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 705, Wed Sep 27 18:44:44 2000 UTC revision 938, Thu Sep 20 21:27:33 2001 UTC
# Line 13  Line 13 
13  Description:  Description:
14    
15  ----------------------------------------------------------------------  ----------------------------------------------------------------------
16    Name: Matthias Blume
17    Date: 2001/09/20 17:28:00 EDT
18    Tag: blume-20010920-slowfp
19    Description:
20    
21    Allen says that x86-fast-fp is not safe yet, so I turned it off again...
22    
23    ----------------------------------------------------------------------
24    Name: Matthias Blume
25    Date: 2001/09/20 17:20:00 EDT
26    Tag: blume-20010920-canonicalpaths
27    Description:
28    
29    0. Updated the BOOT file (something that I forgot to do earlier).
30    
31    1. Small internal change to CM so that it avoids "/../" in filenames
32       as much as possible (but only where it is safe).
33    
34    2. Changed config/_run-sml (resulting in a changed bin/.run-sml) so
35       that arguments that contain delimiters are passed through correctly.
36       This change also means that all "special" arguments of the form
37       @SMLxxx... must come first.
38    
39    3. Changed install script to put relative anchor names for tool commands
40       into pathconfig.
41    
42    ----------------------------------------------------------------------
43    Name: Matthias Blume
44    Date: 2001/09/18 15:35:00 EDT
45    Tag: blume-20010918-readme11036
46    Description:
47    
48    Added README files.
49    
50    ----------------------------------------------------------------------
51    Name: Matthias Blume
52    Date: 2001/09/18 11:45:00 EDT
53    Tag: Release_110_36 (retag)
54    Description:
55    
56    Fixed mistake in config/preloads. Retagged as 110.36.
57    
58    ----------------------------------------------------------------------
59    Name: Matthias Blume
60    Date: 2001/09/18 09:40:00 EDT
61    Tag: Release_110_36_orig (tag changed)
62    Description:
63    
64    New version (110.36).  New bootfiles.
65    
66    ----------------------------------------------------------------------
67    Name: Matthias Blume
68    Date: 2001/09/14 16:15:00 EDT
69    Tag: blume-20010914-x86fastfp
70    Description:
71    
72    John committed some changes that Allen made, in particular a (hopefully)
73    correctly working version of the x86-fp module.
74    
75    I changed the default setting of the Control.MLRISC.getFlag "x86-fast-fp"
76    flag to "true".  Everything seems to compile to a fixpoint ok, and
77    "mandelbrot" speeds up by about 15%.
78    
79    ----------------------------------------------------------------------
80    Name: Matthias Blume
81    Date: 2001/09/13 11:20:00 EDT
82    Tag: blume-20010913-minimal
83    Description:
84    
85    1. Stefan Monnier's patch to fix a miscompilation problem that
86       was brought to light by John Reppy's work on Moby.
87    
88    2. Implemented a minimal "structure Compiler" that contains just
89       "version" and "architecture".  The minimal version will be
90       available when the full version is not.  This is for backward-
91       compatibility with code that wants to test Compiler.version.
92    
93    ----------------------------------------------------------------------
94    Name: Matthias Blume
95    Date: 2001/08/28 14:03:00 EDT
96    Tag: blume-20010828-ml-lex
97    Description:
98    
99    Fix for bug 1581, received from Neophytos Michael.
100    
101    ----------------------------------------------------------------------
102    Name: Matthias Blume
103    Date: 2001/08/27 11:20:00 EDT
104    Tag: blume-20010827-readme11035
105    Description:
106    
107    Fleshed out the README file for 110.35.
108    
109    ----------------------------------------------------------------------
110    Name: Matthias Blume
111    Date: 2001/08/24 17:10:00 EDT
112    Tag: Release_110_35
113    Description:
114    
115    New version number (110.35).  New bootfiles.
116    
117    ----------------------------------------------------------------------
118    Name: Lal George
119    Date: 2001/08/24 13:47:18 EDT 2001
120    Tag: george-20010824-MLRISC-graphs
121    Description:
122    
123     removed clusters from MLRISC completely and replaced with graphs.
124    
125    ----------------------------------------------------------------------
126    Name: Matthias Blume
127    Date: 2001/08/23 17:50:00 EDT
128    Tag: blume-20010823-toplevel
129    Description:
130    
131    - some reorganization of the code that implements various kinds of
132      environments in the compiler (static, dynamic, symbolic, combined)
133    - re-implemented the EnvRef module so that evalStream works properly
134      (if the stream contains references to "use", "CM.make", etc.)
135    - cleaned up evalloop.sml and interact.sml (but they need more cleaning)
136    
137    ----------------------------------------------------------------------
138    Name: Matthias Blume
139    Date: 2001/08/20 15:50 EDT
140    Tag: blume20010820-slipup
141    Description:
142    
143    I forgot to commit a few files.  Here they are...
144    
145    ----------------------------------------------------------------------
146    Name: Matthias Blume
147    Date: 2001/08/20 15:35:00 EDT
148    Tag: blume-20010820-debugprof
149    Description:
150    
151    !!!! NEW BOOTFILES !!!!
152    
153    This is another round of reorganizing the compiler sources.  This
154    time the main goal was to factor out all the "instrumentation"
155    passes (for profiling and backtracing) into their own library.
156    The difficulty was to do it in such a way that it does not depend
157    on elaborate.cm but only on elabdata.cm.
158    
159    Therefore there have been further changes to both elaborate.cm and
160    elabdata.cm -- more "generic" things have been moved from the former
161    to the latter.  As a result, I was forced to split the assignment
162    of numbers indicating "primtyc"s into two portions: SML-generic and
163    SML/NJ-specific.  Since it would have been awkward to maintain,
164    I bit the bullet and actually _changed_ the mapping between these
165    numbers and primtycs.  The bottom line of this is that you need
166    a new set of bin- and bootfiles.
167    
168    I have built new bootfiles for all architectures, so doing a fresh
169    checkout and config/install.sh should be all you need.
170    
171    The newly created library's name is
172    
173        $smlnj/viscomp/debugprof.cm
174    
175    and its sources live under
176    
177        src/compiler/DebugProf
178    
179    ----------------------------------------------------------------------
180    Name: Matthias Blume
181    Date: 2001/08/15 17:15:00 EDT
182    Tag: blume-20010815-compreorg
183    Description:
184    
185    This is a first cut at reorganizing the CM libraries that make up the
186    core of the compiler.  The idea is to separate out pieces that could
187    be used independently by tools, e.g., the parser, the typechecker, etc.
188    
189    The current status is a step in this direction, but it is not quite
190    satisfactory yet.  Expect more changes in the future.
191    
192    Here is the current (new) organization...
193    
194        What used to be $smlnj/viscomp/core.cm is now divided into
195        six CM libraries:
196    
197             $smlnj/viscomp/basics.cm
198                           /parser.cm
199                           /elabdata.cm
200                           /elaborate.cm
201                           /execute.cm
202                           /core.cm
203    
204        The CM files for these libraries live under src/system/smlnj/viscomp.
205        All these libraries are proxy libraries that contain precisely
206        one CM library component.  Here are the locations of the components
207        (all within the src/compiler tree):
208    
209             Basics/basics.cm
210             Parse/parser.cm
211             ElabData/elabdata.cm
212             Elaborator/elaborate.cm
213             Execution/execute.cm
214             core.cm
215    
216         [This organization is the same that has been used already
217         for a while for the architecture-specific parts of the visible
218         compiler and for the old version of core.cm.]
