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 680, Mon Jul 3 06:35:55 2000 UTC revision 918, Mon Aug 27 15:18:01 2001 UTC
# Line 11  Line 11 
11  Date:  Date:
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    
15    ----------------------------------------------------------------------
16    Name: Matthias Blume
17    Date: 2001/08/27 11:20:00 EDT
18    Tag: blume-20010827-readme11035
19    Description:
20    
21    Fleshed out the README file for 110.35.
22    
23    ----------------------------------------------------------------------
24    Name: Matthias Blume
25    Date: 2001/08/24 17:10:00 EDT
26    Tag: Release_110_35
27    Description:
28    
29    New version number (110.35).  New bootfiles.
30    
31    ----------------------------------------------------------------------
32    Name: Lal George
33    Date: 2001/08/24 13:47:18 EDT 2001
34    Tag: george-20010824-MLRISC-graphs
35    Description:
36    
37     removed clusters from MLRISC completely and replaced with graphs.
38    
39    ----------------------------------------------------------------------
40    Name: Matthias Blume
41    Date: 2001/08/23 17:50:00 EDT
42    Tag: blume-20010823-toplevel
43    Description:
44    
45    - some reorganization of the code that implements various kinds of
46      environments in the compiler (static, dynamic, symbolic, combined)
47    - re-implemented the EnvRef module so that evalStream works properly
48      (if the stream contains references to "use", "CM.make", etc.)
49    - cleaned up evalloop.sml and interact.sml (but they need more cleaning)
50    
51    ----------------------------------------------------------------------
52    Name: Matthias Blume
53    Date: 2001/08/20 15:50 EDT
54    Tag: blume20010820-slipup
55    Description:
56    
57    I forgot to commit a few files.  Here they are...
58    
59    ----------------------------------------------------------------------
60    Name: Matthias Blume
61    Date: 2001/08/20 15:35:00 EDT
62    Tag: blume-20010820-debugprof
63    Description:
64    
65    !!!! NEW BOOTFILES !!!!
66    
67    This is another round of reorganizing the compiler sources.  This
68    time the main goal was to factor out all the "instrumentation"
69    passes (for profiling and backtracing) into their own library.
70    The difficulty was to do it in such a way that it does not depend
71    on elaborate.cm but only on elabdata.cm.
72    
73    Therefore there have been further changes to both elaborate.cm and
74    elabdata.cm -- more "generic" things have been moved from the former
75    to the latter.  As a result, I was forced to split the assignment
76    of numbers indicating "primtyc"s into two portions: SML-generic and
77    SML/NJ-specific.  Since it would have been awkward to maintain,
78    I bit the bullet and actually _changed_ the mapping between these
79    numbers and primtycs.  The bottom line of this is that you need
80    a new set of bin- and bootfiles.
81    
82    I have built new bootfiles for all architectures, so doing a fresh
83    checkout and config/install.sh should be all you need.
84    
85    The newly created library's name is
86    
87        $smlnj/viscomp/debugprof.cm
88    
89    and its sources live under
90    
91        src/compiler/DebugProf
92    
93    ----------------------------------------------------------------------
94    Name: Matthias Blume
95    Date: 2001/08/15 17:15:00 EDT
96    Tag: blume-20010815-compreorg
97    Description:
98    
99    This is a first cut at reorganizing the CM libraries that make up the
100    core of the compiler.  The idea is to separate out pieces that could
101    be used independently by tools, e.g., the parser, the typechecker, etc.
102    
103    The current status is a step in this direction, but it is not quite
104    satisfactory yet.  Expect more changes in the future.
105    
106    Here is the current (new) organization...
107    
108        What used to be $smlnj/viscomp/core.cm is now divided into
109        six CM libraries:
110    
111             $smlnj/viscomp/basics.cm
112                           /parser.cm
113                           /elabdata.cm
114                           /elaborate.cm
115                           /execute.cm
116                           /core.cm
117    
118        The CM files for these libraries live under src/system/smlnj/viscomp.
119        All these libraries are proxy libraries that contain precisely
120        one CM library component.  Here are the locations of the components
121        (all within the src/compiler tree):
122    
123             Basics/basics.cm
124             Parse/parser.cm
125             ElabData/elabdata.cm
126             Elaborator/elaborate.cm
127             Execution/execute.cm
128             core.cm
129    
130         [This organization is the same that has been used already
131         for a while for the architecture-specific parts of the visible
132         compiler and for the old version of core.cm.]
133    
134         As you will notice, many source files have been moved from their
135         respective original locations to a new home in one of the above
136         subtrees.
137    
138         The division of labor between the new libraries is the following:
139    
140             basics.cm:
141                - Simple, basic definitions that pertain to many (or all) of
142                  the other libraries.
143             parser.cm:
144                - The SML parser, producing output of type Ast.dec.
145                - The type family for Ast is also defined and exported here.
146             elabdata.cm:
147                - The datatypes that describe input and output of the elaborator.
148                  This includes types, absyn, and static environments.
149             elaborator.cm:
150                - The SML/NJ type checker and elaborator.
151                  This maps an Ast.dec (with a given static environment) to
152                  an Absyn.dec (with a new static environment).
153                - This libraries implements certain modules that used to be
154                  structures as functors (to remove dependencies on FLINT).
155             execute.cm:
156                - Everything having to do with executing binary code objects.
157                - Dynamic environments.
158             core.cm:
159                - SML/NJ-specific instantiations of the elaborator and MLRISC.
160                - Top-level modules.
161                - FLINT (this should eventually become its own library)
162    
163    Notes:
164    
165    I am not 100% happy with the way I separated the elaborator (and its
166    data structures) from FLINT.  Two instances of the same problem:
167    
168        1. Data structures contain certain fields that carry FLINT-specific
169           information.  I hacked around this using exn and the property list
170           module from smlnj-lib.  But the fact that there are middle-end
171           specific fields around at all is a bit annoying.
172    
173        2. The elaborator calculates certain FLINT-related information.  I tried
174           to make this as abstract as I could using functorization, but, again,
175           the fact that the elaborator has to perform calculations on behalf
176           of the middle-end at all is not nice.
177    
178        3. Having to used exn and property lists is unfortunate because it
179           weakens type checking.  The other alternative (parameterizing
180           nearly *everything*) is not appealing, though.
181    
182    I removed the "rebinding =" warning hack because due to the new organization
183    it was awkward to maintain it.  As a result, the compiler now issues some of
184    these warnings when compiling init.cmi during bootstrap compilation. On
185    the plus side, you also get a warning when you do, for example:
186       val op = = Int32.+
187    which was not the case up to now.
188    
189    I placed "assign" and "deref" into the _Core structure so that the
190    code that deals with the "lazy" keyword can find them there.  This
191    removes the need for having access to the primitive environment
192    during elaboration.
193    
194    ----------------------------------------------------------------------
195    Name: Matthias Blume
196    Date: 2001/08/13
197    Tag: blume-20010813-closures
198    Description:
199    
200    This fix was sent to us by Zhong Shao.  It is supposed to improve the
201    performance of certain loops by avoiding needless closure allocation.
202    
203    ----------------------------------------------------------------------
204    Name: Lal George
205    Date: 2001/07/31 10:03:23 EDT 2001
206    Tag: george-20010731-x86-fmalloc
207    Description: Fixed bug in x86 calls
208    
209        There was a bug where call instructions would mysteriously
210        vanish. The call instruction had to be one that returned
211        a floating point value.
212    
213    ----------------------------------------------------------------------
214    Name: Lal George
215    Date: 2001/07/19 16:36:29 EDT 2001
216    Tag: george-20010719-simple-cells
217    Description:
218    
219    I have dramatically simplified the interface for CELLS in MLRISC.
220    
221    In summary, the cells interface is broken up into three parts:
222    
223      1. CellsBasis : CELLS_BASIS
224    
225            CellsBasis is a top level structure and common for all
226            architectures.  it contains the definitions of basic datatypes
227            and utility  functions over these types.
