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 760, Fri Dec 22 14:15:24 2000 UTC revision 903, Mon Aug 20 19:34:06 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  Name: Matthias Blume
17  Date: 2000/12/22 21:20:00 JST  Date: 2001/08/20 15:35:00 EDT
18  Tag: blume-20001222-anchorenv  Tag: blume-20010820-debugprof
19    Description:
20    
21    !!!! NEW BOOTFILES !!!!
22    
23    This is another round of reorganizing the compiler sources.  This
24    time the main goal was to factor out all the "instrumentation"
25    passes (for profiling and backtracing) into their own library.
26    The difficulty was to do it in such a way that it does not depend
27    on elaborate.cm but only on elabdata.cm.
28    
29    Therefore there have been further changes to both elaborate.cm and
30    elabdata.cm -- more "generic" things have been moved from the former
31    to the latter.  As a result, I was forced to split the assignment
32    of numbers indicating "primtyc"s into two portions: SML-generic and
33    SML/NJ-specific.  Since it would have been awkward to maintain,
34    I bit the bullet and actually _changed_ the mapping between these
35    numbers and primtycs.  The bottom line of this is that you need
36    a new set of bin- and bootfiles.
37    
38    I have built new bootfiles for all architectures, so doing a fresh
39    checkout and config/install.sh should be all you need.
40    
41    The newly created library's name is
42    
43        $smlnj/viscomp/debugprof.cm
44    
45    and its sources live under
46    
47        src/compiler/DebugProf
48    
49    ----------------------------------------------------------------------
50    Name: Matthias Blume
51    Date: 2001/08/15 17:15:00 EDT
52    Tag: blume-20010815-compreorg
53    Description:
54    
55    This is a first cut at reorganizing the CM libraries that make up the
56    core of the compiler.  The idea is to separate out pieces that could
57    be used independently by tools, e.g., the parser, the typechecker, etc.
58    
59    The current status is a step in this direction, but it is not quite
60    satisfactory yet.  Expect more changes in the future.
61    
62    Here is the current (new) organization...
63    
64        What used to be $smlnj/viscomp/core.cm is now divided into
65        six CM libraries:
66    
67             $smlnj/viscomp/basics.cm
68                           /parser.cm
69                           /elabdata.cm
70                           /elaborate.cm
71                           /execute.cm
72                           /core.cm
73    
74        The CM files for these libraries live under src/system/smlnj/viscomp.
75        All these libraries are proxy libraries that contain precisely
76        one CM library component.  Here are the locations of the components
77        (all within the src/compiler tree):
78    
79             Basics/basics.cm
80             Parse/parser.cm
81             ElabData/elabdata.cm
82             Elaborator/elaborate.cm
83             Execution/execute.cm
84             core.cm
85    
86         [This organization is the same that has been used already
87         for a while for the architecture-specific parts of the visible
88         compiler and for the old version of core.cm.]
89    
90         As you will notice, many source files have been moved from their
91         respective original locations to a new home in one of the above
92         subtrees.
93    
94         The division of labor between the new libraries is the following:
95    
96             basics.cm:
97                - Simple, basic definitions that pertain to many (or all) of
98                  the other libraries.
99             parser.cm:
100                - The SML parser, producing output of type Ast.dec.
101                - The type family for Ast is also defined and exported here.
102             elabdata.cm:
103                - The datatypes that describe input and output of the elaborator.
104                  This includes types, absyn, and static environments.
105             elaborator.cm:
106                - The SML/NJ type checker and elaborator.
107                  This maps an Ast.dec (with a given static environment) to
108                  an Absyn.dec (with a new static environment).
109                - This libraries implements certain modules that used to be
110                  structures as functors (to remove dependencies on FLINT).
111             execute.cm:
112                - Everything having to do with executing binary code objects.
113                - Dynamic environments.
114             core.cm:
115                - SML/NJ-specific instantiations of the elaborator and MLRISC.
116                - Top-level modules.
117                - FLINT (this should eventually become its own library)
118    
119    Notes:
120    
121    I am not 100% happy with the way I separated the elaborator (and its
122    data structures) from FLINT.  Two instances of the same problem:
123    
124        1. Data structures contain certain fields that carry FLINT-specific
125           information.  I hacked around this using exn and the property list
126           module from smlnj-lib.  But the fact that there are middle-end
127           specific fields around at all is a bit annoying.
128    
129        2. The elaborator calculates certain FLINT-related information.  I tried
130           to make this as abstract as I could using functorization, but, again,
131           the fact that the elaborator has to perform calculations on behalf
132           of the middle-end at all is not nice.
133    
134        3. Having to used exn and property lists is unfortunate because it
135           weakens type checking.  The other alternative (parameterizing
136           nearly *everything*) is not appealing, though.
137    
138    I removed the "rebinding =" warning hack because due to the new organization
139    it was awkward to maintain it.  As a result, the compiler now issues some of
140    these warnings when compiling init.cmi during bootstrap compilation. On
141    the plus side, you also get a warning when you do, for example:
142       val op = = Int32.+
143    which was not the case up to now.
144    
145    I placed "assign" and "deref" into the _Core structure so that the
146    code that deals with the "lazy" keyword can find them there.  This
147    removes the need for having access to the primitive environment
148    during elaboration.
