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 773, Mon Jan 8 16:18:37 2001 UTC revision 904, Mon Aug 20 19:50:05 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: 2001/08/20 15:50 EDT
18    Tag: blume20010820-slipup
19    Description:
20    
21    I forgot to commit a few files.  Here they are...
22    
23    ----------------------------------------------------------------------
24    Name: Matthias Blume
25    Date: 2001/08/20 15:35:00 EDT
26    Tag: blume-20010820-debugprof
27    Description:
28    
29    !!!! NEW BOOTFILES !!!!
30    
31    This is another round of reorganizing the compiler sources.  This
32    time the main goal was to factor out all the "instrumentation"
33    passes (for profiling and backtracing) into their own library.
34    The difficulty was to do it in such a way that it does not depend
35    on elaborate.cm but only on elabdata.cm.
36    
37    Therefore there have been further changes to both elaborate.cm and
38    elabdata.cm -- more "generic" things have been moved from the former
39    to the latter.  As a result, I was forced to split the assignment
40    of numbers indicating "primtyc"s into two portions: SML-generic and
41    SML/NJ-specific.  Since it would have been awkward to maintain,
42    I bit the bullet and actually _changed_ the mapping between these
43    numbers and primtycs.  The bottom line of this is that you need
44    a new set of bin- and bootfiles.
45    
46    I have built new bootfiles for all architectures, so doing a fresh
47    checkout and config/install.sh should be all you need.
48    
49    The newly created library's name is
50    
51        $smlnj/viscomp/debugprof.cm
52    
53    and its sources live under
54    
55        src/compiler/DebugProf
56    
57    ----------------------------------------------------------------------
58    Name: Matthias Blume
59    Date: 2001/08/15 17:15:00 EDT
60    Tag: blume-20010815-compreorg
61    Description:
62    
63    This is a first cut at reorganizing the CM libraries that make up the
64    core of the compiler.  The idea is to separate out pieces that could
65    be used independently by tools, e.g., the parser, the typechecker, etc.
66    
67    The current status is a step in this direction, but it is not quite
68    satisfactory yet.  Expect more changes in the future.
69    
70    Here is the current (new) organization...
71    
72        What used to be $smlnj/viscomp/core.cm is now divided into
73        six CM libraries:
74    
75             $smlnj/viscomp/basics.cm
76                           /parser.cm
77                           /elabdata.cm
78                           /elaborate.cm
79                           /execute.cm
80                           /core.cm
81    
82        The CM files for these libraries live under src/system/smlnj/viscomp.
83        All these libraries are proxy libraries that contain precisely
84        one CM library component.  Here are the locations of the components
85        (all within the src/compiler tree):
86    
87             Basics/basics.cm
88             Parse/parser.cm
89             ElabData/elabdata.cm
90             Elaborator/elaborate.cm
91             Execution/execute.cm
92             core.cm
93    
94         [This organization is the same that has been used already
95         for a while for the architecture-specific parts of the visible
96         compiler and for the old version of core.cm.]
97    
98         As you will notice, many source files have been moved from their
99         respective original locations to a new home in one of the above
100         subtrees.
101    
102         The division of labor between the new libraries is the following:
103    
104             basics.cm:
105                - Simple, basic definitions that pertain to many (or all) of
106                  the other libraries.
107             parser.cm:
108                - The SML parser, producing output of type Ast.dec.
109                - The type family for Ast is also defined and exported here.
110             elabdata.cm:
111                - The datatypes that describe input and output of the elaborator.
112                  This includes types, absyn, and static environments.
113             elaborator.cm:
114                - The SML/NJ type checker and elaborator.
115                  This maps an Ast.dec (with a given static environment) to
116                  an Absyn.dec (with a new static environment).
117                - This libraries implements certain modules that used to be
118                  structures as functors (to remove dependencies on FLINT).
119             execute.cm:
120                - Everything having to do with executing binary code objects.
121                - Dynamic environments.
122             core.cm:
123                - SML/NJ-specific instantiations of the elaborator and MLRISC.
124                - Top-level modules.
125                - FLINT (this should eventually become its own library)
126    
127    Notes:
128    
129    I am not 100% happy with the way I separated the elaborator (and its
130    data structures) from FLINT.  Two instances of the same problem:
131    
132        1. Data structures contain certain fields that carry FLINT-specific
133           information.  I hacked around this using exn and the property list
134           module from smlnj-lib.  But the fact that there are middle-end
135           specific fields around at all is a bit annoying.
136    
137        2. The elaborator calculates certain FLINT-related information.  I tried
138           to make this as abstract as I could using functorization, but, again,
139           the fact that the elaborator has to perform calculations on behalf
140           of the middle-end at all is not nice.
141    
142        3. Having to used exn and property lists is unfortunate because it
143           weakens type checking.  The other alternative (parameterizing
144           nearly *everything*) is not appealing, though.