228    
229      2. functor Cells() : CELLS
230    
231            Cells generates an interface for CELLS that incorporates the
232            specific resources on the target architecture, such as the
233            presence of special register classes, their number and size,
234            and various useful substructures.
235    
236      3. <ARCH>CELLS
237    
238            e.g. SparcCells: SPARCCELLS
239    
240            <ARCH>CELLS usually contains additional bindings for special
241            registers  on the architecture, such as:
242    
243                    val r0 : cell           (* register zero *)
244                    val y : cell            (* Y register *)
245                    val psr : cell          (* processor status register *)
246                    ...
247    
248            The structure returned by applying the Cells functor is opened
249            in this interface.
250    
251    The main implication of all this is that the datatypes for cells is
252    split between CellsBasis and CELLS -- a fairly simple change for user
253    code.
254    
255    In the old scheme the CELLS interface had a definitional binding of
256    the form:
257    
258            signature CELLS = sig
259    
260               structure CellsBasis = CellsBasis
261    
262               ...
263    
264            end
265    
266    With all the sharing constraints that goes on in MLRISC, this old
267    design  quickly leads to errors such as:
268    
269            "structure definition spec inside of sharing ... "
270    
271    
272    and appears to require an unacceptable amount of sharing and where
273    constraint hackery.
274    
275    I think this error message (the interaction of definitional specs and
276    sharing) requires more explanation on our web page.
277    
278    ----------------------------------------------------------------------
279    Name: Matthias Blume
280    Date: 2001/07/19 15:00:00 EDT
281    Tag: blume-20010719-libreorg
282    Description:
283    
284    This update puts together a fairly extensive but straightforward change
285    to the way the libraries that implement the interactive system are
286    organized:
287    
288       The biggest change is the elimination of structure Compiler.  As a
289       replacement for this structure, there is now a CM library
290       (known as $smlnj/compiler.cm or $smlnj/compiler/current.cm)
291       that exports all the substructures of the original structure Compiler
292       directly.  So instead of saying Compiler.Foo.bar one now simply
293       says Foo.bar.  (The CM libraries actually export a collection of
294       structures that is richer than the collection of substructures of
295       structure Compiler.)
296    
297       To make the transition smooth, there is a separate library called
298       $smlnj/compiler/compiler.cm which puts together and exports the
299       original structure Compiler (or at least something very close to it).
300    
301       There are five members of the original structure Compiler
302       that are not exported directly but which instead became members
303       of a new structure Backend (described by signature BACKEND).  These are:
304       structure Profile (: PROFILE), structure Compile (: COMPILE), structure
305       Interact (: INTERACT), structure Machine (: MACHINE), and val
306       architecture (: string).
307    
308       Structure Compiler.Version has become structure CompilerVersion.
309    
310       Cross-compilers for alpha32, hppa, ppc, sparc, and x86 are provided
311       by $smlnj/compiler/<arch>.cm where <arch> is alpha32, hppa, ppc, sparc,
312       or x86, respectively.
313       Each of these exports the same frontend structures that
314       $smlnj/compiler.cm exports.  But they do not have a structure Backend
315       and instead export some structure <Arch>Backend where <Arch> is Alpha32,
316       Hppa, PPC, Sparc, or X86, respectively.
317    
318       Library $smlnj/compiler/all.cm exports the union of the exports of
319       $smlnj/compiler/<arch>.cm
320    
321       There are no structures <Arch>Compiler anymore, use
322       $smlnj/compiler/<arch>.cm instead.
323    
324       Library host-compiler-0.cm is gone.  Instead, the internal library
325       that instantiates CM is now called cm0.cm.  Selection of the host
326       compiler (backend) is no longer done here but. (Responsibility for it
327       now lies with $smlnj/compiler/current.cm.  This seems to be more
328       logical.)
329    
330       Many individual files have been moved or renamed.  Some files have
331       been split into multiple files, and some "dead" files have been deleted.
332    
333    Aside from these changes to library organization, there are also changes
334    to the way the code itself is organized:
335    
336       Structure Binfile has been re-implemented in such a way that it no
337       longer needs any knowledge of the compiler.  It exclusively deals
338       with the details of binfile layout.  It no longer invokes the
339       compiler (for the purpose of creating new prospective binfile
340       content), and it no longer has any knowledge of how to interpret
341       pickles.
342    
343       Structure Compile (: COMPILE) has been stripped down to the bare
344       essentials of compilation.  It no longer deals with linking/execution.
345       The interface has been cleaned up considerably.
346    
347       Utility routines for dealing with linking and execution have been
348       moved into their own substructures.
349    
350       (The ultimate goal of these changes is to provide a light-weight
351       binfile loader/linker (at least for, e.g., stable libraries) that
352       does not require CM or the compiler to be present.)
353    
354    CM documentation has been updated to reflect the changes to library
355    organization.
356    
357    ----------------------------------------------------------------------
358    Name: Matthias Blume
359    Date: 2001/07/10 17:30:00 EDT
360    Tag: Release_110_34
361    Description:
362    
363    Minor tweak to 110.34 (re-tagged):
364    
365      - README.html file added to CVS repository
366      - runtime compiles properly under FreeBSD 3.X and 4.X
367    
368    ----------------------------------------------------------------------
369    Name: Matthias Blume
370    Date: 2001/07/10 17:30:00 EDT
371    Tag: Release_110_34
372    Description:
373    
374    New version number (110.34). New bootfiles.
375    
376    ----------------------------------------------------------------------
377    Name: Matthias Blume
378    Date: 2001/07/09 16:00:00 EDT
379    Tag: blume-20010709-more-varargs
380    Description:
381    
382    I changed the handling of varargs in ml-nlffigen again:
383    The ellipsis ... will now simply be ignored (with an accompanying warning).
384    
385    The immediate effect is that you can actually call a varargs function
386    from ML -- but you can't actually supply any arguments beyond the ones
387    specified explicitly.  (For example, you can call printf with its format
388    string, but you cannot pass additional arguments.)
389    
390    This behavior is only marginally more useful than the one before, but
391    it has the advantage that a function or, more importantly, a function
392    type never gets dropped on the floor, thus avoiding follow-up problems with
393    other types that refer to the offending one.
394    
395    ----------------------------------------------------------------------
396    Name: Matthias Blume
397    Date: 2001/07/09 11:25:00 EDT
398    Tag: blume-20010709-varargs
399    Description:
400    
401    1. ckit-lib.cm now exports structure Error
402    2. ml-nlffigen reports occurences of "..." (i.e., varargs function types)
403       with a warning accompanied by a source location.  Moreover, it
404       merely skips the offending function or type and proceeds with the
405       rest of its work.u  As a result, one can safely feed C code containing
406       "..." to ml-nlffigen.
407    3. There are some internal improvements to CM, providing slightly
408       more general string substitutions in the tools subsystem.
409    
410    ----------------------------------------------------------------------
411    Name: Matthias Blume
412    Date: 2001/06/27 15:10:00 EDT
413    Tag: blume-20010627-concur
414    Description:
415    
416    Fixed a small bug in CM's handling of parallel compilation.
417    (You could observe the bug by Control-C-interrupting an ordinary
418    CMB.make or CM.stabilize and then attaching some compile servers.
419    The result was that all of a sudden the previously interrupted
420    compilation would continue on its own.  This was because of
421    an over-optimization: CM did not bother to clean out certain queues
422    when no servers were attached "anyway", resulting in the contents
423    of these queues to grab control when new servers did get attached.)