219    
220         As you will notice, many source files have been moved from their
221         respective original locations to a new home in one of the above
222         subtrees.
223    
224         The division of labor between the new libraries is the following:
225    
226             basics.cm:
227                - Simple, basic definitions that pertain to many (or all) of
228                  the other libraries.
229             parser.cm:
230                - The SML parser, producing output of type Ast.dec.
231                - The type family for Ast is also defined and exported here.
232             elabdata.cm:
233                - The datatypes that describe input and output of the elaborator.
234                  This includes types, absyn, and static environments.
235             elaborator.cm:
236                - The SML/NJ type checker and elaborator.
237                  This maps an Ast.dec (with a given static environment) to
238                  an Absyn.dec (with a new static environment).
239                - This libraries implements certain modules that used to be
240                  structures as functors (to remove dependencies on FLINT).
241             execute.cm:
242                - Everything having to do with executing binary code objects.
243                - Dynamic environments.
244             core.cm:
245                - SML/NJ-specific instantiations of the elaborator and MLRISC.
246                - Top-level modules.
247                - FLINT (this should eventually become its own library)
248    
249    Notes:
250    
251    I am not 100% happy with the way I separated the elaborator (and its
252    data structures) from FLINT.  Two instances of the same problem:
253    
254        1. Data structures contain certain fields that carry FLINT-specific
255           information.  I hacked around this using exn and the property list
256           module from smlnj-lib.  But the fact that there are middle-end
257           specific fields around at all is a bit annoying.
258    
259        2. The elaborator calculates certain FLINT-related information.  I tried
260           to make this as abstract as I could using functorization, but, again,
261           the fact that the elaborator has to perform calculations on behalf
262           of the middle-end at all is not nice.
263    
264        3. Having to used exn and property lists is unfortunate because it
265           weakens type checking.  The other alternative (parameterizing
266           nearly *everything*) is not appealing, though.
267    
268    I removed the "rebinding =" warning hack because due to the new organization
269    it was awkward to maintain it.  As a result, the compiler now issues some of
270    these warnings when compiling init.cmi during bootstrap compilation. On
271    the plus side, you also get a warning when you do, for example:
272       val op = = Int32.+
273    which was not the case up to now.
274    
275    I placed "assign" and "deref" into the _Core structure so that the
276    code that deals with the "lazy" keyword can find them there.  This
277    removes the need for having access to the primitive environment
278    during elaboration.
279    
280    ----------------------------------------------------------------------
281    Name: Matthias Blume
282    Date: 2001/08/13
283    Tag: blume-20010813-closures
284    Description:
285    
286    This fix was sent to us by Zhong Shao.  It is supposed to improve the
287    performance of certain loops by avoiding needless closure allocation.
288    
289    ----------------------------------------------------------------------
290    Name: Lal George
291    Date: 2001/07/31 10:03:23 EDT 2001
292    Tag: george-20010731-x86-fmalloc
293    Description: Fixed bug in x86 calls
294    
295        There was a bug where call instructions would mysteriously
296        vanish. The call instruction had to be one that returned
297        a floating point value.
298    
299    ----------------------------------------------------------------------
300    Name: Lal George
301    Date: 2001/07/19 16:36:29 EDT 2001
302    Tag: george-20010719-simple-cells
303    Description:
304    
305    I have dramatically simplified the interface for CELLS in MLRISC.
306    
307    In summary, the cells interface is broken up into three parts:
308    
309      1. CellsBasis : CELLS_BASIS
310    
311            CellsBasis is a top level structure and common for all
312            architectures.  it contains the definitions of basic datatypes
313            and utility  functions over these types.
314    
315      2. functor Cells() : CELLS
316    
317            Cells generates an interface for CELLS that incorporates the
318            specific resources on the target architecture, such as the
319            presence of special register classes, their number and size,
320            and various useful substructures.
321    
322      3. <ARCH>CELLS
323    
324            e.g. SparcCells: SPARCCELLS
325    
326            <ARCH>CELLS usually contains additional bindings for special
327            registers  on the architecture, such as:
328    
329                    val r0 : cell           (* register zero *)
330                    val y : cell            (* Y register *)
331                    val psr : cell          (* processor status register *)
332                    ...
333    
334            The structure returned by applying the Cells functor is opened
335            in this interface.
336    
337    The main implication of all this is that the datatypes for cells is
338    split between CellsBasis and CELLS -- a fairly simple change for user
339    code.
340    
341    In the old scheme the CELLS interface had a definitional binding of
342    the form:
343    
344            signature CELLS = sig
345    
346               structure CellsBasis = CellsBasis
347    
348               ...
349    
350            end
351    
352    With all the sharing constraints that goes on in MLRISC, this old
353    design  quickly leads to errors such as:
354    
355            "structure definition spec inside of sharing ... "
356    
357    
358    and appears to require an unacceptable amount of sharing and where
359    constraint hackery.
360    
361    I think this error message (the interaction of definitional specs and
362    sharing) requires more explanation on our web page.
363    
364    ----------------------------------------------------------------------
365    Name: Matthias Blume
366    Date: 2001/07/19 15:00:00 EDT
367    Tag: blume-20010719-libreorg
368    Description:
369    
370    This update puts together a fairly extensive but straightforward change
371    to the way the libraries that implement the interactive system are
372    organized:
373    
374       The biggest change is the elimination of structure Compiler.  As a
375       replacement for this structure, there is now a CM library
376       (known as $smlnj/compiler.cm or $smlnj/compiler/current.cm)
377       that exports all the substructures of the original structure Compiler
378       directly.  So instead of saying Compiler.Foo.bar one now simply
379       says Foo.bar.  (The CM libraries actually export a collection of
380       structures that is richer than the collection of substructures of
381       structure Compiler.)
382    
383       To make the transition smooth, there is a separate library called
384       $smlnj/compiler/compiler.cm which puts together and exports the
385       original structure Compiler (or at least something very close to it).
386    
387       There are five members of the original structure Compiler
388       that are not exported directly but which instead became members
389       of a new structure Backend (described by signature BACKEND).  These are:
390       structure Profile (: PROFILE), structure Compile (: COMPILE), structure
391       Interact (: INTERACT), structure Machine (: MACHINE), and val
392       architecture (: string).
393    
394       Structure Compiler.Version has become structure CompilerVersion.
395    
396       Cross-compilers for alpha32, hppa, ppc, sparc, and x86 are provided
397       by $smlnj/compiler/<arch>.cm where <arch> is alpha32, hppa, ppc, sparc,
398       or x86, respectively.
399       Each of these exports the same frontend structures that
400       $smlnj/compiler.cm exports.  But they do not have a structure Backend
401       and instead export some structure <Arch>Backend where <Arch> is Alpha32,
402       Hppa, PPC, Sparc, or X86, respectively.
403    
404       Library $smlnj/compiler/all.cm exports the union of the exports of
405       $smlnj/compiler/<arch>.cm
406    
407       There are no structures <Arch>Compiler anymore, use
408       $smlnj/compiler/<arch>.cm instead.
409    
410       Library host-compiler-0.cm is gone.  Instead, the internal library
411       that instantiates CM is now called cm0.cm.  Selection of the host
412       compiler (backend) is no longer done here but. (Responsibility for it
413       now lies with $smlnj/compiler/current.cm.  This seems to be more
414       logical.)