145    
146    I removed the "rebinding =" warning hack because due to the new organization
147    it was awkward to maintain it.  As a result, the compiler now issues some of
148    these warnings when compiling init.cmi during bootstrap compilation. On
149    the plus side, you also get a warning when you do, for example:
150       val op = = Int32.+
151    which was not the case up to now.
152    
153    I placed "assign" and "deref" into the _Core structure so that the
154    code that deals with the "lazy" keyword can find them there.  This
155    removes the need for having access to the primitive environment
156    during elaboration.
157    
158    ----------------------------------------------------------------------
159    Name: Matthias Blume
160    Date: 2001/08/13
161    Tag: blume-20010813-closures
162    Description:
163    
164    This fix was sent to us by Zhong Shao.  It is supposed to improve the
165    performance of certain loops by avoiding needless closure allocation.
166    
167    ----------------------------------------------------------------------
168    Name: Lal George
169    Date: 2001/07/31 10:03:23 EDT 2001
170    Tag: george-20010731-x86-fmalloc
171    Description: Fixed bug in x86 calls
172    
173        There was a bug where call instructions would mysteriously
174        vanish. The call instruction had to be one that returned
175        a floating point value.
176    
177    ----------------------------------------------------------------------
178    Name: Lal George
179    Date: 2001/07/19 16:36:29 EDT 2001
180    Tag: george-20010719-simple-cells
181    Description:
182    
183    I have dramatically simplified the interface for CELLS in MLRISC.
184    
185    In summary, the cells interface is broken up into three parts:
186    
187      1. CellsBasis : CELLS_BASIS
188    
189            CellsBasis is a top level structure and common for all
190            architectures.  it contains the definitions of basic datatypes
191            and utility  functions over these types.
192    
193      2. functor Cells() : CELLS
194    
195            Cells generates an interface for CELLS that incorporates the
196            specific resources on the target architecture, such as the
197            presence of special register classes, their number and size,
198            and various useful substructures.
199    
200      3. <ARCH>CELLS
201    
202            e.g. SparcCells: SPARCCELLS
203    
204            <ARCH>CELLS usually contains additional bindings for special
205            registers  on the architecture, such as:
206    
207                    val r0 : cell           (* register zero *)
208                    val y : cell            (* Y register *)
209                    val psr : cell          (* processor status register *)
210                    ...
211    
212            The structure returned by applying the Cells functor is opened
213            in this interface.
214    
215    The main implication of all this is that the datatypes for cells is
216    split between CellsBasis and CELLS -- a fairly simple change for user
217    code.
218    
219    In the old scheme the CELLS interface had a definitional binding of
220    the form:
221    
222            signature CELLS = sig
223    
224               structure CellsBasis = CellsBasis
225    
226               ...
227    
228            end
229    
230    With all the sharing constraints that goes on in MLRISC, this old
231    design  quickly leads to errors such as:
232    
233            "structure definition spec inside of sharing ... "
234    
235    
236    and appears to require an unacceptable amount of sharing and where
237    constraint hackery.
238    
239    I think this error message (the interaction of definitional specs and
240    sharing) requires more explanation on our web page.
241    
242    ----------------------------------------------------------------------
243    Name: Matthias Blume
244    Date: 2001/07/19 15:00:00 EDT
245    Tag: blume-20010719-libreorg
246    Description:
247    
248    This update puts together a fairly extensive but straightforward change
249    to the way the libraries that implement the interactive system are
250    organized:
251    
252       The biggest change is the elimination of structure Compiler.  As a
253       replacement for this structure, there is now a CM library
254       (known as $smlnj/compiler.cm or $smlnj/compiler/current.cm)
255       that exports all the substructures of the original structure Compiler
256       directly.  So instead of saying Compiler.Foo.bar one now simply
257       says Foo.bar.  (The CM libraries actually export a collection of
258       structures that is richer than the collection of substructures of
259       structure Compiler.)
260    
261       To make the transition smooth, there is a separate library called
262       $smlnj/compiler/compiler.cm which puts together and exports the
263       original structure Compiler (or at least something very close to it).
264    
265       There are five members of the original structure Compiler
266       that are not exported directly but which instead became members
267       of a new structure Backend (described by signature BACKEND).  These are:
268       structure Profile (: PROFILE), structure Compile (: COMPILE), structure
269       Interact (: INTERACT), structure Machine (: MACHINE), and val
270       architecture (: string).
271    
272       Structure Compiler.Version has become structure CompilerVersion.
273    
274       Cross-compilers for alpha32, hppa, ppc, sparc, and x86 are provided
275       by $smlnj/compiler/<arch>.cm where <arch> is alpha32, hppa, ppc, sparc,
276       or x86, respectively.
