------------------------------------------------
3153    Name: Matthias Blume
3154    Date: 2000/04/05 12:30:00 JST
3155    Tag: blume_main_v110p26p2_3
3156    Description:
3157    
3158    This update just merges three minor cosmetic updates to CM's sources
3159    to get ready for the 110.27 code freeze on Friday.  No functionality
3160    has changed.
3161    
3162    ----------------------------------------------------------------------
3163    Name: Allen Leung
3164    Date: 2000/04/04 19:39:00 EST
3165    Tag: leunga-20000404-x86-asm
3166    Description:
3167    
3168    1.  Fixed a problem in X86 assembly.
3169    
3170        Things like
3171    
3172           jmp %eax
3173           jmp (%eax)
3174    
3175        should be output as
3176    
3177           jmp *%eax
3178           jmp *(%eax)
3179    
3180    2.  Assembly output
3181    
3182          Added a new flag
3183    
3184              "asm-indent-copies" (default to false)
3185    
3186          When this flag is on, parallel copies will be indented an extra level.
3187    
3188    ----------------------------------------------------------------------
3189    Name: Allen Leung
3190    Date: 2000/04/04 03:18:00 EST
3191    Tag: leunga-20000404-C--Moby
3192    Description:
3193    
3194        All of these fixes are related to C--, Moby, and my own optimization
3195        stuff; so they shouldn't affect SML/NJ.
3196    
3197    1.  X86
3198    
3199        Various fixes related floating point, and extensions.
3200    
3201    2.  Alpha
3202    
3203        Some extra patterns related to loads with signed/zero extension
3204        provided by Fermin.
3205    
3206    3.  Assembly
3207    
3208        When generating assembly, resolve the value of client defined constants,
3209        instead of generating symbolic values.  This is controlled by the
3210        new flag "asm-resolve-constants", which is default to true.
3211    
3212    4.  Machine Descriptions
3213    
3214        a. The precedence parser was slightly broken when parsing infixr symbols.
3215        b. The type generalizing code had the bound variables reversed, resulting
3216           in a problem during arity raising.
3217        c. Various fixes in machine descriptions.
3218    
3219    ----------------------------------------------------------------------
3220    Name: Matthias Blume
3221    Date: 2000/04/03 16:05:00 JST
3222    Tag: blume_main_v110p26p2_2
3223    Description:
3224    
3225    I eliminated coreEnv from compInfo.  Access to the "Core" structure is
3226    now done via the ordinary static environment that is context to each
3227    compilation unit.
3228    
3229    To this end, I arranged that instead of "structure Core" as "structure
3230    _Core" is bound in the pervasive environment.  Core access is done via
3231    _Core (which can never be accidentally rebound because _Core is not a
3232    legal surface-syntax symbol).
3233    
3234    The current solution is much cleaner because the core environment is
3235    now simply part of the pervasive environment which is part of every
3236    compilation unit's context anyway.  In particular, this eliminates all
3237    special-case handling that was necessary until now in order to deal
3238    with dynamic and symbolic parts of the core environment.
3239    
3240    Remaining hackery (to bind the "magic" symbol _Core) is localized in the
3241    compilation manager's bootstrap compiler (actually: in the "init group"
3242    handling).  See the comments in src/system/smlnj/init/init.cmi for
3243    more details.
3244    
3245    I also tried to track down all mentions of "Core" (as string argument
3246    to Symbol.strSymbol) in the compiler and replaced them with a
3247    reference to the new CoreSym.coreSym.  Seems cleaner since the actual
3248    name appears in one place only.
3249    
3250    Binfile and bootfile format have not changed, but the switchover from
3251    the old "init.cmi" to the new one is a bit tricky, so I supplied new
3252    bootfiles anyway.
3253    
3254    ----------------------------------------------------------------------
3255    Name: Allen Leung
3256    Date: 2000/04/02 21:17:00 EST
3257    Tag: leunga-20000402-mltree
3258    Description:
3259    
3260       1. Renamed the constructor CALL in MLTREE by popular demand.
3261       2. Added a bunch of files from my repository.  These are currently
3262          used by other non-SMLNJ backends.