415    
416       Many individual files have been moved or renamed.  Some files have
417       been split into multiple files, and some "dead" files have been deleted.
418    
419    Aside from these changes to library organization, there are also changes
420    to the way the code itself is organized:
421    
422       Structure Binfile has been re-implemented in such a way that it no
423       longer needs any knowledge of the compiler.  It exclusively deals
424       with the details of binfile layout.  It no longer invokes the
425       compiler (for the purpose of creating new prospective binfile
426       content), and it no longer has any knowledge of how to interpret
427       pickles.
428    
429       Structure Compile (: COMPILE) has been stripped down to the bare
430       essentials of compilation.  It no longer deals with linking/execution.
431       The interface has been cleaned up considerably.
432    
433       Utility routines for dealing with linking and execution have been
434       moved into their own substructures.
435    
436       (The ultimate goal of these changes is to provide a light-weight
437       binfile loader/linker (at least for, e.g., stable libraries) that
438       does not require CM or the compiler to be present.)
439    
440    CM documentation has been updated to reflect the changes to library
441    organization.
442    
443    ----------------------------------------------------------------------
444    Name: Matthias Blume
445    Date: 2001/07/10 17:30:00 EDT
446    Tag: Release_110_34
447    Description:
448    
449    Minor tweak to 110.34 (re-tagged):
450    
451      - README.html file added to CVS repository
452      - runtime compiles properly under FreeBSD 3.X and 4.X
453    
454    ----------------------------------------------------------------------
455    Name: Matthias Blume
456    Date: 2001/07/10 17:30:00 EDT
457    Tag: Release_110_34
458    Description:
459    
460    New version number (110.34). New bootfiles.
461    
462    ----------------------------------------------------------------------
463    Name: Matthias Blume
464    Date: 2001/07/09 16:00:00 EDT
465    Tag: blume-20010709-more-varargs
466    Description:
467    
468    I changed the handling of varargs in ml-nlffigen again:
469    The ellipsis ... will now simply be ignored (with an accompanying warning).
470    
471    The immediate effect is that you can actually call a varargs function
472    from ML -- but you can't actually supply any arguments beyond the ones
473    specified explicitly.  (For example, you can call printf with its format
474    string, but you cannot pass additional arguments.)
475    
476    This behavior is only marginally more useful than the one before, but
477    it has the advantage that a function or, more importantly, a function
478    type never gets dropped on the floor, thus avoiding follow-up problems with
479    other types that refer to the offending one.
480    
481    ----------------------------------------------------------------------
482    Name: Matthias Blume
483    Date: 2001/07/09 11:25:00 EDT
484    Tag: blume-20010709-varargs
485    Description:
486    
487    1. ckit-lib.cm now exports structure Error
488    2. ml-nlffigen reports occurences of "..." (i.e., varargs function types)
489       with a warning accompanied by a source location.  Moreover, it
490       merely skips the offending function or type and proceeds with the
491       rest of its work.u  As a result, one can safely feed C code containing
492       "..." to ml-nlffigen.
493    3. There are some internal improvements to CM, providing slightly
494       more general string substitutions in the tools subsystem.
495    
496    ----------------------------------------------------------------------
497    Name: Matthias Blume
498    Date: 2001/06/27 15:10:00 EDT
499    Tag: blume-20010627-concur
500    Description:
501    
502    Fixed a small bug in CM's handling of parallel compilation.
503    (You could observe the bug by Control-C-interrupting an ordinary
504    CMB.make or CM.stabilize and then attaching some compile servers.
505    The result was that all of a sudden the previously interrupted
506    compilation would continue on its own.  This was because of
507    an over-optimization: CM did not bother to clean out certain queues
508    when no servers were attached "anyway", resulting in the contents
509    of these queues to grab control when new servers did get attached.)
510    
511    There is also another minor update to the CM manual.
512    
513    ----------------------------------------------------------------------
514    Name: Matthias Blume
515    Date: 2001/06/26 16:15:00 EDT
516    Tag: blume-20010626-cmdoc
517    Description:
518    
519    Minor typo fixed in CM manual (syntax diagram for libraries).
520    
521    ----------------------------------------------------------------------
522    Name: Matthias Blume
523    Date: 2001/06/25 22:55:00 EDT
524    Tag: blume-20010625-x86pc
525    Description:
526    
527    Fixed a nasty bug in the X86 assembly code that caused signal
528    handlers to fail (crash) randomly.
529    
530    ----------------------------------------------------------------------
531    Name: Matthias Blume
532    Date: 2001/06/25 12:05:00 EDT
533    Tag: blume-20010625-nlffigen
534    Description:
535    
536    This update fixes a number of minor bugs in ml-nlffigen as reported by
537    Nick Carter <nbc@andrew.cmu.edu>.
538    
539      1. Silly but ok typedefs of the form "typedef void myvoid;" are now accepted.
540      2. Default names for generated files are now derived from the name of
541         the C file *without its directory*.  In particular, this causes generated
542         files to be placed locally even if the C file is in some system directory.
543      3. Default names for generated signatures and structures are also derived
544         from the C file name without its directory.  This avoids silly things
545         like "structure GL/GL".
546         (Other silly names are still possible because ml-nlffigen does not do
547          a thorough check of whether generated names are legal ML identifiers.
548          When in doubt, use command line arguments to force particular names.)
549    
550    ----------------------------------------------------------------------
551    Name: Matthias Blume
552    Date: 2001/06/21 12:25:00 EDT
553    Tag: blume-20010621-eXene
554    Description:
555    
556    eXene now compiles and (sort of) works again.
557    
558    The library name (for version > 110.33) is $/eXene.cm.
559    
560    I also added an new example in src/eXene/examples/nbody.  See the
561    README file there for details.
562    
563    ----------------------------------------------------------------------
564    Name: Matthias Blume
565    Date: 2001/06/20 16:40:00 EDT
566    Tag: blume-20010620-cml
567    Description:
568    
569    CML now compiles and works again.
570    
571    Libraries (for version > 110.33):
572    
573      $cml/cml.cm            Main CML library.
574      $cml/basis.cm          CML's version of $/basis.cm.
575      $cml/cml-internal.cm   Internal helper library.
576      $cml/core-cml.cm       Internal helper library.
577      $cml-lib/trace-cml.cm  Tracing facility.
578      $cml-lib/smlnj-lib.cm  CML's version of $/smlnj-lib.cm
579    
580    The installer (config/install.sh) has been taught how to properly
581    install this stuff.
582    
583    ----------------------------------------------------------------------
584    Name: Matthias Blume
585    Date: 2001/06/19 17:55:00 EDT
586    Tag: blume-20010619-instantiate
587    Description:
588    
589    This un-breaks the fix for bug 1432.
590    (The bug was originally fixed in 110.9 but I broke it again some
591    time after that.)
592    
593    ----------------------------------------------------------------------
594    Name: Matthias Blume
595    Date: 2001/06/19 17:25:00 EDT
596    Tag: blume-20010619-signals
597    Description:
598    
599    This should (hopefully) fix the long-standing signal handling bug.
600    (The runtime system was constructing a continuation record with an
601    incorrect descriptor which would cause the GC to drop data on the floor...)