149    
150    ----------------------------------------------------------------------
151    Name: Matthias Blume
152    Date: 2001/08/13
153    Tag: blume-20010813-closures
154    Description:
155    
156    This fix was sent to us by Zhong Shao.  It is supposed to improve the
157    performance of certain loops by avoiding needless closure allocation.
158    
159    ----------------------------------------------------------------------
160    Name: Lal George
161    Date: 2001/07/31 10:03:23 EDT 2001
162    Tag: george-20010731-x86-fmalloc
163    Description: Fixed bug in x86 calls
164    
165        There was a bug where call instructions would mysteriously
166        vanish. The call instruction had to be one that returned
167        a floating point value.
168    
169    ----------------------------------------------------------------------
170    Name: Lal George
171    Date: 2001/07/19 16:36:29 EDT 2001
172    Tag: george-20010719-simple-cells
173    Description:
174    
175    I have dramatically simplified the interface for CELLS in MLRISC.
176    
177    In summary, the cells interface is broken up into three parts:
178    
179      1. CellsBasis : CELLS_BASIS
180    
181            CellsBasis is a top level structure and common for all
182            architectures.  it contains the definitions of basic datatypes
183            and utility  functions over these types.
184    
185      2. functor Cells() : CELLS
186    
187            Cells generates an interface for CELLS that incorporates the
188            specific resources on the target architecture, such as the
189            presence of special register classes, their number and size,
190            and various useful substructures.
191    
192      3. <ARCH>CELLS
193    
194            e.g. SparcCells: SPARCCELLS
195    
196            <ARCH>CELLS usually contains additional bindings for special
197            registers  on the architecture, such as:
198    
199                    val r0 : cell           (* register zero *)
200                    val y : cell            (* Y register *)
201                    val psr : cell          (* processor status register *)
202                    ...
203    
204            The structure returned by applying the Cells functor is opened
205            in this interface.
206    
207    The main implication of all this is that the datatypes for cells is
208    split between CellsBasis and CELLS -- a fairly simple change for user
209    code.
210    
211    In the old scheme the CELLS interface had a definitional binding of
212    the form:
213    
214            signature CELLS = sig
215    
216               structure CellsBasis = CellsBasis
217    
218               ...
219    
220            end
221    
222    With all the sharing constraints that goes on in MLRISC, this old
223    design  quickly leads to errors such as:
224    
225            "structure definition spec inside of sharing ... "
226    
227    
228    and appears to require an unacceptable amount of sharing and where
229    constraint hackery.
230    
231    I think this error message (the interaction of definitional specs and
232    sharing) requires more explanation on our web page.
233    
234    ----------------------------------------------------------------------
235    Name: Matthias Blume
236    Date: 2001/07/19 15:00:00 EDT
237    Tag: blume-20010719-libreorg
238    Description:
239    
240    This update puts together a fairly extensive but straightforward change
241    to the way the libraries that implement the interactive system are
242    organized:
243    
244       The biggest change is the elimination of structure Compiler.  As a
245       replacement for this structure, there is now a CM library
246       (known as $smlnj/compiler.cm or $smlnj/compiler/current.cm)
247       that exports all the substructures of the original structure Compiler
248       directly.  So instead of saying Compiler.Foo.bar one now simply
249       says Foo.bar.  (The CM libraries actually export a collection of
250       structures that is richer than the collection of substructures of
251       structure Compiler.)
252    
253       To make the transition smooth, there is a separate library called
254       $smlnj/compiler/compiler.cm which puts together and exports the
255       original structure Compiler (or at least something very close to it).
256    
257       There are five members of the original structure Compiler
258       that are not exported directly but which instead became members
259       of a new structure Backend (described by signature BACKEND).  These are:
260       structure Profile (: PROFILE), structure Compile (: COMPILE), structure
261       Interact (: INTERACT), structure Machine (: MACHINE), and val
262       architecture (: string).
263    
264       Structure Compiler.Version has become structure CompilerVersion.
265    
266       Cross-compilers for alpha32, hppa, ppc, sparc, and x86 are provided
267       by $smlnj/compiler/<arch>.cm where <arch> is alpha32, hppa, ppc, sparc,
268       or x86, respectively.
269       Each of these exports the same frontend structures that
270       $smlnj/compiler.cm exports.  But they do not have a structure Backend
271       and instead export some structure <Arch>Backend where <Arch> is Alpha32,
272       Hppa, PPC, Sparc, or X86, respectively.
273    
274       Library $smlnj/compiler/all.cm exports the union of the exports of
275       $smlnj/compiler/<arch>.cm
276    
277       There are no structures <Arch>Compiler anymore, use
278       $smlnj/compiler/<arch>.cm instead.
279    
280       Library host-compiler-0.cm is gone.  Instead, the internal library
281       that instantiates CM is now called cm0.cm.  Selection of the host
282       compiler (backend) is no longer done here but. (Responsibility for it
283       now lies with $smlnj/compiler/current.cm.  This seems to be more
284       logical.)
285    
286       Many individual files have been moved or renamed.  Some files have
287       been split into multiple files, and some "dead" files have been deleted.
288    
289    Aside from these changes to library organization, there are also changes
290    to the way the code itself is organized:
291    
292       Structure Binfile has been re-implemented in such a way that it no
293       longer needs any knowledge of the compiler.  It exclusively deals
294       with the details of binfile layout.  It no longer invokes the
295       compiler (for the purpose of creating new prospective binfile
296       content), and it no longer has any knowledge of how to interpret
297       pickles.