424    
425    There is also another minor update to the CM manual.
426    
427    ----------------------------------------------------------------------
428    Name: Matthias Blume
429    Date: 2001/06/26 16:15:00 EDT
430    Tag: blume-20010626-cmdoc
431    Description:
432    
433    Minor typo fixed in CM manual (syntax diagram for libraries).
434    
435    ----------------------------------------------------------------------
436    Name: Matthias Blume
437    Date: 2001/06/25 22:55:00 EDT
438    Tag: blume-20010625-x86pc
439    Description:
440    
441    Fixed a nasty bug in the X86 assembly code that caused signal
442    handlers to fail (crash) randomly.
443    
444    ----------------------------------------------------------------------
445    Name: Matthias Blume
446    Date: 2001/06/25 12:05:00 EDT
447    Tag: blume-20010625-nlffigen
448    Description:
449    
450    This update fixes a number of minor bugs in ml-nlffigen as reported by
451    Nick Carter <nbc@andrew.cmu.edu>.
452    
453      1. Silly but ok typedefs of the form "typedef void myvoid;" are now accepted.
454      2. Default names for generated files are now derived from the name of
455         the C file *without its directory*.  In particular, this causes generated
456         files to be placed locally even if the C file is in some system directory.
457      3. Default names for generated signatures and structures are also derived
458         from the C file name without its directory.  This avoids silly things
459         like "structure GL/GL".
460         (Other silly names are still possible because ml-nlffigen does not do
461          a thorough check of whether generated names are legal ML identifiers.
462          When in doubt, use command line arguments to force particular names.)
463    
464    ----------------------------------------------------------------------
465    Name: Matthias Blume
466    Date: 2001/06/21 12:25:00 EDT
467    Tag: blume-20010621-eXene
468    Description:
469    
470    eXene now compiles and (sort of) works again.
471    
472    The library name (for version > 110.33) is $/eXene.cm.
473    
474    I also added an new example in src/eXene/examples/nbody.  See the
475    README file there for details.
476    
477    ----------------------------------------------------------------------
478    Name: Matthias Blume
479    Date: 2001/06/20 16:40:00 EDT
480    Tag: blume-20010620-cml
481    Description:
482    
483    CML now compiles and works again.
484    
485    Libraries (for version > 110.33):
486    
487      $cml/cml.cm            Main CML library.
488      $cml/basis.cm          CML's version of $/basis.cm.
489      $cml/cml-internal.cm   Internal helper library.
490      $cml/core-cml.cm       Internal helper library.
491      $cml-lib/trace-cml.cm  Tracing facility.
492      $cml-lib/smlnj-lib.cm  CML's version of $/smlnj-lib.cm
493    
494    The installer (config/install.sh) has been taught how to properly
495    install this stuff.
496    
497    ----------------------------------------------------------------------
498    Name: Matthias Blume
499    Date: 2001/06/19 17:55:00 EDT
500    Tag: blume-20010619-instantiate
501    Description:
502    
503    This un-breaks the fix for bug 1432.
504    (The bug was originally fixed in 110.9 but I broke it again some
505    time after that.)
506    
507    ----------------------------------------------------------------------
508    Name: Matthias Blume
509    Date: 2001/06/19 17:25:00 EDT
510    Tag: blume-20010619-signals
511    Description:
512    
513    This should (hopefully) fix the long-standing signal handling bug.
514    (The runtime system was constructing a continuation record with an
515    incorrect descriptor which would cause the GC to drop data on the floor...)
516    
517    ----------------------------------------------------------------------
518    Name: Matthias Blume
519    Date: 2001/06/15 15:05:00 EDT
520    Tag: blume-20010615-moresparc
521    Description:
522    
523    Here is a short late-hour update related to Sparc c-calls:
524    
525     -- made handling of double-word arguments a bit smarter
526    
527     -- instruction selection phase tries to collapse certain clumsily
528        constructed ML-Trees; typical example:
529    
530            ADD(ty,ADD(_,e,LI d1),LI d2)  ->  ADD(ty,e,LI(d1+d2))
531    
532        This currently has no further impact on SML/NJ since mlriscGen does
533        not seem to generate such patterns in the first place, and c-calls
534        (which did generate them in the beginning) has meanwhile been fixed
535        so as to avoid them as well.
536    
537    ----------------------------------------------------------------------
538    Name: Matthias Blume
539    Date: 2001/06/15 15:05:00 EDT
540    Tag: blume-20010615-sparc
541    Description:
542    
543    The purpose of this update is to provide an implementation of NLFFI
544    on Sparc machines.
545    
546    Here are the changes in detail:
547    
548       * src/MLRISC/sparc/c-calls/sparc-c-calls.sml is a new file containing
549       the Sparc implementation of the c-calls API.
550       * The Sparc backend of SML/NJ has been modified to uniformely use %fp
551       for accessing the ML frame.  Thus, we have a real frame pointer and
552       can freely modify %sp without need for an omit-frame-ptr phase.
553       The vfp logic in src/compiler/CodeGen/* has been changed to accomodate
554       this case.
555       * ml-nlffigen has been taught to produce code for different architectures
556       and calling conventions.
557       * In a way similar to what was done in the x86 case, the Sparc
558       backend uses its own specific extension to mltree.  (For example,
559       it needs to be able to generate UNIMP instructions which are part
560       of the calling convention.)
561       * ml-nlffi-lib was reorganized to make it more modular (in particular,
562       to make it easier to plug in new machine- and os-dependent parts).
563    
564    There are some other fairly unrelated bug fixes and cleanups as well:
565    
566       * I further hacked the .cm files for MLRISC tools (like MDLGen) so
567       that they properly share their libraries with existing SML/NJ libraries.
568       * I fixed a minor cosmetic bug in CM, supressing certain spurious
569       follow-up error messages.
570       * Updates to CM/CMB documentation.
571    
572    TODO items:
573    
574       * MLRISC should use a different register as its asmTemp on the Sparc.
575         (The current %o2 is a really bad choice because it is part of the
576         calling conventions, so things might interfere in unexpected ways.)
577    
578    ----------------------------------------------------------------------
579    Name: Matthias Blume
580    Date: 2001/06/07
581    Tag: blume-20010607-calls
582    Description:
583    
584    A number of internal changes related to C calls and calling conventions:
585    
586    1. ML-Tree CALL statements now carry a "pops" field.  It indicates the
587       number of bytes popped implicitly (by the callee).  In most cases
588       this field is 0 but on x86/win32 it is some non-zero value.  This
589       is information provided for the benefit of the "omit-frameptr" pass.
590    2. The CALL instruction on the x86 carries a similar "pops" field.
591       The instruction selection phase copies its value from the ML-Tree
592       CALL statement.
593    3. On all other architectures, the instruction selection phase checks
594       whether "pops=0" and complains if not.
595    4. The c-calls implementation for x86 now accepts two calling conventions:
596       "ccall" and "stdcall".  When "ccall" is selected, the caller cleans
597       up after the call and pops is set to 0.  For "stdcall", the caller
598       does nothing, leaving the cleanup to the callee; pops is set to
599       the number of bytes that were pushed onto the stack.
600    5. The cproto decoder (compiler/Semant/types/cproto.sml) now can
601       distinguish between "ccall" and "stdcall".
602    6. The UNIMP instruction has been added to the supported Sparc instruction
603       set. (This is needed for implementing the official C calling convention
604       on this architecture.)
605    7. I fixed some of the .cm files under src/MLRISC/Tools to make them
606       work with the latest CM.
607    
608    ----------------------------------------------------------------------
609    Name: Matthias Blume
610    Date: 2001/06/05 15:10:00 EDT
611    Tag: blume-20010605-cm-index
612    Description:
613    
614    0. The "lambdasplit" parameter for class "sml" in CM has been documented.
615    
616    1. CM can now generate "index files".  These are human-readable files
617       that list on a per-.cm-file basis each toplevel symbol defined or
618       imported.  The location of the index file for
619       <p>/<d>.cm is <p>/CM/INDEX/<d>.cm.
620       To enable index-file generation, set CM.Control.generate_index to true
621       or export an environment-symbol: export CM_GENERATE_INDEX=true.