602    
603    ----------------------------------------------------------------------
604    Name: Matthias Blume
605    Date: 2001/06/15 15:05:00 EDT
606    Tag: blume-20010615-moresparc
607    Description:
608    
609    Here is a short late-hour update related to Sparc c-calls:
610    
611     -- made handling of double-word arguments a bit smarter
612    
613     -- instruction selection phase tries to collapse certain clumsily
614        constructed ML-Trees; typical example:
615    
616            ADD(ty,ADD(_,e,LI d1),LI d2)  ->  ADD(ty,e,LI(d1+d2))
617    
618        This currently has no further impact on SML/NJ since mlriscGen does
619        not seem to generate such patterns in the first place, and c-calls
620        (which did generate them in the beginning) has meanwhile been fixed
621        so as to avoid them as well.
622    
623    ----------------------------------------------------------------------
624    Name: Matthias Blume
625    Date: 2001/06/15 15:05:00 EDT
626    Tag: blume-20010615-sparc
627    Description:
628    
629    The purpose of this update is to provide an implementation of NLFFI
630    on Sparc machines.
631    
632    Here are the changes in detail:
633    
634       * src/MLRISC/sparc/c-calls/sparc-c-calls.sml is a new file containing
635       the Sparc implementation of the c-calls API.
636       * The Sparc backend of SML/NJ has been modified to uniformely use %fp
637       for accessing the ML frame.  Thus, we have a real frame pointer and
638       can freely modify %sp without need for an omit-frame-ptr phase.
639       The vfp logic in src/compiler/CodeGen/* has been changed to accomodate
640       this case.
641       * ml-nlffigen has been taught to produce code for different architectures
642       and calling conventions.
643       * In a way similar to what was done in the x86 case, the Sparc
644       backend uses its own specific extension to mltree.  (For example,
645       it needs to be able to generate UNIMP instructions which are part
646       of the calling convention.)
647       * ml-nlffi-lib was reorganized to make it more modular (in particular,
648       to make it easier to plug in new machine- and os-dependent parts).
649    
650    There are some other fairly unrelated bug fixes and cleanups as well:
651    
652       * I further hacked the .cm files for MLRISC tools (like MDLGen) so
653       that they properly share their libraries with existing SML/NJ libraries.
654       * I fixed a minor cosmetic bug in CM, supressing certain spurious
655       follow-up error messages.
656       * Updates to CM/CMB documentation.
657    
658    TODO items:
659    
660       * MLRISC should use a different register as its asmTemp on the Sparc.
661         (The current %o2 is a really bad choice because it is part of the
662         calling conventions, so things might interfere in unexpected ways.)
663    
664    ----------------------------------------------------------------------
665    Name: Matthias Blume
666    Date: 2001/06/07
667    Tag: blume-20010607-calls
668    Description:
669    
670    A number of internal changes related to C calls and calling conventions:
671    
672    1. ML-Tree CALL statements now carry a "pops" field.  It indicates the
673       number of bytes popped implicitly (by the callee).  In most cases
674       this field is 0 but on x86/win32 it is some non-zero value.  This
675       is information provided for the benefit of the "omit-frameptr" pass.
676    2. The CALL instruction on the x86 carries a similar "pops" field.
677       The instruction selection phase copies its value from the ML-Tree
678       CALL statement.
679    3. On all other architectures, the instruction selection phase checks
680       whether "pops=0" and complains if not.
681    4. The c-calls implementation for x86 now accepts two calling conventions:
682       "ccall" and "stdcall".  When "ccall" is selected, the caller cleans
683       up after the call and pops is set to 0.  For "stdcall", the caller
684       does nothing, leaving the cleanup to the callee; pops is set to
685       the number of bytes that were pushed onto the stack.
686    5. The cproto decoder (compiler/Semant/types/cproto.sml) now can
687       distinguish between "ccall" and "stdcall".
688    6. The UNIMP instruction has been added to the supported Sparc instruction
689       set. (This is needed for implementing the official C calling convention
690       on this architecture.)
691    7. I fixed some of the .cm files under src/MLRISC/Tools to make them
692       work with the latest CM.
693    
694    ----------------------------------------------------------------------
695    Name: Matthias Blume
696    Date: 2001/06/05 15:10:00 EDT
697    Tag: blume-20010605-cm-index
698    Description:
699    
700    0. The "lambdasplit" parameter for class "sml" in CM has been documented.
701    
702    1. CM can now generate "index files".  These are human-readable files
703       that list on a per-.cm-file basis each toplevel symbol defined or
704       imported.  The location of the index file for
705       <p>/<d>.cm is <p>/CM/INDEX/<d>.cm.
706       To enable index-file generation, set CM.Control.generate_index to true
707       or export an environment-symbol: export CM_GENERATE_INDEX=true.
708    
709       The CM manual has been updated accordingly.
710    
711    2. I made some slight modifications to the c-calls API in MLRISC.
712    
713         a) There is now a callback to support saving/restoring of
714            dedicated but caller-save registers around the actual call
715            instruction.
716         b) One can optionally specify a comment-annotation for the
717            call instruction.
718    
719    3. SML/NJ (mlriscGen.sml) uses this new API for the rawccall primop.
720       (For example, the comment annotation shows the C prototype of
721        the function being called.)
722    
723    ----------------------------------------------------------------------
724    Name: Matthias Blume
725    Date: 2001/06/01 13:30:00 EDT
726    Tag: blume-20010601-nlffi-cleanup
727    Description:
728    
729    This is mostly a cleanup of MLFFI stuff:
730    
731       - some signature files have been put into a more exposed place
732       - the ugly 'f type parameter is gone (simplifies types tremendously!)
733       - ml-nlffigen changed accordingly
734       - tutorial updated
735    
736    Other changes:
737    
738       - author's affiliation in CM manual(s) updated
739       - some more recognized keywords added to Allen's sml.sty
740    
741    ----------------------------------------------------------------------
742    Name: Matthias Blume
743    Date: 2001/05/25 15:30:00 EDT
744    Tag: blume-20010525-iptr
745    Description:
746    
747      - put the official 110.33-README (as it appears on the ftp server) under
748        CVS
749      - fixed a small bug related to incomplete pointer types in
750        ml-nlffigen
751      - small cosmetic change to the ml-nlffi-lib's "arr" type constructor
752        (it does not need the 'f type parameter)
753    
754    ----------------------------------------------------------------------
755    Name: Matthias Blume
756    Date: 2001/05/23 14:30:00 EDT
757    Tag: Release_110_33
758    Description:
759    
760    New version number (110.33).  New bootfiles.
761    
762    ----------------------------------------------------------------------
763    Name: Matthias Blume
764    Date: 2001/05/22 18:06:00 EDT
765    Tag: blume-20010522-targets
766    Description:
767    
768    Made install.sh use file config/targets.customized if it exists, falling
769    back to config/targets if it doesn't.  This way one can have a customized
770    version of the targets file without touching the "real thing", thus
771    eliminating the constant fear of accidentally checking something bogus
772    back into the CVS repository...  (File config/targets.customized must
773    not be added to the repository!)
774    
775    ----------------------------------------------------------------------
776    Name: Matthias Blume
777    Date: 2001/05/22 16:30:00 EDT
778    Tag: blume-20010522-minitut
779    Description:
780    
781    1. Bug fix in ml-nlffigen; now (hopefully) correctly handling
782       struct returns.
783    2. Added src/ml-nlffi-lib/Doc/mini-tutorial.txt.  This is some very
784       incomplete, preliminary documentation for NLFFI.