298    
299       Structure Compile (: COMPILE) has been stripped down to the bare
300       essentials of compilation.  It no longer deals with linking/execution.
301       The interface has been cleaned up considerably.
302    
303       Utility routines for dealing with linking and execution have been
304       moved into their own substructures.
305    
306       (The ultimate goal of these changes is to provide a light-weight
307       binfile loader/linker (at least for, e.g., stable libraries) that
308       does not require CM or the compiler to be present.)
309    
310    CM documentation has been updated to reflect the changes to library
311    organization.
312    
313    ----------------------------------------------------------------------
314    Name: Matthias Blume
315    Date: 2001/07/10 17:30:00 EDT
316    Tag: Release_110_34
317    Description:
318    
319    Minor tweak to 110.34 (re-tagged):
320    
321      - README.html file added to CVS repository
322      - runtime compiles properly under FreeBSD 3.X and 4.X
323    
324    ----------------------------------------------------------------------
325    Name: Matthias Blume
326    Date: 2001/07/10 17:30:00 EDT
327    Tag: Release_110_34
328    Description:
329    
330    New version number (110.34). New bootfiles.
331    
332    ----------------------------------------------------------------------
333    Name: Matthias Blume
334    Date: 2001/07/09 16:00:00 EDT
335    Tag: blume-20010709-more-varargs
336    Description:
337    
338    I changed the handling of varargs in ml-nlffigen again:
339    The ellipsis ... will now simply be ignored (with an accompanying warning).
340    
341    The immediate effect is that you can actually call a varargs function
342    from ML -- but you can't actually supply any arguments beyond the ones
343    specified explicitly.  (For example, you can call printf with its format
344    string, but you cannot pass additional arguments.)
345    
346    This behavior is only marginally more useful than the one before, but
347    it has the advantage that a function or, more importantly, a function
348    type never gets dropped on the floor, thus avoiding follow-up problems with
349    other types that refer to the offending one.
350    
351    ----------------------------------------------------------------------
352    Name: Matthias Blume
353    Date: 2001/07/09 11:25:00 EDT
354    Tag: blume-20010709-varargs
355    Description:
356    
357    1. ckit-lib.cm now exports structure Error
358    2. ml-nlffigen reports occurences of "..." (i.e., varargs function types)
359       with a warning accompanied by a source location.  Moreover, it
360       merely skips the offending function or type and proceeds with the
361       rest of its work.u  As a result, one can safely feed C code containing
362       "..." to ml-nlffigen.
363    3. There are some internal improvements to CM, providing slightly
364       more general string substitutions in the tools subsystem.
365    
366    ----------------------------------------------------------------------
367    Name: Matthias Blume
368    Date: 2001/06/27 15:10:00 EDT
369    Tag: blume-20010627-concur
370    Description:
371    
372    Fixed a small bug in CM's handling of parallel compilation.
373    (You could observe the bug by Control-C-interrupting an ordinary
374    CMB.make or CM.stabilize and then attaching some compile servers.
375    The result was that all of a sudden the previously interrupted
376    compilation would continue on its own.  This was because of
377    an over-optimization: CM did not bother to clean out certain queues
378    when no servers were attached "anyway", resulting in the contents
379    of these queues to grab control when new servers did get attached.)
380    
381    There is also another minor update to the CM manual.
382    
383    ----------------------------------------------------------------------
384    Name: Matthias Blume
385    Date: 2001/06/26 16:15:00 EDT
386    Tag: blume-20010626-cmdoc
387    Description:
388    
389    Minor typo fixed in CM manual (syntax diagram for libraries).
390    
391    ----------------------------------------------------------------------
392    Name: Matthias Blume
393    Date: 2001/06/25 22:55:00 EDT
394    Tag: blume-20010625-x86pc
395    Description:
396    
397    Fixed a nasty bug in the X86 assembly code that caused signal
398    handlers to fail (crash) randomly.
399    
400    ----------------------------------------------------------------------
401    Name: Matthias Blume
402    Date: 2001/06/25 12:05:00 EDT
403    Tag: blume-20010625-nlffigen
404    Description:
405    
406    This update fixes a number of minor bugs in ml-nlffigen as reported by
407    Nick Carter <nbc@andrew.cmu.edu>.
408    
409      1. Silly but ok typedefs of the form "typedef void myvoid;" are now accepted.
410      2. Default names for generated files are now derived from the name of
411         the C file *without its directory*.  In particular, this causes generated
412         files to be placed locally even if the C file is in some system directory.
413      3. Default names for generated signatures and structures are also derived
414         from the C file name without its directory.  This avoids silly things
415         like "structure GL/GL".
416         (Other silly names are still possible because ml-nlffigen does not do
417          a thorough check of whether generated names are legal ML identifiers.
418          When in doubt, use command line arguments to force particular names.)
419    
420    ----------------------------------------------------------------------
421    Name: Matthias Blume
422    Date: 2001/06/21 12:25:00 EDT
423    Tag: blume-20010621-eXene
424    Description:
425    
426    eXene now compiles and (sort of) works again.
427    
428    The library name (for version > 110.33) is $/eXene.cm.
429    
430    I also added an new example in src/eXene/examples/nbody.  See the
431    README file there for details.
432    
433    ----------------------------------------------------------------------
434    Name: Matthias Blume
435    Date: 2001/06/20 16:40:00 EDT
436    Tag: blume-20010620-cml
437    Description:
438    
439    CML now compiles and works again.
440    
441    Libraries (for version > 110.33):
442    
443      $cml/cml.cm            Main CML library.