277       Each of these exports the same frontend structures that
278       $smlnj/compiler.cm exports.  But they do not have a structure Backend
279       and instead export some structure <Arch>Backend where <Arch> is Alpha32,
280       Hppa, PPC, Sparc, or X86, respectively.
281    
282       Library $smlnj/compiler/all.cm exports the union of the exports of
283       $smlnj/compiler/<arch>.cm
284    
285       There are no structures <Arch>Compiler anymore, use
286       $smlnj/compiler/<arch>.cm instead.
287    
288       Library host-compiler-0.cm is gone.  Instead, the internal library
289       that instantiates CM is now called cm0.cm.  Selection of the host
290       compiler (backend) is no longer done here but. (Responsibility for it
291       now lies with $smlnj/compiler/current.cm.  This seems to be more
292       logical.)
293    
294       Many individual files have been moved or renamed.  Some files have
295       been split into multiple files, and some "dead" files have been deleted.
296    
297    Aside from these changes to library organization, there are also changes
298    to the way the code itself is organized:
299    
300       Structure Binfile has been re-implemented in such a way that it no
301       longer needs any knowledge of the compiler.  It exclusively deals
302       with the details of binfile layout.  It no longer invokes the
303       compiler (for the purpose of creating new prospective binfile
304       content), and it no longer has any knowledge of how to interpret
305       pickles.
306    
307       Structure Compile (: COMPILE) has been stripped down to the bare
308       essentials of compilation.  It no longer deals with linking/execution.
309       The interface has been cleaned up considerably.
310    
311       Utility routines for dealing with linking and execution have been
312       moved into their own substructures.
313    
314       (The ultimate goal of these changes is to provide a light-weight
315       binfile loader/linker (at least for, e.g., stable libraries) that
316       does not require CM or the compiler to be present.)
317    
318    CM documentation has been updated to reflect the changes to library
319    organization.
320    
321    ----------------------------------------------------------------------
322    Name: Matthias Blume
323    Date: 2001/07/10 17:30:00 EDT
324    Tag: Release_110_34
325    Description:
326    
327    Minor tweak to 110.34 (re-tagged):
328    
329      - README.html file added to CVS repository
330      - runtime compiles properly under FreeBSD 3.X and 4.X
331    
332    ----------------------------------------------------------------------
333    Name: Matthias Blume
334    Date: 2001/07/10 17:30:00 EDT
335    Tag: Release_110_34
336    Description:
337    
338    New version number (110.34). New bootfiles.
339    
340    ----------------------------------------------------------------------
341    Name: Matthias Blume
342    Date: 2001/07/09 16:00:00 EDT
343    Tag: blume-20010709-more-varargs
344    Description:
345    
346    I changed the handling of varargs in ml-nlffigen again:
347    The ellipsis ... will now simply be ignored (with an accompanying warning).
348    
349    The immediate effect is that you can actually call a varargs function
350    from ML -- but you can't actually supply any arguments beyond the ones
351    specified explicitly.  (For example, you can call printf with its format
352    string, but you cannot pass additional arguments.)
353    
354    This behavior is only marginally more useful than the one before, but
355    it has the advantage that a function or, more importantly, a function
356    type never gets dropped on the floor, thus avoiding follow-up problems with
357    other types that refer to the offending one.
358    
359    ----------------------------------------------------------------------
360    Name: Matthias Blume
361    Date: 2001/07/09 11:25:00 EDT
362    Tag: blume-20010709-varargs
363    Description:
364    
365    1. ckit-lib.cm now exports structure Error
366    2. ml-nlffigen reports occurences of "..." (i.e., varargs function types)
367       with a warning accompanied by a source location.  Moreover, it
368       merely skips the offending function or type and proceeds with the
369       rest of its work.u  As a result, one can safely feed C code containing
370       "..." to ml-nlffigen.
371    3. There are some internal improvements to CM, providing slightly
372       more general string substitutions in the tools subsystem.
373    
374    ----------------------------------------------------------------------
375    Name: Matthias Blume
376    Date: 2001/06/27 15:10:00 EDT
377    Tag: blume-20010627-concur
378    Description:
379    
380    Fixed a small bug in CM's handling of parallel compilation.
381    (You could observe the bug by Control-C-interrupting an ordinary
382    CMB.make or CM.stabilize and then attaching some compile servers.
383    The result was that all of a sudden the previously interrupted
384    compilation would continue on its own.  This was because of
385    an over-optimization: CM did not bother to clean out certain queues
386    when no servers were attached "anyway", resulting in the contents
387    of these queues to grab control when new servers did get attached.)
388    
389    There is also another minor update to the CM manual.
390    
391    ----------------------------------------------------------------------
392    Name: Matthias Blume
393    Date: 2001/06/26 16:15:00 EDT
394    Tag: blume-20010626-cmdoc
395    Description:
396    
397    Minor typo fixed in CM manual (syntax diagram for libraries).
398    
399    ----------------------------------------------------------------------
400    Name: Matthias Blume
401    Date: 2001/06/25 22:55:00 EDT
402    Tag: blume-20010625-x86pc
403    Description:
404    
405    Fixed a nasty bug in the X86 assembly code that caused signal
406    handlers to fail (crash) randomly.