622    
623       The CM manual has been updated accordingly.
624    
625    2. I made some slight modifications to the c-calls API in MLRISC.
626    
627         a) There is now a callback to support saving/restoring of
628            dedicated but caller-save registers around the actual call
629            instruction.
630         b) One can optionally specify a comment-annotation for the
631            call instruction.
632    
633    3. SML/NJ (mlriscGen.sml) uses this new API for the rawccall primop.
634       (For example, the comment annotation shows the C prototype of
635        the function being called.)
636    
637    ----------------------------------------------------------------------
638    Name: Matthias Blume
639    Date: 2001/06/01 13:30:00 EDT
640    Tag: blume-20010601-nlffi-cleanup
641    Description:
642    
643    This is mostly a cleanup of MLFFI stuff:
644    
645       - some signature files have been put into a more exposed place
646       - the ugly 'f type parameter is gone (simplifies types tremendously!)
647       - ml-nlffigen changed accordingly
648       - tutorial updated
649    
650    Other changes:
651    
652       - author's affiliation in CM manual(s) updated
653       - some more recognized keywords added to Allen's sml.sty
654    
655    ----------------------------------------------------------------------
656    Name: Matthias Blume
657    Date: 2001/05/25 15:30:00 EDT
658    Tag: blume-20010525-iptr
659    Description:
660    
661      - put the official 110.33-README (as it appears on the ftp server) under
662        CVS
663      - fixed a small bug related to incomplete pointer types in
664        ml-nlffigen
665      - small cosmetic change to the ml-nlffi-lib's "arr" type constructor
666        (it does not need the 'f type parameter)
667    
668    ----------------------------------------------------------------------
669    Name: Matthias Blume
670    Date: 2001/05/23 14:30:00 EDT
671    Tag: Release_110_33
672    Description:
673    
674    New version number (110.33).  New bootfiles.
675    
676    ----------------------------------------------------------------------
677    Name: Matthias Blume
678    Date: 2001/05/22 18:06:00 EDT
679    Tag: blume-20010522-targets
680    Description:
681    
682    Made install.sh use file config/targets.customized if it exists, falling
683    back to config/targets if it doesn't.  This way one can have a customized
684    version of the targets file without touching the "real thing", thus
685    eliminating the constant fear of accidentally checking something bogus
686    back into the CVS repository...  (File config/targets.customized must
687    not be added to the repository!)
688    
689    ----------------------------------------------------------------------
690    Name: Matthias Blume
691    Date: 2001/05/22 16:30:00 EDT
692    Tag: blume-20010522-minitut
693    Description:
694    
695    1. Bug fix in ml-nlffigen; now (hopefully) correctly handling
696       struct returns.
697    2. Added src/ml-nlffi-lib/Doc/mini-tutorial.txt.  This is some very
698       incomplete, preliminary documentation for NLFFI.
699    
700    ----------------------------------------------------------------------
701    Name: Matthias Blume
702    Date: 2001/05/14 11:30:00 EDT
703    Tag: blume-20010514-script
704    Description:
705    
706    Some bugs in install script fixed.
707    
708    In addition to that I also made a slight change to the NLFFI API:
709    Functors generated by ml-nlffigen now take the dynamic library as a
710    straight functor argument, not as a suspended one.  (The original
711    functor code used to force the suspension right away anyway, so there
712    was nothing gained by this complication of the interface.)
713    
714    ----------------------------------------------------------------------
715    Name: Matthias Blume
716    Date: 2001/05/11 14:35:00 EDT
717    Tag: blume-20010511-ml-nlffi
718    Description:
719    
720    I finally took the plunge and added my new FFI code to the main
721    repository.  For x86-linux it is now ready for prime-time.
722    
723    There are two new subdirectories of "src":
724    
725      - ml-nlffi-lib:
726           The utility library for programs using the FFI interface.
727           Here is the implementation of $/c.cm and its associated low-level
728           partners $/c-int.cm and $/memory.cm.
729      - ml-nlffigen:
730           A stand-alone program for generating ML glue code from C source
731           code.
732    
733    Building ml-nlffigen requires $/ckit-lib.cm.
734    
735    The config/install.sh script has been updates to do the Right Thing
736    (hopefully).
737    
738    Notice that the source tree for the C-Kit will not be put under "src"
739    but directly under the installation root directory.  (This is the
740    structure that currently exists on the CVS server when you check out
741    module "sml".)  Fortunately, config/install.sh knows about this oddity.
742    
743    Bugs: No documentation yet.
744    
745    ----------------------------------------------------------------------
746    Name: Matthias Blume
747    Date: 2001/05/09 16:35:00 EDT
748    Tag: blume-20010509-cpscontract
749    Description:
750    
751    Fixed a bug in the accounting code in cpsopt/contract.sml.  (The
752    wrapper/unwrapper elimination did not decrement usage counts and some
753    dead variables got overlooked by the dead-up logic.)
754    
755    ----------------------------------------------------------------------
756    Name: Lal George
757    Date: 2001/05/08  17:26:09 EDT
758    Tag: george-20010508-omit-frameptr
759    Description:
760    
761    Changes to implement the omit-frame-pointer optimization to support
762    raw C calls. For now, there is only support on the Intel x86, but
763    other architectures will follow as more experience is gained with this.
764    
765    
766    ----------------------------------------------------------------------
767    Name: Matthias Blume
768    Date: 2001/05/07 14:40:00 EDT
769    Tag: blume-20010507-proxies
770    Description:
771    
772    I made into "proxy libraries" all libraries that qualify for such a
773    change.  (A qualifying library is a library that has another library or
774    groups as its sole member and repeats that member's export list
775    verbatim.  A proxy library avoids this repetition by omitting its export
776    list, effectively inheriting the list that its (only) member exports.
777    See the CM manual for more explanation.)
778    The main effect is that explicit export lists for these libraries
779    do not have to be kepts in sync, making maintenance a bit easier.
780    
781    I also added copyright notices to many .cm-files.
782    
783    Last but not least, I made a new set of bootfiles.
784    
785    ----------------------------------------------------------------------
786    Name: Matthias Blume
787    Date: 2001/05/04 17:00:00 EDT
788    Tag: blume-20010504-cm-lsplit
789    Description:
790    
791    0. John merged pending changes to $/smlnj-lib.cm
792    
793    1. Allen's previous change accidentally backed out of one of Lal's
794       earlier changes.  I undid this mistake (re-introducing Lal's change).
795    
796    2. I used the new topOrder' function from graph-scc.sml (from $/smlnj-lib.cm)
797       within the compiler where applicable.  There is some code simplification
798       because of that.
799    
800    3. The "split" phase (in FLINT) is now part of the default list of phases.
801       Compiler.Control.LambdaSplitting.* can be used to globally control the
802       lambda-splitting (cross-module-inlining) engine.  In addition to that,
803       it can now also be controlled on a per-source basis: CM has been taught
804       a new tool parameter applicable to ML source files.
805    
806       - To turn lambda-splitting off completely:
807            local open Compiler.Control.LambdaSplitting in
808                val _ = set Off
809            end
810       - To make "no lambda-splitting" the global default (but allow per-source
811         overriding); this is the initial setting:
812            local open Compiler.Control.LambdaSplitting in
813                val _ = set (Default NONE)
814            end
815       - To make "lambda-splitting with aggressiveness a" the global default
816         (and allow per-source overriding):
817            local open Compiler.Control.LambdaSplitting in
818                val _ = set (Default (SOME a))
819            end
820    
821       - To turn lambda-splitting off for a given ML souce file (say: a.sml)
822         write (in the respective .cm-file):
823            a.sml (lambdasplitting:off)
824       - To turn lambda-splitting for a.sml on with minimal aggressiveness:
825            a.sml (lambdasplitting:on)
826       - To turn lambda-splitting for a.sml on with aggressiveness <a> (where
827         <a> is a decimal non-negative integer):
828            a.sml (lambdasplitting:<a>)
829       - To turn lambda-splitting for a.sml on with maximal aggressiveness:
830            a.sml (lambdasplitting:infinity)
831       - To use the global default for a.sml:
832            a.sml (lambdasplitting:default)
833         or simply
834            a.sml
835    
836    ----------------------------------------------------------------------
837    Name: Allen Leung
838    Date: 2001/05/04 01:57:00 EDT
839    Tag: leunga-20010504-sync
840    Description:
841    
842      MLRISC features.
843    
844      1. Fix to CMPXCHG instructions.