444      $cml/basis.cm          CML's version of $/basis.cm.
445      $cml/cml-internal.cm   Internal helper library.
446      $cml/core-cml.cm       Internal helper library.
447      $cml-lib/trace-cml.cm  Tracing facility.
448      $cml-lib/smlnj-lib.cm  CML's version of $/smlnj-lib.cm
449    
450    The installer (config/install.sh) has been taught how to properly
451    install this stuff.
452    
453    ----------------------------------------------------------------------
454    Name: Matthias Blume
455    Date: 2001/06/19 17:55:00 EDT
456    Tag: blume-20010619-instantiate
457    Description:
458    
459    This un-breaks the fix for bug 1432.
460    (The bug was originally fixed in 110.9 but I broke it again some
461    time after that.)
462    
463    ----------------------------------------------------------------------
464    Name: Matthias Blume
465    Date: 2001/06/19 17:25:00 EDT
466    Tag: blume-20010619-signals
467    Description:
468    
469    This should (hopefully) fix the long-standing signal handling bug.
470    (The runtime system was constructing a continuation record with an
471    incorrect descriptor which would cause the GC to drop data on the floor...)
472    
473    ----------------------------------------------------------------------
474    Name: Matthias Blume
475    Date: 2001/06/15 15:05:00 EDT
476    Tag: blume-20010615-moresparc
477    Description:
478    
479    Here is a short late-hour update related to Sparc c-calls:
480    
481     -- made handling of double-word arguments a bit smarter
482    
483     -- instruction selection phase tries to collapse certain clumsily
484        constructed ML-Trees; typical example:
485    
486            ADD(ty,ADD(_,e,LI d1),LI d2)  ->  ADD(ty,e,LI(d1+d2))
487    
488        This currently has no further impact on SML/NJ since mlriscGen does
489        not seem to generate such patterns in the first place, and c-calls
490        (which did generate them in the beginning) has meanwhile been fixed
491        so as to avoid them as well.
492    
493    ----------------------------------------------------------------------
494    Name: Matthias Blume
495    Date: 2001/06/15 15:05:00 EDT
496    Tag: blume-20010615-sparc
497    Description:
498    
499    The purpose of this update is to provide an implementation of NLFFI
500    on Sparc machines.
501    
502    Here are the changes in detail:
503    
504       * src/MLRISC/sparc/c-calls/sparc-c-calls.sml is a new file containing
505       the Sparc implementation of the c-calls API.
506       * The Sparc backend of SML/NJ has been modified to uniformely use %fp
507       for accessing the ML frame.  Thus, we have a real frame pointer and
508       can freely modify %sp without need for an omit-frame-ptr phase.
509       The vfp logic in src/compiler/CodeGen/* has been changed to accomodate
510       this case.
511       * ml-nlffigen has been taught to produce code for different architectures
512       and calling conventions.
513       * In a way similar to what was done in the x86 case, the Sparc
514       backend uses its own specific extension to mltree.  (For example,
515       it needs to be able to generate UNIMP instructions which are part
516       of the calling convention.)
517       * ml-nlffi-lib was reorganized to make it more modular (in particular,
518       to make it easier to plug in new machine- and os-dependent parts).
519    
520    There are some other fairly unrelated bug fixes and cleanups as well:
521    
522       * I further hacked the .cm files for MLRISC tools (like MDLGen) so
523       that they properly share their libraries with existing SML/NJ libraries.
524       * I fixed a minor cosmetic bug in CM, supressing certain spurious
525       follow-up error messages.
526       * Updates to CM/CMB documentation.
527    
528    TODO items:
529    
530       * MLRISC should use a different register as its asmTemp on the Sparc.
531         (The current %o2 is a really bad choice because it is part of the
532         calling conventions, so things might interfere in unexpected ways.)
533    
534    ----------------------------------------------------------------------
535    Name: Matthias Blume
536    Date: 2001/06/07
537    Tag: blume-20010607-calls
538    Description:
539    
540    A number of internal changes related to C calls and calling conventions:
541    
542    1. ML-Tree CALL statements now carry a "pops" field.  It indicates the
543       number of bytes popped implicitly (by the callee).  In most cases
544       this field is 0 but on x86/win32 it is some non-zero value.  This
545       is information provided for the benefit of the "omit-frameptr" pass.
546    2. The CALL instruction on the x86 carries a similar "pops" field.
547       The instruction selection phase copies its value from the ML-Tree
548       CALL statement.
549    3. On all other architectures, the instruction selection phase checks
550       whether "pops=0" and complains if not.
551    4. The c-calls implementation for x86 now accepts two calling conventions:
552       "ccall" and "stdcall".  When "ccall" is selected, the caller cleans
553       up after the call and pops is set to 0.  For "stdcall", the caller
554       does nothing, leaving the cleanup to the callee; pops is set to
555       the number of bytes that were pushed onto the stack.
556    5. The cproto decoder (compiler/Semant/types/cproto.sml) now can
557       distinguish between "ccall" and "stdcall".
558    6. The UNIMP instruction has been added to the supported Sparc instruction
559       set. (This is needed for implementing the official C calling convention
560       on this architecture.)