407    
408    ----------------------------------------------------------------------
409    Name: Matthias Blume
410    Date: 2001/06/25 12:05:00 EDT
411    Tag: blume-20010625-nlffigen
412    Description:
413    
414    This update fixes a number of minor bugs in ml-nlffigen as reported by
415    Nick Carter <nbc@andrew.cmu.edu>.
416    
417      1. Silly but ok typedefs of the form "typedef void myvoid;" are now accepted.
418      2. Default names for generated files are now derived from the name of
419         the C file *without its directory*.  In particular, this causes generated
420         files to be placed locally even if the C file is in some system directory.
421      3. Default names for generated signatures and structures are also derived
422         from the C file name without its directory.  This avoids silly things
423         like "structure GL/GL".
424         (Other silly names are still possible because ml-nlffigen does not do
425          a thorough check of whether generated names are legal ML identifiers.
426          When in doubt, use command line arguments to force particular names.)
427    
428    ----------------------------------------------------------------------
429    Name: Matthias Blume
430    Date: 2001/06/21 12:25:00 EDT
431    Tag: blume-20010621-eXene
432    Description:
433    
434    eXene now compiles and (sort of) works again.
435    
436    The library name (for version > 110.33) is $/eXene.cm.
437    
438    I also added an new example in src/eXene/examples/nbody.  See the
439    README file there for details.
440    
441    ----------------------------------------------------------------------
442    Name: Matthias Blume
443    Date: 2001/06/20 16:40:00 EDT
444    Tag: blume-20010620-cml
445    Description:
446    
447    CML now compiles and works again.
448    
449    Libraries (for version > 110.33):
450    
451      $cml/cml.cm            Main CML library.
452      $cml/basis.cm          CML's version of $/basis.cm.
453      $cml/cml-internal.cm   Internal helper library.
454      $cml/core-cml.cm       Internal helper library.
455      $cml-lib/trace-cml.cm  Tracing facility.
456      $cml-lib/smlnj-lib.cm  CML's version of $/smlnj-lib.cm
457    
458    The installer (config/install.sh) has been taught how to properly
459    install this stuff.
460    
461    ----------------------------------------------------------------------
462    Name: Matthias Blume
463    Date: 2001/06/19 17:55:00 EDT
464    Tag: blume-20010619-instantiate
465    Description:
466    
467    This un-breaks the fix for bug 1432.
468    (The bug was originally fixed in 110.9 but I broke it again some
469    time after that.)
470    
471    ----------------------------------------------------------------------
472    Name: Matthias Blume
473    Date: 2001/06/19 17:25:00 EDT
474    Tag: blume-20010619-signals
475    Description:
476    
477    This should (hopefully) fix the long-standing signal handling bug.
478    (The runtime system was constructing a continuation record with an
479    incorrect descriptor which would cause the GC to drop data on the floor...)
480    
481    ----------------------------------------------------------------------
482    Name: Matthias Blume
483    Date: 2001/06/15 15:05:00 EDT
484    Tag: blume-20010615-moresparc
485    Description:
486    
487    Here is a short late-hour update related to Sparc c-calls:
488    
489     -- made handling of double-word arguments a bit smarter
490    
491     -- instruction selection phase tries to collapse certain clumsily
492        constructed ML-Trees; typical example:
493    
494            ADD(ty,ADD(_,e,LI d1),LI d2)  ->  ADD(ty,e,LI(d1+d2))
495    
496        This currently has no further impact on SML/NJ since mlriscGen does
497        not seem to generate such patterns in the first place, and c-calls
498        (which did generate them in the beginning) has meanwhile been fixed
499        so as to avoid them as well.
500    
501    ----------------------------------------------------------------------
502    Name: Matthias Blume
503    Date: 2001/06/15 15:05:00 EDT
504    Tag: blume-20010615-sparc
505    Description:
506    
507    The purpose of this update is to provide an implementation of NLFFI
508    on Sparc machines.
509    
510    Here are the changes in detail:
511    
512       * src/MLRISC/sparc/c-calls/sparc-c-calls.sml is a new file containing
513       the Sparc implementation of the c-calls API.
514       * The Sparc backend of SML/NJ has been modified to uniformely use %fp
515       for accessing the ML frame.  Thus, we have a real frame pointer and
516       can freely modify %sp without need for an omit-frame-ptr phase.
517       The vfp logic in src/compiler/CodeGen/* has been changed to accomodate
518       this case.
519       * ml-nlffigen has been taught to produce code for different architectures
520       and calling conventions.