3263    
3264    ----------------------------------------------------------------------
3265    Name: Allen Leung
3266    Date: 2000/03/31 21:15:00 EST
3267    Tag: leunga-20000331-aliasing
3268    Description:
3269    
3270    This update contains a rewritten (and hopefully more correct) module
3271    for extracting aliasing information from CPS.
3272    
3273       To turn on this feature:
3274    
3275            Compiler.Control.CG.memDisambiguate := true
3276    
3277       To pretty print the region information with assembly
3278    
3279           Compiler.Control.MLRISC.getFlag "asm-show-region" := true;
3280    
3281       To control how many levels of aliasing information are printed, use:
3282    
3283           Compiler.Control.MLRISC.getInt "points-to-show-level" := n
3284    
3285       The default of n is 3.
3286    
3287    ----------------------------------------------------------------------
3288    Name: David MacQueen
3289    Date: 2000/03/31 11:15:00 EST
3290    Tag: dbm-20000331-runtime_fix
3291    Description:
3292    
3293    This update contains:
3294    
3295    1. runtime/c-lib/c-libraries.c
3296       includes added in revision 1.2 caused compilation errors on hppa-hpux
3297    
3298    2. fix for bug 1556
3299       system/Basis/Implementation/NJ/internal-signals.sml
3300    
3301    ----------------------------------------------------------------------
3302    Name: Matthias Blume
3303    Date: 2000/03/31 18:00:00 JST
3304    Tag: blume_main_v110p26p2_1
3305    Description:
3306    
3307    This update contains:
3308    
3309    1. A small change to CM's handling of stable libraries:
3310       CM now maintains one "global" modmap that is used for all stable
3311       libraries.  The use of such a global modmap maximizes sharing and
3312       minimizes the need for re-traversing parts of environments during
3313       modmap construction.  (However, this has minor impact since modmap
3314       construction seems to account for just one percent or less of total
3315       compile time.)
3316    
3317    2. I added a "genmap" phase to the statistics.  This is where I got the
3318       "one percent" number (see above).
3319    
3320    3. CM's new tool parameter mechanism just became _even_ better. :)
3321       - The parser understands named parameters and recursive options.
3322       - The "make" and "shell" tools use these new features.
3323         (This makes it a lot easier to cascade these tools.)
3324       - There is a small syntax change: named parameters use a
3325    
3326           <name> : ( <option> ... )            or
3327           <name> : <string>
3328    
3329         syntax.  Previously, named parameters were implemented in an
3330         ad-hoc fashion by each tool individually (by parsing strings)
3331         and had the form
3332    
3333           <name>=<string>
3334    
3335       See the CM manual for a full description of these issues.
3336    
3337    ----------------------------------------------------------------------
3338    Name: Matthias Blume
3339    Date: 2000/03/30 18:00:00 JST
3340    Tag: blume_main_v110p26p2_0
3341    Description:
3342    
3343    !!!!! WARNING !!!!!!
3344    !!  New binfiles  !!
3345    !!!!!!!!!!!!!!!!!!!!
3346    
3347    This update contains:
3348    
3349    1. Moderate changes to CM:
3350    
3351       - Changes to CM's tools mechanism.  In particular, it is now possible
3352       to have tools that accept additional "command line" parameters
3353       (specified in the .cm file at each instance where the tool's class is
3354       used).
3355    
3356       This was done to accommodate the new "make" and "shell" tools which
3357       facilitate fairly seamless hookup to portions of code managed using
3358       Makefiles or Shell scripts.
3359    
3360       There are no classes "shared" or "private" anymore.  Instead, the
3361       sharing annotation is now a parameter to the "sml" class.
3362    
3363       There is a bit of generic machinery for implementing one's own
3364       tools that accept command-line parameters.  However, I am not yet fully
3365       satisfied with that part, so expect changes here in the future.
3366    
3367       All existing tools are described in the CM manual.
3368    
3369       - Slightly better error handling.  (CM now suppresses many followup
3370       error messages that tended to be more annoying than helpful.)
3371    
3372    2. Major changes to the compiler's static environment data structures.