561    7. I fixed some of the .cm files under src/MLRISC/Tools to make them
562       work with the latest CM.
563    
564    ----------------------------------------------------------------------
565    Name: Matthias Blume
566    Date: 2001/06/05 15:10:00 EDT
567    Tag: blume-20010605-cm-index
568    Description:
569    
570    0. The "lambdasplit" parameter for class "sml" in CM has been documented.
571    
572    1. CM can now generate "index files".  These are human-readable files
573       that list on a per-.cm-file basis each toplevel symbol defined or
574       imported.  The location of the index file for
575       <p>/<d>.cm is <p>/CM/INDEX/<d>.cm.
576       To enable index-file generation, set CM.Control.generate_index to true
577       or export an environment-symbol: export CM_GENERATE_INDEX=true.
578    
579       The CM manual has been updated accordingly.
580    
581    2. I made some slight modifications to the c-calls API in MLRISC.
582    
583         a) There is now a callback to support saving/restoring of
584            dedicated but caller-save registers around the actual call
585            instruction.
586         b) One can optionally specify a comment-annotation for the
587            call instruction.
588    
589    3. SML/NJ (mlriscGen.sml) uses this new API for the rawccall primop.
590       (For example, the comment annotation shows the C prototype of
591        the function being called.)
592    
593    ----------------------------------------------------------------------
594    Name: Matthias Blume
595    Date: 2001/06/01 13:30:00 EDT
596    Tag: blume-20010601-nlffi-cleanup
597    Description:
598    
599    This is mostly a cleanup of MLFFI stuff:
600    
601       - some signature files have been put into a more exposed place
602       - the ugly 'f type parameter is gone (simplifies types tremendously!)
603       - ml-nlffigen changed accordingly
604       - tutorial updated
605    
606    Other changes:
607    
608       - author's affiliation in CM manual(s) updated
609       - some more recognized keywords added to Allen's sml.sty
610    
611    ----------------------------------------------------------------------
612    Name: Matthias Blume
613    Date: 2001/05/25 15:30:00 EDT
614    Tag: blume-20010525-iptr
615    Description:
616    
617      - put the official 110.33-README (as it appears on the ftp server) under
618        CVS
619      - fixed a small bug related to incomplete pointer types in
620        ml-nlffigen
621      - small cosmetic change to the ml-nlffi-lib's "arr" type constructor
622        (it does not need the 'f type parameter)
623    
624    ----------------------------------------------------------------------
625    Name: Matthias Blume
626    Date: 2001/05/23 14:30:00 EDT
627    Tag: Release_110_33
628    Description:
629    
630    New version number (110.33).  New bootfiles.
631    
632    ----------------------------------------------------------------------
633    Name: Matthias Blume
634    Date: 2001/05/22 18:06:00 EDT
635    Tag: blume-20010522-targets
636    Description:
637    
638    Made install.sh use file config/targets.customized if it exists, falling
639    back to config/targets if it doesn't.  This way one can have a customized
640    version of the targets file without touching the "real thing", thus
641    eliminating the constant fear of accidentally checking something bogus
642    back into the CVS repository...  (File config/targets.customized must
643    not be added to the repository!)
644    
645    ----------------------------------------------------------------------
646    Name: Matthias Blume
647    Date: 2001/05/22 16:30:00 EDT
648    Tag: blume-20010522-minitut
649    Description:
650    
651    1. Bug fix in ml-nlffigen; now (hopefully) correctly handling
652       struct returns.
653    2. Added src/ml-nlffi-lib/Doc/mini-tutorial.txt.  This is some very
654       incomplete, preliminary documentation for NLFFI.
655    
656    ----------------------------------------------------------------------
657    Name: Matthias Blume
658    Date: 2001/05/14 11:30:00 EDT
659    Tag: blume-20010514-script
660    Description:
661    
662    Some bugs in install script fixed.
663    
664    In addition to that I also made a slight change to the NLFFI API:
665    Functors generated by ml-nlffigen now take the dynamic library as a
666    straight functor argument, not as a suspended one.  (The original
667    functor code used to force the suspension right away anyway, so there
668    was nothing gained by this complication of the interface.)
669    
670    ----------------------------------------------------------------------
671    Name: Matthias Blume
672    Date: 2001/05/11 14:35:00 EDT
673    Tag: blume-20010511-ml-nlffi
674    Description:
675    
676    I finally took the plunge and added my new FFI code to the main
677    repository.  For x86-linux it is now ready for prime-time.
678    
679    There are two new subdirectories of "src":
680    
681      - ml-nlffi-lib:
682           The utility library for programs using the FFI interface.
683           Here is the implementation of $/c.cm and its associated low-level
684           partners $/c-int.cm and $/memory.cm.
685      - ml-nlffigen:
686           A stand-alone program for generating ML glue code from C source
687           code.
688    
689    Building ml-nlffigen requires $/ckit-lib.cm.
690    
691    The config/install.sh script has been updates to do the Right Thing
692    (hopefully).
693    
694    Notice that the source tree for the C-Kit will not be put under "src"
695    but directly under the installation root directory.  (This is the
696    structure that currently exists on the CVS server when you check out
697    module "sml".)  Fortunately, config/install.sh knows about this oddity.