521       * In a way similar to what was done in the x86 case, the Sparc
522       backend uses its own specific extension to mltree.  (For example,
523       it needs to be able to generate UNIMP instructions which are part
524       of the calling convention.)
525       * ml-nlffi-lib was reorganized to make it more modular (in particular,
526       to make it easier to plug in new machine- and os-dependent parts).
527    
528    There are some other fairly unrelated bug fixes and cleanups as well:
529    
530       * I further hacked the .cm files for MLRISC tools (like MDLGen) so
531       that they properly share their libraries with existing SML/NJ libraries.
532       * I fixed a minor cosmetic bug in CM, supressing certain spurious
533       follow-up error messages.
534       * Updates to CM/CMB documentation.
535    
536    TODO items:
537    
538       * MLRISC should use a different register as its asmTemp on the Sparc.
539         (The current %o2 is a really bad choice because it is part of the
540         calling conventions, so things might interfere in unexpected ways.)
541    
542    ----------------------------------------------------------------------
543    Name: Matthias Blume
544    Date: 2001/06/07
545    Tag: blume-20010607-calls
546    Description:
547    
548    A number of internal changes related to C calls and calling conventions:
549    
550    1. ML-Tree CALL statements now carry a "pops" field.  It indicates the
551       number of bytes popped implicitly (by the callee).  In most cases
552       this field is 0 but on x86/win32 it is some non-zero value.  This
553       is information provided for the benefit of the "omit-frameptr" pass.
554    2. The CALL instruction on the x86 carries a similar "pops" field.
555       The instruction selection phase copies its value from the ML-Tree
556       CALL statement.
557    3. On all other architectures, the instruction selection phase checks
558       whether "pops=0" and complains if not.
559    4. The c-calls implementation for x86 now accepts two calling conventions:
560       "ccall" and "stdcall".  When "ccall" is selected, the caller cleans
561       up after the call and pops is set to 0.  For "stdcall", the caller
562       does nothing, leaving the cleanup to the callee; pops is set to
563       the number of bytes that were pushed onto the stack.
564    5. The cproto decoder (compiler/Semant/types/cproto.sml) now can
565       distinguish between "ccall" and "stdcall".
566    6. The UNIMP instruction has been added to the supported Sparc instruction
567       set. (This is needed for implementing the official C calling convention
568       on this architecture.)
569    7. I fixed some of the .cm files under src/MLRISC/Tools to make them
570       work with the latest CM.
571    
572    ----------------------------------------------------------------------
573    Name: Matthias Blume
574    Date: 2001/06/05 15:10:00 EDT
575    Tag: blume-20010605-cm-index
576    Description:
577    
578    0. The "lambdasplit" parameter for class "sml" in CM has been documented.
579    
580    1. CM can now generate "index files".  These are human-readable files
581       that list on a per-.cm-file basis each toplevel symbol defined or
582       imported.  The location of the index file for
583       <p>/<d>.cm is <p>/CM/INDEX/<d>.cm.
584       To enable index-file generation, set CM.Control.generate_index to true
585       or export an environment-symbol: export CM_GENERATE_INDEX=true.
586    
587       The CM manual has been updated accordingly.
588    
589    2. I made some slight modifications to the c-calls API in MLRISC.
590    
591         a) There is now a callback to support saving/restoring of
592            dedicated but caller-save registers around the actual call
593            instruction.
594         b) One can optionally specify a comment-annotation for the
595            call instruction.
596    
597    3. SML/NJ (mlriscGen.sml) uses this new API for the rawccall primop.
598       (For example, the comment annotation shows the C prototype of
599        the function being called.)
600    
601    ----------------------------------------------------------------------
602    Name: Matthias Blume
603    Date: 2001/06/01 13:30:00 EDT
604    Tag: blume-20010601-nlffi-cleanup
605    Description:
606    
607    This is mostly a cleanup of MLFFI stuff:
608    
609       - some signature files have been put into a more exposed place
610       - the ugly 'f type parameter is gone (simplifies types tremendously!)
611       - ml-nlffigen changed accordingly
612       - tutorial updated
613    
614    Other changes:
615    
616       - author's affiliation in CM manual(s) updated
617       - some more recognized keywords added to Allen's sml.sty
618    
619    ----------------------------------------------------------------------
620    Name: Matthias Blume
621    Date: 2001/05/25 15:30:00 EDT
622    Tag: blume-20010525-iptr
623    Description:
624    
625      - put the official 110.33-README (as it appears on the ftp server) under
626        CVS
627      - fixed a small bug related to incomplete pointer types in
628        ml-nlffigen
629      - small cosmetic change to the ml-nlffi-lib's "arr" type constructor
630        (it does not need the 'f type parameter)
631    
632    ----------------------------------------------------------------------
633    Name: Matthias Blume
634    Date: 2001/05/23 14:30:00 EDT
635    Tag: Release_110_33
636    Description:
637    
638    New version number (110.33).  New bootfiles.
639    
640    ----------------------------------------------------------------------
641    Name: Matthias Blume
642    Date: 2001/05/22 18:06:00 EDT
643    Tag: blume-20010522-targets
644    Description:
645    
646    Made install.sh use file config/targets.customized if it exists, falling
647    back to config/targets if it doesn't.  This way one can have a customized
648    version of the targets file without touching the "real thing", thus
649    eliminating the constant fear of accidentally checking something bogus
650    back into the CVS repository...  (File config/targets.customized must
651    not be added to the repository!)