785    
786    ----------------------------------------------------------------------
787    Name: Matthias Blume
788    Date: 2001/05/14 11:30:00 EDT
789    Tag: blume-20010514-script
790    Description:
791    
792    Some bugs in install script fixed.
793    
794    In addition to that I also made a slight change to the NLFFI API:
795    Functors generated by ml-nlffigen now take the dynamic library as a
796    straight functor argument, not as a suspended one.  (The original
797    functor code used to force the suspension right away anyway, so there
798    was nothing gained by this complication of the interface.)
799    
800    ----------------------------------------------------------------------
801    Name: Matthias Blume
802    Date: 2001/05/11 14:35:00 EDT
803    Tag: blume-20010511-ml-nlffi
804    Description:
805    
806    I finally took the plunge and added my new FFI code to the main
807    repository.  For x86-linux it is now ready for prime-time.
808    
809    There are two new subdirectories of "src":
810    
811      - ml-nlffi-lib:
812           The utility library for programs using the FFI interface.
813           Here is the implementation of $/c.cm and its associated low-level
814           partners $/c-int.cm and $/memory.cm.
815      - ml-nlffigen:
816           A stand-alone program for generating ML glue code from C source
817           code.
818    
819    Building ml-nlffigen requires $/ckit-lib.cm.
820    
821    The config/install.sh script has been updates to do the Right Thing
822    (hopefully).
823    
824    Notice that the source tree for the C-Kit will not be put under "src"
825    but directly under the installation root directory.  (This is the
826    structure that currently exists on the CVS server when you check out
827    module "sml".)  Fortunately, config/install.sh knows about this oddity.
828    
829    Bugs: No documentation yet.
830    
831    ----------------------------------------------------------------------
832    Name: Matthias Blume
833    Date: 2001/05/09 16:35:00 EDT
834    Tag: blume-20010509-cpscontract
835    Description:
836    
837    Fixed a bug in the accounting code in cpsopt/contract.sml.  (The
838    wrapper/unwrapper elimination did not decrement usage counts and some
839    dead variables got overlooked by the dead-up logic.)
840    
841    ----------------------------------------------------------------------
842    Name: Lal George
843    Date: 2001/05/08  17:26:09 EDT
844    Tag: george-20010508-omit-frameptr
845    Description:
846    
847    Changes to implement the omit-frame-pointer optimization to support
848    raw C calls. For now, there is only support on the Intel x86, but
849    other architectures will follow as more experience is gained with this.
850    
851    
852    ----------------------------------------------------------------------
853    Name: Matthias Blume
854    Date: 2001/05/07 14:40:00 EDT
855    Tag: blume-20010507-proxies
856    Description:
857    
858    I made into "proxy libraries" all libraries that qualify for such a
859    change.  (A qualifying library is a library that has another library or
860    groups as its sole member and repeats that member's export list
861    verbatim.  A proxy library avoids this repetition by omitting its export
862    list, effectively inheriting the list that its (only) member exports.
863    See the CM manual for more explanation.)
864    The main effect is that explicit export lists for these libraries
865    do not have to be kepts in sync, making maintenance a bit easier.
866    
867    I also added copyright notices to many .cm-files.
868    
869    Last but not least, I made a new set of bootfiles.
870    
871    ----------------------------------------------------------------------
872    Name: Matthias Blume
873    Date: 2001/05/04 17:00:00 EDT
874    Tag: blume-20010504-cm-lsplit
875    Description:
876    
877    0. John merged pending changes to $/smlnj-lib.cm
878    
879    1. Allen's previous change accidentally backed out of one of Lal's
880       earlier changes.  I undid this mistake (re-introducing Lal's change).
881    
882    2. I used the new topOrder' function from graph-scc.sml (from $/smlnj-lib.cm)
883       within the compiler where applicable.  There is some code simplification
884       because of that.
885    
886    3. The "split" phase (in FLINT) is now part of the default list of phases.
887       Compiler.Control.LambdaSplitting.* can be used to globally control the
888       lambda-splitting (cross-module-inlining) engine.  In addition to that,
889       it can now also be controlled on a per-source basis: CM has been taught
890       a new tool parameter applicable to ML source files.
891    
892       - To turn lambda-splitting off completely:
893            local open Compiler.Control.LambdaSplitting in
894                val _ = set Off
895            end
896       - To make "no lambda-splitting" the global default (but allow per-source
897         overriding); this is the initial setting:
898            local open Compiler.Control.LambdaSplitting in
899                val _ = set (Default NONE)
900            end
901       - To make "lambda-splitting with aggressiveness a" the global default
902         (and allow per-source overriding):
903            local open Compiler.Control.LambdaSplitting in
904                val _ = set (Default (SOME a))
905            end
906    
907       - To turn lambda-splitting off for a given ML souce file (say: a.sml)
908         write (in the respective .cm-file):
909            a.sml (lambdasplitting:off)
910       - To turn lambda-splitting for a.sml on with minimal aggressiveness:
911            a.sml (lambdasplitting:on)
912       - To turn lambda-splitting for a.sml on with aggressiveness <a> (where
913         <a> is a decimal non-negative integer):
914            a.sml (lambdasplitting:<a>)
915       - To turn lambda-splitting for a.sml on with maximal aggressiveness:
916            a.sml (lambdasplitting:infinity)
917       - To use the global default for a.sml:
918            a.sml (lambdasplitting:default)
919         or simply
920            a.sml
921    
922    ----------------------------------------------------------------------
923    Name: Allen Leung
924    Date: 2001/05/04 01:57:00 EDT
925    Tag: leunga-20010504-sync
926    Description:
927    
928      MLRISC features.
929    
930      1. Fix to CMPXCHG instructions.
931      2. Changed RA interface to allow annotations in callbacks.
932      3. Added a new method to the stream interface to allow annotations updates.
933    
934    ----------------------------------------------------------------------
935    Name: Matthias Blume
936    Date: 2001/05/01 11:45:00 EDT
937    Tag: blume-20010501-pcedittmp
938    Description:
939    
940    Changed install.sh to use the current working directory instead of
941    /usr/tmp for a temporary file (pcedittmp).  The previous choice
942    of /usr/tmp caused trouble with MacOS X because of file premission
943    problems.
944    
945    ----------------------------------------------------------------------
946    Name: Matthias Blume
947    Date: 2001/04/20 11:10:00 EDT
948    Tag: blume-20010420-inMLflag
949    Description:
950    
951     - added vp_limitPtrMask to vproc-state.h
952       (for use by the raw-C-calls mechanism to implement proper interrupt
953        handling)
954     - made the ML compiler aware of various data-structure offsets so it
955       can generate code for accessing the vp_inML flag and vp_limitPtrMask
956     - tweaked mlriscGen.sml to have it emit interrupt-handling code for
957       raw C-calls
958    
959    ----------------------------------------------------------------------
960    Name: Lal George
961    Date: 2001/04/20 09:15:28 EDT
962    Tag: george-20010420-macosX
963    Description:
964    
965     - Changes to port to Mac OS X; Darwin.
966    
967     - In the process I found that sqrt was broken on the PPC, because the
968       fsqrt instruction is not implemented.