698    
699    Bugs: No documentation yet.
700    
701    ----------------------------------------------------------------------
702    Name: Matthias Blume
703    Date: 2001/05/09 16:35:00 EDT
704    Tag: blume-20010509-cpscontract
705    Description:
706    
707    Fixed a bug in the accounting code in cpsopt/contract.sml.  (The
708    wrapper/unwrapper elimination did not decrement usage counts and some
709    dead variables got overlooked by the dead-up logic.)
710    
711    ----------------------------------------------------------------------
712    Name: Lal George
713    Date: 2001/05/08  17:26:09 EDT
714    Tag: george-20010508-omit-frameptr
715    Description:
716    
717    Changes to implement the omit-frame-pointer optimization to support
718    raw C calls. For now, there is only support on the Intel x86, but
719    other architectures will follow as more experience is gained with this.
720    
721    
722    ----------------------------------------------------------------------
723    Name: Matthias Blume
724    Date: 2001/05/07 14:40:00 EDT
725    Tag: blume-20010507-proxies
726    Description:
727    
728    I made into "proxy libraries" all libraries that qualify for such a
729    change.  (A qualifying library is a library that has another library or
730    groups as its sole member and repeats that member's export list
731    verbatim.  A proxy library avoids this repetition by omitting its export
732    list, effectively inheriting the list that its (only) member exports.
733    See the CM manual for more explanation.)
734    The main effect is that explicit export lists for these libraries
735    do not have to be kepts in sync, making maintenance a bit easier.
736    
737    I also added copyright notices to many .cm-files.
738    
739    Last but not least, I made a new set of bootfiles.
740    
741    ----------------------------------------------------------------------
742    Name: Matthias Blume
743    Date: 2001/05/04 17:00:00 EDT
744    Tag: blume-20010504-cm-lsplit
745    Description:
746    
747    0. John merged pending changes to $/smlnj-lib.cm
748    
749    1. Allen's previous change accidentally backed out of one of Lal's
750       earlier changes.  I undid this mistake (re-introducing Lal's change).
751    
752    2. I used the new topOrder' function from graph-scc.sml (from $/smlnj-lib.cm)
753       within the compiler where applicable.  There is some code simplification
754       because of that.
755    
756    3. The "split" phase (in FLINT) is now part of the default list of phases.
757       Compiler.Control.LambdaSplitting.* can be used to globally control the
758       lambda-splitting (cross-module-inlining) engine.  In addition to that,
759       it can now also be controlled on a per-source basis: CM has been taught
760       a new tool parameter applicable to ML source files.
761    
762       - To turn lambda-splitting off completely:
763            local open Compiler.Control.LambdaSplitting in
764                val _ = set Off
765            end
766       - To make "no lambda-splitting" the global default (but allow per-source
767         overriding); this is the initial setting:
768            local open Compiler.Control.LambdaSplitting in
769                val _ = set (Default NONE)
770            end
771       - To make "lambda-splitting with aggressiveness a" the global default
772         (and allow per-source overriding):
773            local open Compiler.Control.LambdaSplitting in
774                val _ = set (Default (SOME a))
775            end
776    
777       - To turn lambda-splitting off for a given ML souce file (say: a.sml)
778         write (in the respective .cm-file):
779            a.sml (lambdasplitting:off)
780       - To turn lambda-splitting for a.sml on with minimal aggressiveness:
781            a.sml (lambdasplitting:on)
782       - To turn lambda-splitting for a.sml on with aggressiveness <a> (where
783         <a> is a decimal non-negative integer):
784            a.sml (lambdasplitting:<a>)
785       - To turn lambda-splitting for a.sml on with maximal aggressiveness:
786            a.sml (lambdasplitting:infinity)
787       - To use the global default for a.sml:
788            a.sml (lambdasplitting:default)
789         or simply
790            a.sml
791    
792    ----------------------------------------------------------------------
793    Name: Allen Leung
794    Date: 2001/05/04 01:57:00 EDT
795    Tag: leunga-20010504-sync
796    Description:
797    
798      MLRISC features.
799    
800      1. Fix to CMPXCHG instructions.
801      2. Changed RA interface to allow annotations in callbacks.
802      3. Added a new method to the stream interface to allow annotations updates.
803    
804    ----------------------------------------------------------------------
805    Name: Matthias Blume
806    Date: 2001/05/01 11:45:00 EDT
807    Tag: blume-20010501-pcedittmp
808    Description:
809    
810    Changed install.sh to use the current working directory instead of
811    /usr/tmp for a temporary file (pcedittmp).  The previous choice
812    of /usr/tmp caused trouble with MacOS X because of file premission
813    problems.
814    
815    ----------------------------------------------------------------------
816    Name: Matthias Blume
817    Date: 2001/04/20 11:10:00 EDT
818    Tag: blume-20010420-inMLflag
819    Description:
820    
821     - added vp_limitPtrMask to vproc-state.h
822       (for use by the raw-C-calls mechanism to implement proper interrupt
823        handling)
824     - made the ML compiler aware of various data-structure offsets so it
825       can generate code for accessing the vp_inML flag and vp_limitPtrMask
826     - tweaked mlriscGen.sml to have it emit interrupt-handling code for
827       raw C-calls
828    
829    ----------------------------------------------------------------------
830    Name: Lal George
831    Date: 2001/04/20 09:15:28 EDT
832    Tag: george-20010420-macosX
833    Description:
834    
835     - Changes to port to Mac OS X; Darwin.