652    
653    ----------------------------------------------------------------------
654    Name: Matthias Blume
655    Date: 2001/05/22 16:30:00 EDT
656    Tag: blume-20010522-minitut
657    Description:
658    
659    1. Bug fix in ml-nlffigen; now (hopefully) correctly handling
660       struct returns.
661    2. Added src/ml-nlffi-lib/Doc/mini-tutorial.txt.  This is some very
662       incomplete, preliminary documentation for NLFFI.
663    
664    ----------------------------------------------------------------------
665    Name: Matthias Blume
666    Date: 2001/05/14 11:30:00 EDT
667    Tag: blume-20010514-script
668    Description:
669    
670    Some bugs in install script fixed.
671    
672    In addition to that I also made a slight change to the NLFFI API:
673    Functors generated by ml-nlffigen now take the dynamic library as a
674    straight functor argument, not as a suspended one.  (The original
675    functor code used to force the suspension right away anyway, so there
676    was nothing gained by this complication of the interface.)
677    
678    ----------------------------------------------------------------------
679    Name: Matthias Blume
680    Date: 2001/05/11 14:35:00 EDT
681    Tag: blume-20010511-ml-nlffi
682    Description:
683    
684    I finally took the plunge and added my new FFI code to the main
685    repository.  For x86-linux it is now ready for prime-time.
686    
687    There are two new subdirectories of "src":
688    
689      - ml-nlffi-lib:
690           The utility library for programs using the FFI interface.
691           Here is the implementation of $/c.cm and its associated low-level
692           partners $/c-int.cm and $/memory.cm.
693      - ml-nlffigen:
694           A stand-alone program for generating ML glue code from C source
695           code.
696    
697    Building ml-nlffigen requires $/ckit-lib.cm.
698    
699    The config/install.sh script has been updates to do the Right Thing
700    (hopefully).
701    
702    Notice that the source tree for the C-Kit will not be put under "src"
703    but directly under the installation root directory.  (This is the
704    structure that currently exists on the CVS server when you check out
705    module "sml".)  Fortunately, config/install.sh knows about this oddity.
706    
707    Bugs: No documentation yet.
708    
709    ----------------------------------------------------------------------
710    Name: Matthias Blume
711    Date: 2001/05/09 16:35:00 EDT
712    Tag: blume-20010509-cpscontract
713    Description:
714    
715    Fixed a bug in the accounting code in cpsopt/contract.sml.  (The
716    wrapper/unwrapper elimination did not decrement usage counts and some
717    dead variables got overlooked by the dead-up logic.)
718    
719    ----------------------------------------------------------------------
720    Name: Lal George
721    Date: 2001/05/08  17:26:09 EDT
722    Tag: george-20010508-omit-frameptr
723    Description:
724    
725    Changes to implement the omit-frame-pointer optimization to support
726    raw C calls. For now, there is only support on the Intel x86, but
727    other architectures will follow as more experience is gained with this.
728    
729    
730    ----------------------------------------------------------------------
731    Name: Matthias Blume
732    Date: 2001/05/07 14:40:00 EDT
733    Tag: blume-20010507-proxies
734    Description:
735    
736    I made into "proxy libraries" all libraries that qualify for such a
737    change.  (A qualifying library is a library that has another library or
738    groups as its sole member and repeats that member's export list
739    verbatim.  A proxy library avoids this repetition by omitting its export
740    list, effectively inheriting the list that its (only) member exports.
741    See the CM manual for more explanation.)
742    The main effect is that explicit export lists for these libraries
743    do not have to be kepts in sync, making maintenance a bit easier.
744    
745    I also added copyright notices to many .cm-files.
746    
747    Last but not least, I made a new set of bootfiles.
748    
749    ----------------------------------------------------------------------
750    Name: Matthias Blume
751    Date: 2001/05/04 17:00:00 EDT
752    Tag: blume-20010504-cm-lsplit
753    Description:
754    
755    0. John merged pending changes to $/smlnj-lib.cm
756    
757    1. Allen's previous change accidentally backed out of one of Lal's
758       earlier changes.  I undid this mistake (re-introducing Lal's change).
759    
760    2. I used the new topOrder' function from graph-scc.sml (from $/smlnj-lib.cm)
761       within the compiler where applicable.  There is some code simplification
762       because of that.