969    
970    ----------------------------------------------------------------------
971    Name: Matthias Blume
972    Date: 2001/04/18 12:45:00 EDT
973    Tag: blume-20010418-ccalls
974    Description:
975    
976     - fixed two off-by-4 errors in the x86-specific c-calls implementation
977       (this bug prevented structure arguments containing pointers from being
978        passed correctly)
979     - changed the raw-C-call code in mlriscGen.sml in such a way that
980       structure arguments are represented as a pointer to the beginning
981       of the structure (instead of having a series of synthesized arguments,
982       one for each structure member)
983    
984     - made makeml script's verbosity level configurable via environment
985       variable (MAKEML_VERBOSITY)
986    
987     - eliminated placeholder implementations for f32l, w16s, i16s, and f32s
988       in rawmem-x86.sml; we are now using the real thing
989    
990    ----------------------------------------------------------------------
991    Name: Matthias Blume
992    Date: 2001/03/22 16:25:00 EST
993    Tag: blume-20010322-bootfiles
994    Description:
995    
996    Created a new set of bootfiles (for your automatic installation convenience).
997    
998    ----------------------------------------------------------------------
999    Name: Matthias Blume
1000    Date: 2001/03/22 15:10:00 EST
1001    Tag: blume-20010322-rawmem-parcm
1002    Description:
1003    
1004    1. All "raw memory access" primitives for the new FFI are implemented now
1005       (at least on the x86).
1006    2. Some further cleanup of CM's parallel make mechanism.
1007    
1008    ----------------------------------------------------------------------
1009    Name: Matthias Blume
1010    Date: 2001/03/19 17:53:00 EST
1011    Tag: blume-20010319-parallel
1012    Description:
1013    
1014    Parallel make (using compile servers) now works again.
1015    
1016    To this end, CM.stabilize and CMB.make have been modified to work in
1017    two passes when compile servers are attached:
1018       1. Compile everything, do not perform stabilization; this pass
1019          uses compile servers
1020       2. Stabilize everything; this pass does not use compile servers
1021    If there are no compile servers, the two passes are combined into one
1022    (as before).  Splitting the passes increases the inherent parallelism
1023    in the dependency graph because the entire graph including all
1024    libraries is available at the same time.  This, in turn, improves
1025    server utilization.  The downside is that the master process will
1026    have to do some extra work after compilation is done (because for
1027    technical reasons it must re-read all the binfiles during stabilization).
1028    
1029    ----------------------------------------------------------------------
1030    Name: Matthias Blume
1031    Date: 2001/03/16 12:22:00 EST
1032    Tag: blume-20010316-bootfiles
1033    Description:
1034    
1035    Created a new set of bootfiles (for your automatic installation convenience).
1036    
1037    ----------------------------------------------------------------------
1038    Name: Matthias Blume
1039    Date: 2001/03/16 11:00:00 EST
1040    Tag: blume-20010316-MLTREE-fixup
1041    Description:
1042    
1043    This is a minor fixup for an (untagged) earlier commit by Allen.
1044    (A file was missing).
1045    
1046    ----------------------------------------------------------------------
1047    Name: Allen Leung
1048    Date: Mon Mar  5 18:54:57 EST 2001
1049    Tag: leunga-20010305-cut-support
1050    
1051    1. New support for alternative control-flow in MLTREE.
1052       Currently we support
1053    
1054          FLOW_TO(CALL ...., [k1,...,kn])
1055    
1056       This is needed for 'cuts to' in C-- and try/handle-like constructs
1057       in Moby
1058    
1059       New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
1060    
1061    2. Register Allocator
1062    
1063       Changes in interface [from Fermin, John]
1064    
1065    3. Alpha 8-bit SLL support [Fermin]
1066    
1067    4. All architectures
1068    
1069       A new module (ClusterExpandCopies) for expanding parallel copies.
1070    
1071    ----------------------------------------------------------------------
1072    Name: Allen Leung
1073    Date: 2001/02/27 23:07:00 EST
1074    Tag: leunga-20010227-minor-stuff
1075    
1076    1. Alpha bug fix for CMOVNE
1077    2. Handle mltree COND(..,FCMP ...,...)
1078    3. Bug fix in simplifier
1079    
1080    ----------------------------------------------------------------------
1081    Name: Matthias Blume
1082    Date: 2001/01/30 17:50:00 EST
1083    Tag: blume-20010130-sync
1084    Description:
1085    
1086    This is just a minor update to sync my devel branch with the main brach.
1087    The only visible change is the addition of some README files.
1088    
1089    ----------------------------------------------------------------------
1090    Name: Matthias Blume
1091    Date: 2001/01/12 23:30:00 JST
1092    Tag: blume-20010112-bootfiles
1093    Description:
1094    
1095    Made a new set of bootfiles that goes with the current state of the
1096    repository.
1097    
1098    ----------------------------------------------------------------------
1099    Name: Matthias Blume
1100    Date: 2001/01/12 21:20:00 JST
1101    Tag: blume-20010112-sync
1102    Description:
1103    
1104    I am just flushing out some minor changes that had accumulated in
1105    my private branch in order to sync with the main tree.  (This is
1106    mainly because I had CVS trouble when trying to merge _into_ my
1107    private branch.)
1108    
1109    Most people should be completely unaffected by this.
1110    
1111    ----------------------------------------------------------------------
1112    Name: Allen Leung
1113    Date: Thu Jan 11 21:03:00 EST 2001
1114    Tag: leunga-20010111-labexp=mltree
1115    Description:
1116    
1117    1.  Removed the type LabelExp and replace it by MLTree.
1118    2.  Rewritten mltree-simplify with the pattern matcher tool.
1119    3.  There were some bugs in alpha code generator which would break
1120        64-bit code generation.
1121    4.  Redo the tools to generate code with the
1122    5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
1123        are now generated by perl scripts.
1124    
1125    ----------------------------------------------------------------------
1126    Name: Matthias Blume
1127    Date: 2001/01/10 21:55:00 JST
1128    Tag: blume-20010110-rcc
1129    Description:
1130    
1131    The RCC stuff now seems to work (but only on the x86).
1132    This required hacking of the c-calls interface (and -implementation) in
1133    MLRISC.
1134    
1135    Normal compiler users should be unaffected.
1136    
1137    ----------------------------------------------------------------------
1138    Name: Matthias Blume
1139    Date: 2001/01/09 01:20:00 JST
1140    Tag: blume-20010109-rcc
1141    Description:
1142    
1143    This is a fairly big patch, flushing out a large number of pending
1144    changes that I made to my development copy over the last couple of days.
1145    
1146    Of practical relevance at this moment is a workaround for a pickling
1147    bug that Allen ran into the other day.  The cause of the bug itself is
1148    still unknown and it might be hard to fix it properly, but the
1149    workaround has some merits of its own (namely somewhat reducing pickling
1150    overhead for certain libraries).  Therefore, I think this solution should
1151    be satisfactory at this time.
1152    
1153    The rest of the changes (i.e., the vast majority) has to do with my
1154    ongoing efforts of providing direct support for C function calls from
1155    ML.  At the moment there is a new primop "RAW_CCALL", typing magic
1156    in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
1157    case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
1158    translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
1159    handle RCC.
1160    
1161    The last part (the changes to mlriscGen.sml) are still known to be
1162    wrong on the x86 and not implemented on all other architectures.  But
1163    the infrastructure is in place. I had to change a few functor
1164    signatures in the backend to be able to route the CCalls interface
1165    from MLRISC there, and I had to specialize the mltree type (on the
1166    x86) to include the necessary extensions. (The extensions themselves
1167    were already there and redy to go in MLRISC/x86).