845      2. Changed RA interface to allow annotations in callbacks.
846      3. Added a new method to the stream interface to allow annotations updates.
847    
848    ----------------------------------------------------------------------
849    Name: Matthias Blume
850    Date: 2001/05/01 11:45:00 EDT
851    Tag: blume-20010501-pcedittmp
852    Description:
853    
854    Changed install.sh to use the current working directory instead of
855    /usr/tmp for a temporary file (pcedittmp).  The previous choice
856    of /usr/tmp caused trouble with MacOS X because of file premission
857    problems.
858    
859    ----------------------------------------------------------------------
860    Name: Matthias Blume
861    Date: 2001/04/20 11:10:00 EDT
862    Tag: blume-20010420-inMLflag
863    Description:
864    
865     - added vp_limitPtrMask to vproc-state.h
866       (for use by the raw-C-calls mechanism to implement proper interrupt
867        handling)
868     - made the ML compiler aware of various data-structure offsets so it
869       can generate code for accessing the vp_inML flag and vp_limitPtrMask
870     - tweaked mlriscGen.sml to have it emit interrupt-handling code for
871       raw C-calls
872    
873    ----------------------------------------------------------------------
874    Name: Lal George
875    Date: 2001/04/20 09:15:28 EDT
876    Tag: george-20010420-macosX
877    Description:
878    
879     - Changes to port to Mac OS X; Darwin.
880    
881     - In the process I found that sqrt was broken on the PPC, because the
882       fsqrt instruction is not implemented.
883    
884    ----------------------------------------------------------------------
885    Name: Matthias Blume
886    Date: 2001/04/18 12:45:00 EDT
887    Tag: blume-20010418-ccalls
888    Description:
889    
890     - fixed two off-by-4 errors in the x86-specific c-calls implementation
891       (this bug prevented structure arguments containing pointers from being
892        passed correctly)
893     - changed the raw-C-call code in mlriscGen.sml in such a way that
894       structure arguments are represented as a pointer to the beginning
895       of the structure (instead of having a series of synthesized arguments,
896       one for each structure member)
897    
898     - made makeml script's verbosity level configurable via environment
899       variable (MAKEML_VERBOSITY)
900    
901     - eliminated placeholder implementations for f32l, w16s, i16s, and f32s
902       in rawmem-x86.sml; we are now using the real thing
903    
904    ----------------------------------------------------------------------
905    Name: Matthias Blume
906    Date: 2001/03/22 16:25:00 EST
907    Tag: blume-20010322-bootfiles
908    Description:
909    
910    Created a new set of bootfiles (for your automatic installation convenience).
911    
912    ----------------------------------------------------------------------
913    Name: Matthias Blume
914    Date: 2001/03/22 15:10:00 EST
915    Tag: blume-20010322-rawmem-parcm
916    Description:
917    
918    1. All "raw memory access" primitives for the new FFI are implemented now
919       (at least on the x86).
920    2. Some further cleanup of CM's parallel make mechanism.
921    
922    ----------------------------------------------------------------------
923    Name: Matthias Blume
924    Date: 2001/03/19 17:53:00 EST
925    Tag: blume-20010319-parallel
926    Description:
927    
928    Parallel make (using compile servers) now works again.
929    
930    To this end, CM.stabilize and CMB.make have been modified to work in
931    two passes when compile servers are attached:
932       1. Compile everything, do not perform stabilization; this pass
933          uses compile servers
934       2. Stabilize everything; this pass does not use compile servers
935    If there are no compile servers, the two passes are combined into one
936    (as before).  Splitting the passes increases the inherent parallelism
937    in the dependency graph because the entire graph including all
938    libraries is available at the same time.  This, in turn, improves
939    server utilization.  The downside is that the master process will
940    have to do some extra work after compilation is done (because for
941    technical reasons it must re-read all the binfiles during stabilization).
942    
943    ----------------------------------------------------------------------
944    Name: Matthias Blume
945    Date: 2001/03/16 12:22:00 EST
946    Tag: blume-20010316-bootfiles
947    Description:
948    
949    Created a new set of bootfiles (for your automatic installation convenience).
950    
951    ----------------------------------------------------------------------
952    Name: Matthias Blume
953    Date: 2001/03/16 11:00:00 EST
954    Tag: blume-20010316-MLTREE-fixup
955    Description:
956    
957    This is a minor fixup for an (untagged) earlier commit by Allen.
958    (A file was missing).
959    
960    ----------------------------------------------------------------------
961    Name: Allen Leung
962    Date: Mon Mar  5 18:54:57 EST 2001
963    Tag: leunga-20010305-cut-support
964    
965    1. New support for alternative control-flow in MLTREE.
966       Currently we support
967    
968          FLOW_TO(CALL ...., [k1,...,kn])
969    
970       This is needed for 'cuts to' in C-- and try/handle-like constructs
971       in Moby
972    
973       New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
974    
975    2. Register Allocator
976    
977       Changes in interface [from Fermin, John]
978    
979    3. Alpha 8-bit SLL support [Fermin]
980    
981    4. All architectures
982    
983       A new module (ClusterExpandCopies) for expanding parallel copies.
984    
985    ----------------------------------------------------------------------
986    Name: Allen Leung
987    Date: 2001/02/27 23:07:00 EST
988    Tag: leunga-20010227-minor-stuff
989    
990    1. Alpha bug fix for CMOVNE
991    2. Handle mltree COND(..,FCMP ...,...)
992    3. Bug fix in simplifier
993    
994    ----------------------------------------------------------------------
995    Name: Matthias Blume
996    Date: 2001/01/30 17:50:00 EST
997    Tag: blume-20010130-sync
998    Description:
999    
1000    This is just a minor update to sync my devel branch with the main brach.
1001    The only visible change is the addition of some README files.
1002    
1003    ----------------------------------------------------------------------
1004    Name: Matthias Blume
1005    Date: 2001/01/12 23:30:00 JST
1006    Tag: blume-20010112-bootfiles
1007    Description:
1008    
1009    Made a new set of bootfiles that goes with the current state of the
1010    repository.
1011    
1012    ----------------------------------------------------------------------
1013    Name: Matthias Blume
1014    Date: 2001/01/12 21:20:00 JST
1015    Tag: blume-20010112-sync
1016    Description:
1017    
1018    I am just flushing out some minor changes that had accumulated in
1019    my private branch in order to sync with the main tree.  (This is
1020    mainly because I had CVS trouble when trying to merge _into_ my
1021    private branch.)
1022    
1023    Most people should be completely unaffected by this.
1024    
1025    ----------------------------------------------------------------------
1026    Name: Allen Leung
1027    Date: Thu Jan 11 21:03:00 EST 2001
1028    Tag: leunga-20010111-labexp=mltree
1029    Description:
1030    
1031    1.  Removed the type LabelExp and replace it by MLTree.
1032    2.  Rewritten mltree-simplify with the pattern matcher tool.
1033    3.  There were some bugs in alpha code generator which would break
1034        64-bit code generation.
1035    4.  Redo the tools to generate code with the
1036    5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
1037        are now generated by perl scripts.