763    
764    3. The "split" phase (in FLINT) is now part of the default list of phases.
765       Compiler.Control.LambdaSplitting.* can be used to globally control the
766       lambda-splitting (cross-module-inlining) engine.  In addition to that,
767       it can now also be controlled on a per-source basis: CM has been taught
768       a new tool parameter applicable to ML source files.
769    
770       - To turn lambda-splitting off completely:
771            local open Compiler.Control.LambdaSplitting in
772                val _ = set Off
773            end
774       - To make "no lambda-splitting" the global default (but allow per-source
775         overriding); this is the initial setting:
776            local open Compiler.Control.LambdaSplitting in
777                val _ = set (Default NONE)
778            end
779       - To make "lambda-splitting with aggressiveness a" the global default
780         (and allow per-source overriding):
781            local open Compiler.Control.LambdaSplitting in
782                val _ = set (Default (SOME a))
783            end
784    
785       - To turn lambda-splitting off for a given ML souce file (say: a.sml)
786         write (in the respective .cm-file):
787            a.sml (lambdasplitting:off)
788       - To turn lambda-splitting for a.sml on with minimal aggressiveness:
789            a.sml (lambdasplitting:on)
790       - To turn lambda-splitting for a.sml on with aggressiveness <a> (where
791         <a> is a decimal non-negative integer):
792            a.sml (lambdasplitting:<a>)
793       - To turn lambda-splitting for a.sml on with maximal aggressiveness:
794            a.sml (lambdasplitting:infinity)
795       - To use the global default for a.sml:
796            a.sml (lambdasplitting:default)
797         or simply
798            a.sml
799    
800    ----------------------------------------------------------------------
801    Name: Allen Leung
802    Date: 2001/05/04 01:57:00 EDT
803    Tag: leunga-20010504-sync
804    Description:
805    
806      MLRISC features.
807    
808      1. Fix to CMPXCHG instructions.
809      2. Changed RA interface to allow annotations in callbacks.
810      3. Added a new method to the stream interface to allow annotations updates.
811    
812    ----------------------------------------------------------------------
813    Name: Matthias Blume
814    Date: 2001/05/01 11:45:00 EDT
815    Tag: blume-20010501-pcedittmp
816    Description:
817    
818    Changed install.sh to use the current working directory instead of
819    /usr/tmp for a temporary file (pcedittmp).  The previous choice
820    of /usr/tmp caused trouble with MacOS X because of file premission
821    problems.
822    
823    ----------------------------------------------------------------------
824    Name: Matthias Blume
825    Date: 2001/04/20 11:10:00 EDT
826    Tag: blume-20010420-inMLflag
827    Description:
828    
829     - added vp_limitPtrMask to vproc-state.h
830       (for use by the raw-C-calls mechanism to implement proper interrupt
831        handling)
832     - made the ML compiler aware of various data-structure offsets so it
833       can generate code for accessing the vp_inML flag and vp_limitPtrMask
834     - tweaked mlriscGen.sml to have it emit interrupt-handling code for
835       raw C-calls
836    
837    ----------------------------------------------------------------------
838    Name: Lal George
839    Date: 2001/04/20 09:15:28 EDT
840    Tag: george-20010420-macosX
841    Description:
842    
843     - Changes to port to Mac OS X; Darwin.
844    
845     - In the process I found that sqrt was broken on the PPC, because the
846       fsqrt instruction is not implemented.
847    
848    ----------------------------------------------------------------------
849    Name: Matthias Blume
850    Date: 2001/04/18 12:45:00 EDT
851    Tag: blume-20010418-ccalls
852    Description:
853    
854     - fixed two off-by-4 errors in the x86-specific c-calls implementation
855       (this bug prevented structure arguments containing pointers from being
856        passed correctly)
857     - changed the raw-C-call code in mlriscGen.sml in such a way that
858       structure arguments are represented as a pointer to the beginning
859       of the structure (instead of having a series of synthesized arguments,
860       one for each structure member)
861    
862     - made makeml script's verbosity level configurable via environment
863       variable (MAKEML_VERBOSITY)
864    
865     - eliminated placeholder implementations for f32l, w16s, i16s, and f32s
866       in rawmem-x86.sml; we are now using the real thing
867    
868    ----------------------------------------------------------------------
869    Name: Matthias Blume
870    Date: 2001/03/22 16:25:00 EST
871    Tag: blume-20010322-bootfiles
872    Description:
873    
874    Created a new set of bootfiles (for your automatic installation convenience).
875    
876    ----------------------------------------------------------------------
877    Name: Matthias Blume
878    Date: 2001/03/22 15:10:00 EST
879    Tag: blume-20010322-rawmem-parcm
880    Description:
881    
882    1. All "raw memory access" primitives for the new FFI are implemented now
883       (at least on the x86).
884    2. Some further cleanup of CM's parallel make mechanism.