1168    
1169    Everything should be very happy as soon as someone helps me with
1170    mlriscGen.sml...
1171    
1172    In any case, nothing of this should matter to anyone as long as the
1173    new primop is not being used (which is going to be the case unless you
1174    find it where I hid it :). The rest of the compiler is completely
1175    unaffected.
1176    
1177    ----------------------------------------------------------------------
1178    Name: Matthias Blume
1179    Date: 2001/01/05 00:30:00 JST
1180    Tag: blume-20010105-primops
1181    Description:
1182    
1183    Added some experimental support for work that I am doing right now.
1184    These changes mostly concern added primops, but there is also a new
1185    experimental C library in the runtime system (but currently not enabled
1186    anywhere except on Linux/X86).
1187    
1188    In the course of adding primops (and playing with them), I discovered that
1189    Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
1190    broken.  (Zhong was very right he labeled this stuff as "major gross hack".)
1191    To recover, I made type information in INL_PRIM mandatory and changed
1192    prim.sml as well as built-in.sml accordingly.  The InLine structure now
1193    has complete, correct type information (i.e., no bottom types).
1194    
1195    Since all these changes mean that we need new binfiles, I also bumped the
1196    version number to 110.32.1.
1197    
1198    ----------------------------------------------------------------------
1199    Name: Matthias Blume
1200    Date: 2000/12/30 22:10:00 JST
1201    Tag: blume-20001230-various
1202    Description:
1203    
1204    Added proxy libraries for MLRISC and let MLRISC libraries refer
1205    to each other using path anchors.  (See CM manual for explanation.)
1206    
1207    Updated CM documentation.
1208    
1209    Fixed some bugs in CM.
1210    
1211    Implemented "proxy" libraries (= syntactic sugar for CM).
1212    
1213    Added "-quiet" option to makeml and changed runtime system accordingly.
1214    
1215    Added cleanup handler for exportML to reset timers and compiler stats.
1216    
1217    ----------------------------------------------------------------------
1218    Name: Lal George
1219    Date: 2000/12/22 22:22:58 EST 2000
1220    Tag: Release_110_32
1221    Description:
1222    
1223            Infinite precision used throughout MLRISC.
1224            see MLRISC/mltree/machine-int.sig
1225    
1226    ----------------------------------------------------------------------
1227    Name: Matthias Blume
1228    Date: 2000/12/22 23:16:00 JST
1229    Tag: blume-20001222-warn
1230    Description:
1231    
1232    Corrected wording and formatting of some CM warning message which I
1233    broke in my previous patch.
1234    
1235    ----------------------------------------------------------------------
1236    Name: Matthias Blume
1237    Date: 2000/12/22 21:20:00 JST
1238    Tag: blume-20001222-anchorenv
1239    Description:
1240    
1241    Fixed CM's handling of anchor environments in connection with CMB.make.
1242    
1243    ----------------------------------------------------------------------
1244    Name: Matthias Blume
1245    Date: 2000/12/22 13:15:00 JST
1246    Tag: blume-20001222-cleanup
1247    Description:
1248    
1249    Removed src/cm/ffi which does not (and did not) belong here.
1250    
1251    ----------------------------------------------------------------------
1252    Name: Matthias Blume
1253    Date: 2000/12/21 23:55:00 JST
1254    Tag: blume-20001221-exn
1255    Description:
1256    
1257    Probably most important: CM no longer silently swallows all exceptions
1258    in the compiler.
1259    Plus: some other minor CM changes.  For example, CM now reports some
1260    sizes for generated binfiles (code, data, envpickle, lambdapickle).
1261    
1262    ----------------------------------------------------------------------
1263    Name: Matthias Blume
1264    Date: 2000/12/15 00:01:05 JST
1265    Tag: blume-20001215-dirtool
1266    Description:
1267    
1268    - "dir" tool added.
1269    - improvements and cleanup to Tools structure
1270    - documentation updates
1271    
1272    ----------------------------------------------------------------------
1273    Name: Allen Leung
1274    Date: Thu Dec 14 03:45:24 EST 2000
1275    Description:
1276    Tag:  leunga-20001214-int-inf
1277    Description:
1278    
1279       In IntInf, added these standard functions, which are missing from our
1280    implementation:
1281    
1282        andb : int * int -> int
1283        xorb : int * int -> int
1284        orb  : int * int -> int
1285        notb : int -> int
1286         <<   : int * word -> int
1287        ~>>  : int * word -> int
1288    
1289       Not tested, I hope they are correct.
1290    
1291    ----------------------------------------------------------------------
1292    Name: Allen Leung
1293    Date: Fri Dec  8 19:23:26 EST 2000
1294    Description:
1295    Tag:  leunga-20001208-nowhere
1296    Description:
1297    
1298      Slight improvements to the 'nowhere' tool to handle OR-patterns,
1299    to generate better error messages etc.  Plus a brief manual.
1300    
1301    ----------------------------------------------------------------------
1302    Name: Lal George
1303    Date: 2000/12/08 09:54:02 EST 2000
1304    Tag: Release_110_31
1305    Description:
1306    
1307    - Version 110.31
1308    ----------------------------------------------------------------------
1309    Name: Allen Leung
1310    Date: Thu Dec  7 22:01:04 EST 2000
1311    Tag:  leunga-20001207-cell-monster-hack
1312    Description:
1313    
1314    Major MLRISC internal changes.  Affect all clients.
1315    Summary:
1316    
1317    1.  Type CELLS.cell = int is now replaced by a datatype.
1318        As a result, the old regmap is now gone.  Almost all interfaces
1319        in MLRISC change as a consequence.
1320    
1321    2.  A new brand version of machine description tool (v3.0) that generates
1322        modules expecting the new interface.  The old version is removed.
1323    
1324    3.  The RA interface has been further abstracted into two new functors.
1325        RISC_RA and X86RA.  These functors have much simpler interfaces.
1326        [See also directory MLRISC/demo.]
1327    
1328    4.  Some other new source->source code generation tools are available:
1329    
1330        a. MLRISC/Tools/RewriteGen -- generate rewriters from rules.
1331        b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules.
1332           I use this tool to generate the peephole optimizers---with the new
1333           cell type changes, peephole rules are becoming difficult to write
1334           without conditional pattern matching.
1335    
1336    5.  More Intmap -> IntHashTable change.  Previous changes by Matthias didn't
1337        cover the entire MLRISC source tree so many things broke.
1338    
1339    6.  CM files have been moved to the subdirectory MLRISC/cm.
1340        They are moved because there are a lot of them and they clutter up the
1341        root dir.
1342    
1343    7.  More detailed documentation to come...
1344    
1345        NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do
1346        a makeml -rebuild first.  This is because of other other
1347        changes that Matthias has made (see below).
1348    
1349    
1350    ----------------------------------------------------------------------
1351    Name: Matthias Blume
1352    Date: 2000/11/30 23:12:00 JST
1353    Tag: blume-20001130-filereorg
1354    Description:
1355    
1356    Some manual updates and some file reorganizations in CM.
1357    
1358    ----------------------------------------------------------------------
1359    Name: Matthias Blume
1360    Date: 2000/11/24 17:45:00 JST
1361    Tag: blume-20001124-link
1362    Description:
1363    
1364    Drastically improved link traversal code for the case that the dynamic
1365    value was already loaded at bootstrap time.  As a result, CM and CMB
1366    now both load blazingly fast -- even on a very slow machine.  Also,
1367    memory consumption has been further reduced by this.