1038    
1039    ----------------------------------------------------------------------
1040    Name: Matthias Blume
1041    Date: 2001/01/10 21:55:00 JST
1042    Tag: blume-20010110-rcc
1043    Description:
1044    
1045    The RCC stuff now seems to work (but only on the x86).
1046    This required hacking of the c-calls interface (and -implementation) in
1047    MLRISC.
1048    
1049    Normal compiler users should be unaffected.
1050    
1051    ----------------------------------------------------------------------
1052    Name: Matthias Blume
1053    Date: 2001/01/09 01:20:00 JST
1054    Tag: blume-20010109-rcc
1055    Description:
1056    
1057    This is a fairly big patch, flushing out a large number of pending
1058    changes that I made to my development copy over the last couple of days.
1059    
1060    Of practical relevance at this moment is a workaround for a pickling
1061    bug that Allen ran into the other day.  The cause of the bug itself is
1062    still unknown and it might be hard to fix it properly, but the
1063    workaround has some merits of its own (namely somewhat reducing pickling
1064    overhead for certain libraries).  Therefore, I think this solution should
1065    be satisfactory at this time.
1066    
1067    The rest of the changes (i.e., the vast majority) has to do with my
1068    ongoing efforts of providing direct support for C function calls from
1069    ML.  At the moment there is a new primop "RAW_CCALL", typing magic
1070    in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
1071    case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
1072    translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
1073    handle RCC.
1074    
1075    The last part (the changes to mlriscGen.sml) are still known to be
1076    wrong on the x86 and not implemented on all other architectures.  But
1077    the infrastructure is in place. I had to change a few functor
1078    signatures in the backend to be able to route the CCalls interface
1079    from MLRISC there, and I had to specialize the mltree type (on the
1080    x86) to include the necessary extensions. (The extensions themselves
1081    were already there and redy to go in MLRISC/x86).
1082    
1083    Everything should be very happy as soon as someone helps me with
1084    mlriscGen.sml...
1085    
1086    In any case, nothing of this should matter to anyone as long as the
1087    new primop is not being used (which is going to be the case unless you
1088    find it where I hid it :). The rest of the compiler is completely
1089    unaffected.
1090    
1091    ----------------------------------------------------------------------
1092    Name: Matthias Blume
1093    Date: 2001/01/05 00:30:00 JST
1094    Tag: blume-20010105-primops
1095    Description:
1096    
1097    Added some experimental support for work that I am doing right now.
1098    These changes mostly concern added primops, but there is also a new
1099    experimental C library in the runtime system (but currently not enabled
1100    anywhere except on Linux/X86).
1101    
1102    In the course of adding primops (and playing with them), I discovered that
1103    Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
1104    broken.  (Zhong was very right he labeled this stuff as "major gross hack".)
1105    To recover, I made type information in INL_PRIM mandatory and changed
1106    prim.sml as well as built-in.sml accordingly.  The InLine structure now
1107    has complete, correct type information (i.e., no bottom types).
1108    
1109    Since all these changes mean that we need new binfiles, I also bumped the
1110    version number to 110.32.1.
1111    
1112    ----------------------------------------------------------------------
1113    Name: Matthias Blume
1114    Date: 2000/12/30 22:10:00 JST
1115    Tag: blume-20001230-various
1116    Description:
1117    
1118    Added proxy libraries for MLRISC and let MLRISC libraries refer
1119    to each other using path anchors.  (See CM manual for explanation.)
1120    
1121    Updated CM documentation.
1122    
1123    Fixed some bugs in CM.
1124    
1125    Implemented "proxy" libraries (= syntactic sugar for CM).
1126    
1127    Added "-quiet" option to makeml and changed runtime system accordingly.
1128    
1129    Added cleanup handler for exportML to reset timers and compiler stats.
1130    
1131    ----------------------------------------------------------------------
1132    Name: Lal George
1133    Date: 2000/12/22 22:22:58 EST 2000
1134    Tag: Release_110_32
1135    Description:
1136    
1137            Infinite precision used throughout MLRISC.
1138            see MLRISC/mltree/machine-int.sig
1139    
1140    ----------------------------------------------------------------------
1141    Name: Matthias Blume
1142    Date: 2000/12/22 23:16:00 JST
1143    Tag: blume-20001222-warn
1144    Description:
1145    
1146    Corrected wording and formatting of some CM warning message which I
1147    broke in my previous patch.
1148    
1149    ----------------------------------------------------------------------
1150    Name: Matthias Blume
1151    Date: 2000/12/22 21:20:00 JST
1152    Tag: blume-20001222-anchorenv
1153    Description:
1154    
1155    Fixed CM's handling of anchor environments in connection with CMB.make.
1156    
1157    ----------------------------------------------------------------------
1158    Name: Matthias Blume
1159    Date: 2000/12/22 13:15:00 JST
1160    Tag: blume-20001222-cleanup
1161    Description:
1162    
1163    Removed src/cm/ffi which does not (and did not) belong here.
1164    
1165    ----------------------------------------------------------------------
1166    Name: Matthias Blume
1167    Date: 2000/12/21 23:55:00 JST
1168    Tag: blume-20001221-exn
1169    Description:
1170    
1171    Probably most important: CM no longer silently swallows all exceptions
1172    in the compiler.
1173    Plus: some other minor CM changes.  For example, CM now reports some
1174    sizes for generated binfiles (code, data, envpickle, lambdapickle).
1175    
1176    ----------------------------------------------------------------------
1177    Name: Matthias Blume
1178    Date: 2000/12/15 00:01:05 JST
1179    Tag: blume-20001215-dirtool
1180    Description:
1181    
1182    - "dir" tool added.
1183    - improvements and cleanup to Tools structure
1184    - documentation updates
1185    
1186    ----------------------------------------------------------------------
1187    Name: Allen Leung
1188    Date: Thu Dec 14 03:45:24 EST 2000
1189    Description:
1190    Tag:  leunga-20001214-int-inf
1191    Description:
1192    
1193       In IntInf, added these standard functions, which are missing from our
1194    implementation:
1195    
1196        andb : int * int -> int
1197        xorb : int * int -> int
1198        orb  : int * int -> int
1199        notb : int -> int
1200         <<   : int * word -> int
1201        ~>>  : int * word -> int
1202    
1203       Not tested, I hope they are correct.
1204    
1205    ----------------------------------------------------------------------
1206    Name: Allen Leung
1207    Date: Fri Dec  8 19:23:26 EST 2000
1208    Description:
1209    Tag:  leunga-20001208-nowhere
1210    Description:
1211    
1212      Slight improvements to the 'nowhere' tool to handle OR-patterns,
1213    to generate better error messages etc.  Plus a brief manual.
1214    
1215    ----------------------------------------------------------------------
1216    Name: Lal George
1217    Date: 2000/12/08 09:54:02 EST 2000
1218    Tag: Release_110_31
1219    Description:
1220    
1221    - Version 110.31
1222    ----------------------------------------------------------------------
1223    Name: Allen Leung
1224    Date: Thu Dec  7 22:01:04 EST 2000
1225    Tag:  leunga-20001207-cell-monster-hack
1226    Description:
1227    
1228    Major MLRISC internal changes.  Affect all clients.
1229    Summary:
1230    
1231    1.  Type CELLS.cell = int is now replaced by a datatype.
1232        As a result, the old regmap is now gone.  Almost all interfaces
1233        in MLRISC change as a consequence.
1234    
1235    2.  A new brand version of machine description tool (v3.0) that generates
1236        modules expecting the new interface.  The old version is removed.
1237    
1238    3.  The RA interface has been further abstracted into two new functors.
1239        RISC_RA and X86RA.  These functors have much simpler interfaces.
1240        [See also directory MLRISC/demo.]
1241    
1242    4.  Some other new source->source code generation tools are available:
1243    
1244        a. MLRISC/Tools/RewriteGen -- generate rewriters from rules.
1245        b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules.
1246           I use this tool to generate the peephole optimizers---with the new
1247           cell type changes, peephole rules are becoming difficult to write
1248           without conditional pattern matching.