885    
886    ----------------------------------------------------------------------
887    Name: Matthias Blume
888    Date: 2001/03/19 17:53:00 EST
889    Tag: blume-20010319-parallel
890    Description:
891    
892    Parallel make (using compile servers) now works again.
893    
894    To this end, CM.stabilize and CMB.make have been modified to work in
895    two passes when compile servers are attached:
896       1. Compile everything, do not perform stabilization; this pass
897          uses compile servers
898       2. Stabilize everything; this pass does not use compile servers
899    If there are no compile servers, the two passes are combined into one
900    (as before).  Splitting the passes increases the inherent parallelism
901    in the dependency graph because the entire graph including all
902    libraries is available at the same time.  This, in turn, improves
903    server utilization.  The downside is that the master process will
904    have to do some extra work after compilation is done (because for
905    technical reasons it must re-read all the binfiles during stabilization).
906    
907    ----------------------------------------------------------------------
908    Name: Matthias Blume
909    Date: 2001/03/16 12:22:00 EST
910    Tag: blume-20010316-bootfiles
911    Description:
912    
913    Created a new set of bootfiles (for your automatic installation convenience).
914    
915    ----------------------------------------------------------------------
916    Name: Matthias Blume
917    Date: 2001/03/16 11:00:00 EST
918    Tag: blume-20010316-MLTREE-fixup
919    Description:
920    
921    This is a minor fixup for an (untagged) earlier commit by Allen.
922    (A file was missing).
923    
924    ----------------------------------------------------------------------
925    Name: Allen Leung
926    Date: Mon Mar  5 18:54:57 EST 2001
927    Tag: leunga-20010305-cut-support
928    
929    1. New support for alternative control-flow in MLTREE.
930       Currently we support
931    
932          FLOW_TO(CALL ...., [k1,...,kn])
933    
934       This is needed for 'cuts to' in C-- and try/handle-like constructs
935       in Moby
936    
937       New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
938    
939    2. Register Allocator
940    
941       Changes in interface [from Fermin, John]
942    
943    3. Alpha 8-bit SLL support [Fermin]
944    
945    4. All architectures
946    
947       A new module (ClusterExpandCopies) for expanding parallel copies.
948    
949    ----------------------------------------------------------------------
950    Name: Allen Leung
951    Date: 2001/02/27 23:07:00 EST
952    Tag: leunga-20010227-minor-stuff
953    
954    1. Alpha bug fix for CMOVNE
955    2. Handle mltree COND(..,FCMP ...,...)
956    3. Bug fix in simplifier
957    
958    ----------------------------------------------------------------------
959    Name: Matthias Blume
960    Date: 2001/01/30 17:50:00 EST
961    Tag: blume-20010130-sync
962    Description:
963    
964    This is just a minor update to sync my devel branch with the main brach.
965    The only visible change is the addition of some README files.
966    
967    ----------------------------------------------------------------------
968    Name: Matthias Blume
969    Date: 2001/01/12 23:30:00 JST
970    Tag: blume-20010112-bootfiles
971    Description:
972    
973    Made a new set of bootfiles that goes with the current state of the
974    repository.
975    
976    ----------------------------------------------------------------------
977    Name: Matthias Blume
978    Date: 2001/01/12 21:20:00 JST
979    Tag: blume-20010112-sync
980    Description:
981    
982    I am just flushing out some minor changes that had accumulated in
983    my private branch in order to sync with the main tree.  (This is
984    mainly because I had CVS trouble when trying to merge _into_ my
985    private branch.)
986    
987    Most people should be completely unaffected by this.
988    
989    ----------------------------------------------------------------------
990    Name: Allen Leung
991    Date: Thu Jan 11 21:03:00 EST 2001
992    Tag: leunga-20010111-labexp=mltree
993    Description:
994    
995    1.  Removed the type LabelExp and replace it by MLTree.
996    2.  Rewritten mltree-simplify with the pattern matcher tool.
997    3.  There were some bugs in alpha code generator which would break
998        64-bit code generation.
999    4.  Redo the tools to generate code with the
1000    5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
1001        are now generated by perl scripts.
1002    
1003    ----------------------------------------------------------------------
1004    Name: Matthias Blume
1005    Date: 2001/01/10 21:55:00 JST
1006    Tag: blume-20010110-rcc
1007    Description:
1008    
1009    The RCC stuff now seems to work (but only on the x86).
1010    This required hacking of the c-calls interface (and -implementation) in
1011    MLRISC.
1012    
1013    Normal compiler users should be unaffected.
1014    
1015  ----------------------------------------------------------------------  ----------------------------------------------------------------------
1016  Name: Matthias Blume  Name: Matthias Blume
1017  Date: 2001/01/09 01:20:00 JST  Date: 2001/01/09 01:20:00 JST

Legend:
Removed from v.773  
changed lines
  Added in v.904

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