836    
837     - In the process I found that sqrt was broken on the PPC, because the
838       fsqrt instruction is not implemented.
839    
840    ----------------------------------------------------------------------
841    Name: Matthias Blume
842    Date: 2001/04/18 12:45:00 EDT
843    Tag: blume-20010418-ccalls
844    Description:
845    
846     - fixed two off-by-4 errors in the x86-specific c-calls implementation
847       (this bug prevented structure arguments containing pointers from being
848        passed correctly)
849     - changed the raw-C-call code in mlriscGen.sml in such a way that
850       structure arguments are represented as a pointer to the beginning
851       of the structure (instead of having a series of synthesized arguments,
852       one for each structure member)
853    
854     - made makeml script's verbosity level configurable via environment
855       variable (MAKEML_VERBOSITY)
856    
857     - eliminated placeholder implementations for f32l, w16s, i16s, and f32s
858       in rawmem-x86.sml; we are now using the real thing
859    
860    ----------------------------------------------------------------------
861    Name: Matthias Blume
862    Date: 2001/03/22 16:25:00 EST
863    Tag: blume-20010322-bootfiles
864    Description:
865    
866    Created a new set of bootfiles (for your automatic installation convenience).
867    
868    ----------------------------------------------------------------------
869    Name: Matthias Blume
870    Date: 2001/03/22 15:10:00 EST
871    Tag: blume-20010322-rawmem-parcm
872    Description:
873    
874    1. All "raw memory access" primitives for the new FFI are implemented now
875       (at least on the x86).
876    2. Some further cleanup of CM's parallel make mechanism.
877    
878    ----------------------------------------------------------------------
879    Name: Matthias Blume
880    Date: 2001/03/19 17:53:00 EST
881    Tag: blume-20010319-parallel
882    Description:
883    
884    Parallel make (using compile servers) now works again.
885    
886    To this end, CM.stabilize and CMB.make have been modified to work in
887    two passes when compile servers are attached:
888       1. Compile everything, do not perform stabilization; this pass
889          uses compile servers
890       2. Stabilize everything; this pass does not use compile servers
891    If there are no compile servers, the two passes are combined into one
892    (as before).  Splitting the passes increases the inherent parallelism
893    in the dependency graph because the entire graph including all
894    libraries is available at the same time.  This, in turn, improves
895    server utilization.  The downside is that the master process will
896    have to do some extra work after compilation is done (because for
897    technical reasons it must re-read all the binfiles during stabilization).
898    
899    ----------------------------------------------------------------------
900    Name: Matthias Blume
901    Date: 2001/03/16 12:22:00 EST
902    Tag: blume-20010316-bootfiles
903    Description:
904    
905    Created a new set of bootfiles (for your automatic installation convenience).
906    
907    ----------------------------------------------------------------------
908    Name: Matthias Blume
909    Date: 2001/03/16 11:00:00 EST
910    Tag: blume-20010316-MLTREE-fixup
911    Description:
912    
913    This is a minor fixup for an (untagged) earlier commit by Allen.
914    (A file was missing).
915    
916    ----------------------------------------------------------------------
917    Name: Allen Leung
918    Date: Mon Mar  5 18:54:57 EST 2001
919    Tag: leunga-20010305-cut-support
920    
921    1. New support for alternative control-flow in MLTREE.
922       Currently we support
923    
924          FLOW_TO(CALL ...., [k1,...,kn])
925    
926       This is needed for 'cuts to' in C-- and try/handle-like constructs
927       in Moby
928    
929       New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
930    
931    2. Register Allocator
932    
933       Changes in interface [from Fermin, John]
934    
935    3. Alpha 8-bit SLL support [Fermin]
936    
937    4. All architectures
938    
939       A new module (ClusterExpandCopies) for expanding parallel copies.
940    
941    ----------------------------------------------------------------------
942    Name: Allen Leung
943    Date: 2001/02/27 23:07:00 EST
944    Tag: leunga-20010227-minor-stuff
945    
946    1. Alpha bug fix for CMOVNE
947    2. Handle mltree COND(..,FCMP ...,...)
948    3. Bug fix in simplifier
949    
950    ----------------------------------------------------------------------
951    Name: Matthias Blume
952    Date: 2001/01/30 17:50:00 EST
953    Tag: blume-20010130-sync
954  Description:  Description:
955    
956    This is just a minor update to sync my devel branch with the main brach.
957    The only visible change is the addition of some README files.
958    
959    ----------------------------------------------------------------------
960    Name: Matthias Blume
961    Date: 2001/01/12 23:30:00 JST
962    Tag: blume-20010112-bootfiles
963    Description:
964    
965    Made a new set of bootfiles that goes with the current state of the
966    repository.
967    
968    ----------------------------------------------------------------------
969    Name: Matthias Blume
970    Date: 2001/01/12 21:20:00 JST
971    Tag: blume-20010112-sync
972    Description:
973    
974    I am just flushing out some minor changes that had accumulated in
975    my private branch in order to sync with the main tree.  (This is
976    mainly because I had CVS trouble when trying to merge _into_ my
977    private branch.)
978    
979    Most people should be completely unaffected by this.
980    
981    ----------------------------------------------------------------------
982    Name: Allen Leung
983    Date: Thu Jan 11 21:03:00 EST 2001
984    Tag: leunga-20010111-labexp=mltree
985    Description:
986    
987    1.  Removed the type LabelExp and replace it by MLTree.