1368    
1369    Warning: The format of the PIDMAP file has changed.  THerefore, to
1370    bootstrap you have to do this:
1371    
1372    1. Run CMB.make
1373    2. Make a symbolic link for the boot directory:
1374         ln -s sml.boot.ARCH-OS xxx
1375    3. "Rebuild" the boot directory:
1376         ./makeml -boot xxx -rebuild sml ; rm xxx
1377    4. Boot normally:
1378          ./makeml
1379    
1380    ----------------------------------------------------------------------
1381    Name: Matthias Blume
1382    Date: 2000/11/21 21:20:00 JST
1383    Tag: blume-20001121-tools
1384    Description:
1385    
1386    Continued hacking on autoloading problem -- with success this time.
1387    Also changed tool-plugin mechanism.  See new CM manual.
1388    
1389    ----------------------------------------------------------------------
1390    Name: Matthias Blume
1391    Date: 2000/11/19 14:30:00 JST
1392    Tag:  blume-20001119-autoload
1393    Description:
1394    
1395    Some hacking to make autoloading faster.  Success for CMB, no success
1396    so far for CM.  There is a reduced structure CM' that autoloads faster.
1397    (This is a temporary, non-documented hack to be eliminated again when
1398    the general problem is solved.)
1399    
1400    ----------------------------------------------------------------------
1401    Name: Matthias Blume
1402    Date: 2000/11/17 14:10:00 JST
1403    Tag: blume-20001117-pickle-lib
1404    Description:
1405    
1406    1. Eliminated comp-lib.cm
1407    2. Made pickle-lib.cm
1408    3. Eliminated all uses of intset.sml (from comp-lib.cm)
1409    4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
1410       equivalent constructs from smlnj-lib.cm (INtHashTable).
1411    5. Point 4. also goes for those uses of intmap.* in MLRISC.
1412       Duplicated intmap modules thrown out.
1413    6. Hunted down all duplicated SCC code and replaced it with
1414       equivalent stuff (GraphSCCFn from smlnj-lib.cm).
1415    7. Rewrote Feedback module.
1416    8. Moved sortedlist.sml into viscomp-lib.cm.  Eventually it
1417       should be thrown out and equivalent modules from smlnj-lib.cm
1418       should be used (IntRedBlackSet, IntListSet, ...).
1419    
1420    Confirmed that compiler compiles to fixpoint.
1421    
1422    ----------------------------------------------------------------------
1423    Name: Allen Leung
1424    Date: 2000/11/10 18:00:00
1425    Tag: leunga-20001110-new-x86-fp
1426    
1427    A new x86 floating point code generator has been added.
1428    By default this is turned off.  To turn this on, do:
1429    
1430        CM.autoload "$smlnj/compiler.cm";
1431        Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
1432    
1433    Changes:
1434    
1435    1.  Changed FTAN to FPTAN so that the assembly output is correct.
1436    2.  Changed the extension callback for FTANGENT to generate:
1437    
1438              fptan
1439              fstp  %st(0)
1440        instead of
1441              fptan
1442              fstpl ftempmem
1443    
1444    3.  Numerous assembly fixes for x86.
1445    
1446    5.  Cleaned up the machine code output module x86/x86MC.sml and added
1447        support for a whole bunch of instructions and addressing modes:
1448    
1449          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st, %st(n)
1450          faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp  %st, %st(n)
1451          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st(n), %st
1452          fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
1453          fxch %st(n)
1454          fld %st(n)
1455          fst %st(n)
1456          fst mem
1457          fstp %st(n)
1458          fucom %st(n)
1459          fucomp %st(n)
1460    
1461        All these are now generated when the fast fp mode is turned on.
1462    
1463    6.  Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
1464    
1465    ----------------------------------------------------------------------
1466    Name: Matthias Blume
1467    Date: 2000/11/09 11:20:00 JST
1468    Tag: blume-20001109-scc
1469    Description:
1470    
1471    Eliminated some code duplication:
1472    
1473    1. Added "where" clause to GraphSCCFn in SML/NJ Library.
1474       (Otherwise the functor is useless.)
1475    2. Used GraphSCCFn where SCCUtilFun was used previously.
1476    3. Got rid of SCCUtilFun (in comp-lib.cm).
1477    
1478    ----------------------------------------------------------------------
1479    Name: Lal George
1480    Date: 2000/11/06 09:02:21 EST 2000
1481    Tag: Release_110_30
1482    Description:
1483    
1484    - Version 110.30
1485    ----------------------------------------------------------------------
1486    Name: Matthias Blume
1487    Date: 2000/11/04 14:45:00
1488    Tag: blume-20001104-mlbuild
1489    Description:
1490    
1491    - Made ml-build faster on startup.
1492    - Documentation fixes.
1493    
1494    ----------------------------------------------------------------------
1495    Name: Matthias Blume
1496    Date: 2000/11/02 17:00:00 JST
1497    Tag: blume-20001102-condcomp
1498    Description:
1499    
1500    - Small tweaks to pickler -- new BOOTFILES!
1501    - Version bumped to 110.29.2.
1502    - Added conditional compilation facility to init.cmi (see comment there).
1503    ----------------------------------------------------------------------
1504    Name: Allen Leung
1505    Date: 2000/10/23 19:31:00
1506    Tag: leunga-20001023-demo-ra
1507    
1508    1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
1509    2. Test programs for the graph library updated
1510    3. Some new MLRISC demo programs added
1511    
1512    ----------------------------------------------------------------------
1513    Name: Matthias Blume
1514    Date: 2000/08/31 22:15:00 JST
1515    Tag: blume-20001017-errmsg
1516    Description:
1517    
1518    More error message grief: Where there used to be no messages, there
1519    now were some that had bogus error regions.  Fixed.
1520    
1521    ----------------------------------------------------------------------
1522    Name: Matthias Blume
1523    Date: 2000/08/31 17:30:00 JST
1524    Tag: blume-20001017-v110p29p1
1525    Description:
1526    
1527    I made a version 110.29.1 with new bootfiles.
1528    
1529    Changes:  Modified pickler/unpickler for faster and leaner unpickling.
1530              CM documentation changes and a small bugfix in CM's error reporting.
1531    
1532    ----------------------------------------------------------------------
1533  Name: Lal George  Name: Lal George
1534  Date: 2000/09/27 14:42:35 EDT  Date: 2000/09/27 14:42:35 EDT
1535  Tag: george-20000927-nodestatus  Tag: george-20000927-nodestatus
# Line 1500  Line 3017 
3017       elaborator). There were a lot of changes during my "linkpath" trials       elaborator). There were a lot of changes during my "linkpath" trials
3018       that could have been reverted to their original state but weren't.       that could have been reverted to their original state but weren't.
3019       Please, don't be too harsh on me for messing with this code a bit more       Please, don't be too harsh on me for messing with this code a bit more
3020       than what was strictly necessary...  (I _did_ resist the tempation       than what was strictly necessary...  (I _did_ resist the temptation
3021       of doing any "global reformatting" to avoid an untimely death at       of doing any "global reformatting" to avoid an untimely death at
3022       Dave's hands. :)       Dave's hands. :)
3023    

Legend:
Removed from v.705  
changed lines
  Added in v.938

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