1249    
1250    5.  More Intmap -> IntHashTable change.  Previous changes by Matthias didn't
1251        cover the entire MLRISC source tree so many things broke.
1252    
1253    6.  CM files have been moved to the subdirectory MLRISC/cm.
1254        They are moved because there are a lot of them and they clutter up the
1255        root dir.
1256    
1257    7.  More detailed documentation to come...
1258    
1259        NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do
1260        a makeml -rebuild first.  This is because of other other
1261        changes that Matthias has made (see below).
1262    
1263    
1264    ----------------------------------------------------------------------
1265    Name: Matthias Blume
1266    Date: 2000/11/30 23:12:00 JST
1267    Tag: blume-20001130-filereorg
1268    Description:
1269    
1270    Some manual updates and some file reorganizations in CM.
1271    
1272    ----------------------------------------------------------------------
1273    Name: Matthias Blume
1274    Date: 2000/11/24 17:45:00 JST
1275    Tag: blume-20001124-link
1276    Description:
1277    
1278    Drastically improved link traversal code for the case that the dynamic
1279    value was already loaded at bootstrap time.  As a result, CM and CMB
1280    now both load blazingly fast -- even on a very slow machine.  Also,
1281    memory consumption has been further reduced by this.
1282    
1283    Warning: The format of the PIDMAP file has changed.  THerefore, to
1284    bootstrap you have to do this:
1285    
1286    1. Run CMB.make
1287    2. Make a symbolic link for the boot directory:
1288         ln -s sml.boot.ARCH-OS xxx
1289    3. "Rebuild" the boot directory:
1290         ./makeml -boot xxx -rebuild sml ; rm xxx
1291    4. Boot normally:
1292          ./makeml
1293    
1294    ----------------------------------------------------------------------
1295    Name: Matthias Blume
1296    Date: 2000/11/21 21:20:00 JST
1297    Tag: blume-20001121-tools
1298    Description:
1299    
1300    Continued hacking on autoloading problem -- with success this time.
1301    Also changed tool-plugin mechanism.  See new CM manual.
1302    
1303    ----------------------------------------------------------------------
1304    Name: Matthias Blume
1305    Date: 2000/11/19 14:30:00 JST
1306    Tag:  blume-20001119-autoload
1307    Description:
1308    
1309    Some hacking to make autoloading faster.  Success for CMB, no success
1310    so far for CM.  There is a reduced structure CM' that autoloads faster.
1311    (This is a temporary, non-documented hack to be eliminated again when
1312    the general problem is solved.)
1313    
1314    ----------------------------------------------------------------------
1315    Name: Matthias Blume
1316    Date: 2000/11/17 14:10:00 JST
1317    Tag: blume-20001117-pickle-lib
1318    Description:
1319    
1320    1. Eliminated comp-lib.cm
1321    2. Made pickle-lib.cm
1322    3. Eliminated all uses of intset.sml (from comp-lib.cm)
1323    4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
1324       equivalent constructs from smlnj-lib.cm (INtHashTable).
1325    5. Point 4. also goes for those uses of intmap.* in MLRISC.
1326       Duplicated intmap modules thrown out.
1327    6. Hunted down all duplicated SCC code and replaced it with
1328       equivalent stuff (GraphSCCFn from smlnj-lib.cm).
1329    7. Rewrote Feedback module.
1330    8. Moved sortedlist.sml into viscomp-lib.cm.  Eventually it
1331       should be thrown out and equivalent modules from smlnj-lib.cm
1332       should be used (IntRedBlackSet, IntListSet, ...).
1333    
1334    Confirmed that compiler compiles to fixpoint.
1335    
1336    ----------------------------------------------------------------------
1337    Name: Allen Leung
1338    Date: 2000/11/10 18:00:00
1339    Tag: leunga-20001110-new-x86-fp
1340    
1341    A new x86 floating point code generator has been added.
1342    By default this is turned off.  To turn this on, do:
1343    
1344        CM.autoload "$smlnj/compiler.cm";
1345        Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
1346    
1347    Changes:
1348    
1349    1.  Changed FTAN to FPTAN so that the assembly output is correct.
1350    2.  Changed the extension callback for FTANGENT to generate:
1351    
1352              fptan
1353              fstp  %st(0)
1354        instead of
1355              fptan
1356              fstpl ftempmem
1357    
1358    3.  Numerous assembly fixes for x86.
1359    
1360    5.  Cleaned up the machine code output module x86/x86MC.sml and added
1361        support for a whole bunch of instructions and addressing modes:
1362    
1363          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st, %st(n)
1364          faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp  %st, %st(n)
1365          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st(n), %st
1366          fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
1367          fxch %st(n)
1368          fld %st(n)
1369          fst %st(n)
1370          fst mem
1371          fstp %st(n)
1372          fucom %st(n)
1373          fucomp %st(n)
1374    
1375        All these are now generated when the fast fp mode is turned on.
1376    
1377    6.  Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
1378    
1379    ----------------------------------------------------------------------
1380    Name: Matthias Blume
1381    Date: 2000/11/09 11:20:00 JST
1382    Tag: blume-20001109-scc
1383    Description:
1384    
1385    Eliminated some code duplication:
1386    
1387    1. Added "where" clause to GraphSCCFn in SML/NJ Library.
1388       (Otherwise the functor is useless.)
1389    2. Used GraphSCCFn where SCCUtilFun was used previously.
1390    3. Got rid of SCCUtilFun (in comp-lib.cm).
1391    
1392    ----------------------------------------------------------------------
1393    Name: Lal George
1394    Date: 2000/11/06 09:02:21 EST 2000
1395    Tag: Release_110_30
1396    Description:
1397    
1398    - Version 110.30
1399    ----------------------------------------------------------------------
1400    Name: Matthias Blume
1401    Date: 2000/11/04 14:45:00
1402    Tag: blume-20001104-mlbuild
1403    Description:
1404    
1405    - Made ml-build faster on startup.
1406    - Documentation fixes.
1407    
1408    ----------------------------------------------------------------------
1409    Name: Matthias Blume
1410    Date: 2000/11/02 17:00:00 JST
1411    Tag: blume-20001102-condcomp
1412    Description:
1413    
1414    - Small tweaks to pickler -- new BOOTFILES!
1415    - Version bumped to 110.29.2.
1416    - Added conditional compilation facility to init.cmi (see comment there).
1417    ----------------------------------------------------------------------
1418    Name: Allen Leung
1419    Date: 2000/10/23 19:31:00
1420    Tag: leunga-20001023-demo-ra
1421    
1422    1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
1423    2. Test programs for the graph library updated
1424    3. Some new MLRISC demo programs added
1425    
1426    ----------------------------------------------------------------------
1427    Name: Matthias Blume
1428    Date: 2000/08/31 22:15:00 JST
1429    Tag: blume-20001017-errmsg
1430    Description:
1431    
1432    More error message grief: Where there used to be no messages, there
1433    now were some that had bogus error regions.  Fixed.
1434    
1435    ----------------------------------------------------------------------
1436    Name: Matthias Blume
1437    Date: 2000/08/31 17:30:00 JST
1438    Tag: blume-20001017-v110p29p1
1439    Description:
1440    
1441    I made a version 110.29.1 with new bootfiles.
1442    
1443    Changes:  Modified pickler/unpickler for faster and leaner unpickling.
1444              CM documentation changes and a small bugfix in CM's error reporting.
1445    
1446    ----------------------------------------------------------------------
1447    Name: Lal George
1448    Date: 2000/09/27 14:42:35 EDT
1449    Tag: george-20000927-nodestatus
1450    Description:
1451    
1452    Changed the type of the nodestatus, so that:
1453    
1454            SPILLED(~1)             is now SPILLED
1455            SPILLED(m) where m>=0   is now MEMREG(m)
1456            SPILLED(s) where s<~1   is now SPILL_LOC(~s)
1457    
1458    ----------------------------------------------------------------------
1459    Name: Matthias Blume
1460    Date: 2000/09/07 14:45:00 JST
1461    Tag: blume-20000907-cmerrmsg
1462    Description:
1463    
1464    Small tweak to CM to avoid getting ML syntax error messages twice.
1465    
1466    ----------------------------------------------------------------------
1467    Name: Matthias Blume
1468    Date: 2000/08/31 18:00:00 JST
1469    Tag: blume-20000831-cvsbootfiles
1470    Description:
1471    
1472    New URL for boot files (because the 110.29 files on the BL server do
1473    now work correctly with my updated install scripts for yacc and lex).