988    2.  Rewritten mltree-simplify with the pattern matcher tool.
989    3.  There were some bugs in alpha code generator which would break
990        64-bit code generation.
991    4.  Redo the tools to generate code with the
992    5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
993        are now generated by perl scripts.
994    
995    ----------------------------------------------------------------------
996    Name: Matthias Blume
997    Date: 2001/01/10 21:55:00 JST
998    Tag: blume-20010110-rcc
999    Description:
1000    
1001    The RCC stuff now seems to work (but only on the x86).
1002    This required hacking of the c-calls interface (and -implementation) in
1003    MLRISC.
1004    
1005    Normal compiler users should be unaffected.
1006    
1007    ----------------------------------------------------------------------
1008    Name: Matthias Blume
1009    Date: 2001/01/09 01:20:00 JST
1010    Tag: blume-20010109-rcc
1011    Description:
1012    
1013    This is a fairly big patch, flushing out a large number of pending
1014    changes that I made to my development copy over the last couple of days.
1015    
1016    Of practical relevance at this moment is a workaround for a pickling
1017    bug that Allen ran into the other day.  The cause of the bug itself is
1018    still unknown and it might be hard to fix it properly, but the
1019    workaround has some merits of its own (namely somewhat reducing pickling
1020    overhead for certain libraries).  Therefore, I think this solution should
1021    be satisfactory at this time.
1022    
1023    The rest of the changes (i.e., the vast majority) has to do with my
1024    ongoing efforts of providing direct support for C function calls from
1025    ML.  At the moment there is a new primop "RAW_CCALL", typing magic
1026    in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
1027    case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
1028    translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
1029    handle RCC.
1030    
1031    The last part (the changes to mlriscGen.sml) are still known to be
1032    wrong on the x86 and not implemented on all other architectures.  But
1033    the infrastructure is in place. I had to change a few functor
1034    signatures in the backend to be able to route the CCalls interface
1035    from MLRISC there, and I had to specialize the mltree type (on the
1036    x86) to include the necessary extensions. (The extensions themselves
1037    were already there and redy to go in MLRISC/x86).
1038    
1039    Everything should be very happy as soon as someone helps me with
1040    mlriscGen.sml...
1041    
1042    In any case, nothing of this should matter to anyone as long as the
1043    new primop is not being used (which is going to be the case unless you
1044    find it where I hid it :). The rest of the compiler is completely
1045    unaffected.
1046    
1047    ----------------------------------------------------------------------
1048    Name: Matthias Blume
1049    Date: 2001/01/05 00:30:00 JST
1050    Tag: blume-20010105-primops
1051    Description:
1052    
1053    Added some experimental support for work that I am doing right now.
1054    These changes mostly concern added primops, but there is also a new
1055    experimental C library in the runtime system (but currently not enabled
1056    anywhere except on Linux/X86).
1057    
1058    In the course of adding primops (and playing with them), I discovered that
1059    Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
1060    broken.  (Zhong was very right he labeled this stuff as "major gross hack".)
1061    To recover, I made type information in INL_PRIM mandatory and changed
1062    prim.sml as well as built-in.sml accordingly.  The InLine structure now
1063    has complete, correct type information (i.e., no bottom types).
1064    
1065    Since all these changes mean that we need new binfiles, I also bumped the
1066    version number to 110.32.1.
1067    
1068    ----------------------------------------------------------------------
1069    Name: Matthias Blume
1070    Date: 2000/12/30 22:10:00 JST
1071    Tag: blume-20001230-various
1072    Description:
1073    
1074    Added proxy libraries for MLRISC and let MLRISC libraries refer
1075    to each other using path anchors.  (See CM manual for explanation.)
1076    
1077    Updated CM documentation.
1078    
1079    Fixed some bugs in CM.
1080    
1081    Implemented "proxy" libraries (= syntactic sugar for CM).
1082    
1083    Added "-quiet" option to makeml and changed runtime system accordingly.
1084    
1085    Added cleanup handler for exportML to reset timers and compiler stats.
1086    
1087    ----------------------------------------------------------------------
1088    Name: Lal George
1089    Date: 2000/12/22 22:22:58 EST 2000
1090    Tag: Release_110_32
1091    Description:
1092    
1093            Infinite precision used throughout MLRISC.
1094            see MLRISC/mltree/machine-int.sig
1095    
1096  ----------------------------------------------------------------------  ----------------------------------------------------------------------
1097  Name: Matthias Blume  Name: Matthias Blume
1098  Date: 2000/12/22 23:16:00 JST  Date: 2000/12/22 23:16:00 JST
# Line 27  Line 1104 
1104    
1105  ----------------------------------------------------------------------  ----------------------------------------------------------------------
1106  Name: Matthias Blume  Name: Matthias Blume
1107    Date: 2000/12/22 21:20:00 JST
1108    Tag: blume-20001222-anchorenv
1109    Description:
1110    
1111    Fixed CM's handling of anchor environments in connection with CMB.make.
1112    
1113    ----------------------------------------------------------------------
1114    Name: Matthias Blume
1115  Date: 2000/12/22 13:15:00 JST  Date: 2000/12/22 13:15:00 JST
1116  Tag: blume-20001222-cleanup  Tag: blume-20001222-cleanup
1117  Description:  Description:

Legend:
Removed from v.760  
changed lines
  Added in v.903

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