3373    
3374       - no CMStaticEnv anymore.
3375            - no CMEnv, no "BareEnvironment" (actually, _only_ BareEnvironment,
3376              but it is called Environment), no conversions between different
3377              kinds of static environments
3378    
3379       - There is still a notion of a "modmap", but such modmaps are generated
3380         on demand at the time when they are needed.  This sounds slow, but I
3381         sped up the code that generates modmaps enough for this not to lead to
3382         a slowdown of the compiler (at least I didn't detect any).
3383    
3384       - To facilitate rapid modmap generation, static environments now
3385         contain an (optional) "modtree" structure.  Modtree annotations are
3386         constructed by the unpickler during unpickling.  (This means that
3387         the elaborator does not have to worry about modtrees at all.)
3388         Modtrees have the advantage that they are compositional in the same
3389         way as the environment data structure itself is compositional.
3390         As a result, modtrees never hang on to parts of an environment that
3391         has already been rendered "stale" by filtering or rebinding.
3392    
3393       - I went through many, many trials and errors before arriving at the
3394         current solution.  (The initial idea of "linkpaths" did not work.)
3395         But the result of all this is that I have touched a lot of files that
3396         depend on the "modules" and "types" data structures (most of the
3397         elaborator). There were a lot of changes during my "linkpath" trials
3398         that could have been reverted to their original state but weren't.
3399         Please, don't be too harsh on me for messing with this code a bit more
3400         than what was strictly necessary...  (I _did_ resist the temptation
3401         of doing any "global reformatting" to avoid an untimely death at
3402         Dave's hands. :)
3403    
3404       - One positive aspect of the previous point:  At least I made sure that
3405         all files that I touched now compile without warnings (other than
3406         "polyEqual").
3407    
3408       - compiler now tends to run "leaner" (i.e., ties up less memory in
3409         redundant modmaps)
3410    
3411  ----------------------------------------------------------------------  ----------------------------------------------------------------------
3412  Name: Allen Leung  Name: Allen Leung
# Line 104  Line 3500 
3500    
3501       I've changed andl to testl in the floating point test sequence       I've changed andl to testl in the floating point test sequence
3502       whenever appropriate.  The Intel optimization guide states that       whenever appropriate.  The Intel optimization guide states that
3503       testl is perferable to andl.       testl is preferable to andl.
3504    
3505  7. RA (x86 only)  7. RA (x86 only)
3506    
# Line 286  Line 3682 
3682    
3683  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
3684  additional argument called "template" which is an optional string that  additional argument called "template" which is an optional string that
3685  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
3686  explanation.  explanation.
3687    
3688  2. A special-purpose tool can be "regisitered" by simply dropping the  2. A special-purpose tool can be "registered" by simply dropping the
3689  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
3690  directory where the .cm file lives that uses this tool.  (The  directory where the .cm file lives that uses this tool.  (The
3691  behavior/misfeature until now was to look for the tool description  behavior/misfeature until now was to look for the tool description
# Line 333  Line 3729 
3729  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
3730  <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,
3731  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
3732  perhaps got accidentially killed).  In this case, fixpt will quickly  perhaps got accidentally killed).  In this case, fixpt will quickly
3733  move through what exists before continuing where it left off earlier,  move through what exists before continuing where it left off earlier,
3734  and, thus, saves a lot of time.  and, thus, saves a lot of time.
3735    
# Line 383  Line 3779 
3779    it from that remote directory.    it from that remote directory.
3780    This should simplify installation further:  For machines that have    This should simplify installation further:  For machines that have
3781    access to the internet, just fetch <version>-config.tgz, unpack it,    access to the internet, just fetch <version>-config.tgz, unpack it,
3782    edit config/targets, and go (run config/install.sh).  The scipt will    edit config/targets, and go (run config/install.sh).  The script will
3783    fetch everything else that it might need all by itself.    fetch everything else that it might need all by itself.
3784    
3785    For CVS users, this mechanism is not relevant for source archives, but    For CVS users, this mechanism is not relevant for source archives, but

Legend:
Removed from v.586  
changed lines
  Added in v.1021

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