1474    
1475    ----------------------------------------------------------------------
1476    Name: Matthias Blume
1477    Date: 2000/08/08 12:33:00 JST
1478    Tag: blume-20000808-manual
1479    Description:
1480    
1481    Tiny update to CM manual.
1482    
1483    ----------------------------------------------------------------------
1484    Name: Allen Leung
1485    Date: 2000/08/7 19:31:00
1486    Tag: leunga-20000807-a-whole-bunch-of-stuff
1487    
1488      Moby, C--, SSA, x86, machine descriptions etc.  Should only affect C--
1489    and Mobdy.
1490    
1491    1.  x86
1492    
1493       a.  Fixes to peephole module by John and Dan.
1494       b.  Assembly fix to SETcc by Allen.
1495       c.  Fix to c-call by John.
1496       d.  Fix to spilling by John.  (This one deals with the missing FSTPT case)
1497       e.  Instruction selection optimization to SETcc as suggested by John.
1498    
1499           For example,
1500    
1501            MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
1502    
1503           should generate:
1504    
1505            MOVL a, x
1506            SUBL b, x
1507            SHRL 31, x
1508    
1509    2.  IR stuff
1510    
1511         A bunch of new DJ-graph related algorithms added.  These
1512         speed up SSA construction.
1513    
1514    3.  SSA + Scheduling
1515    
1516         Added code for SSA and scheduling to the repository
1517    
1518    ----------------------------------------------------------------------
1519    Name: Lal George
1520    Date: 2000/07/27 11:53:14 EDT
1521    
1522    Tag: lal-20000727-linux-ppc
1523    Description:
1524    
1525     Made changes to support Linux PPC.
1526     p.s. I have confirmation that the 110.29 boot files work fine.
1527    
1528    ----------------------------------------------------------------------
1529    Name: Matthias Blume
1530    Date: 2000/07/27 17:40:00 JST
1531    Tag: blume-20000727-scripts
1532    Description:
1533    
1534    !!!! WARNING !!!!
1535    You must recompile the runtime system!
1536    !!!! WARNING !!!!
1537    
1538    This is basically another round of script-enhancements:
1539    
1540    1. sml, ml-build, and ml-makedepend accept options -D and -U to define
1541       and undefine CM preprocessor symbols.
1542    
1543    2. ml-build avoids generating a new heap image if it finds that the
1544       existing one is still ok.  (The condition is that no ML file had to
1545       be recompiled and all ML files are found to be older that the heap
1546       file.)
1547    
1548       To make this work smoothly, I also hacked the runtime system as
1549       well as SMLofNJ.SysInfo to get access to the heap image suffix
1550       (.sparc-solaris, ...) that is currently being used.
1551    
1552       Moreover, the signature of CM.mk_standalone has changed.  See the
1553       CM manual.
1554    
1555    3. ml-makedepend accepts additional options -n, -a, and -o.  (See the
1556       CM manual for details.)
1557    
1558    4. More CM manual updates:
1559        - all of the above has been documented.
1560        - there is now a section describing the (CM-related) command line
1561          arguments that are accepted by the "sml" command
1562    
1563    ----------------------------------------------------------------------
1564    Name: Matthias Blume
1565    Date: 2000/07/25 16:20:00 JST
1566    Tag: blume-20000725-makedepend
1567    Description:
1568    
1569    Added a script called ml-makedepend.  This can be used in makefiles
1570    for Unix' make in a way very similar to the "makedepend" command for
1571    C.
1572    
1573    The script internally uses function CM.sources.
1574    
1575    Synopsis:
1576    
1577        ml-makedepend [-f makefile] cmfile targetname
1578    
1579    The default for the makefile is "makefile" (or "Makefile" should
1580    "makefile" not exist).
1581    
1582    ml-makedepend adds a cmfile/targetname-specific section to this
1583    makefile (after removing the previous version of this section).  The
1584    section contains a single dependency specification with targetname on
1585    the LHS (targetname is an arbitrary name), and a list of files derived
1586    from the cmfile on the RHS.  Some of the files on the RHS are
1587    ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
1588    "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
1589    path names.  The makefile writer is responsible for making sure that
1590    these variables have correct at the time "make" is invoked.
1591    
1592    ----------------------------------------------------------------------
1593    Name: Matthias Blume
1594    Date: 2000/07/22 23:30:00 JST
1595    Tag: blume-20000722-urlupdate
1596    Description:
1597    
1598    Changed BOOT and config/srcarchiveurl to point to BL server:
1599    
1600        ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
1601    
1602    ----------------------------------------------------------------------
1603    Name: Matthias Blume
1604    Date: 2000/07/18 18:00:00 JST
1605    Tag: blume-20000718-Version_110_29
1606    Description:
1607    
1608    1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
1609    
1610    2. Updated config/version to 110.29
1611    
1612    3. Updated config/srcarchiveurl
1613    
1614    3. New boot files!
1615       ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
1616    
1617    ----------------------------------------------------------------------
1618    Name: Matthias Blume
1619    Date: 2000/07/11 13:58:00 JST
1620    Tag: blume-20000711-doctypo
1621    Description:
1622    
1623    Fixed a few typos in CM manual.
1624    
1625    ----------------------------------------------------------------------
1626    Name: Allen Leung
1627    Date: 2000/06/15 00:38:00
1628    Tag: leunga-20000704-sparc-x86
1629    
1630    1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
1631    2. fix to x86 RET bug [found by Dan Grossman]
1632    3. sparc assembly bug fix for ticc instructions [found by Fermin]
1633    
1634       Affects c-- and moby only
1635    
1636    ----------------------------------------------------------------------
1637    Name: Matthias Blume
1638    Date: 2000/07/04 15:26:00
1639    Tag: blume-20000704-trigger
1640    Description:
1641    
1642    1. Improvements to CM manual.
1643    2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
1644       of getting a back-trace.  The function, when called, raises an
1645       internal exception which explicitly carries the full back-trace history,
1646       so it is unaffected by any intervening handle-raise pairs ("trivial"
1647       or not).  The interactive loop will print that history once it arrives
1648       at top level.
1649       Short of having all exceptions implicitly carry the full history, the
1650       recommended way of using this facility is:
1651         - compile your program with instrumentation "on"
1652         - run it, when it raises an exception, look at the history
1653         - if the history is "cut off" because of some handler, go and modify
1654           your program so that it explicitly calls BTrace.trigger
1655         - recompile (still instrumented), and rerun; look at the full history
1656    
1657  ----------------------------------------------------------------------  ----------------------------------------------------------------------
1658  Name: Matthias Blume  Name: Matthias Blume
1659  Date: 2000/07/02 15:36:00 JST  Date: 2000/07/03 15:36:00 JST
1660  Tag: blume-20000702-manual  Tag: blume-20000702-manual
1661  Description:  Description:
1662    
1663  Small corrections and updates to CM manual.  Small corrections and updates to CM manual.
1664    
1665  ----------------------------------------------------------------------  ----------------------------------------------------------------------
1666  Name: Matthias Blumen  Name: Matthias Blume
1667  Date: 2000/06/29 16:04:00 JST  Date: 2000/06/29 16:04:00 JST
1668  Tag: blume-20000629-yacctool  Tag: blume-20000629-yacctool
1669  Description:  Description:

Legend:
Removed from v.680  
changed lines
  Added in v.918

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