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 826, Wed May 9 20:32:17 2001 UTC revision 986, Wed Nov 21 21:03:17 2001 UTC
# Line 14  Line 14 
14    
15  ----------------------------------------------------------------------  ----------------------------------------------------------------------
16  Name: Matthias Blume  Name: Matthias Blume
17    Date: 2001/11/21 16:10:00 EST
18    Tag: Release_110_37
19    Description:
20    
21    1. Modifications to the SML/NJ code generator and to the runtime system
22       so that code object name strings are directly inserted into code
23       objects at code generation time.  The only business the runtime system
24       has with this is now to read the name strings on occasions.
25       (The encoding of the name string has also changed somewhat.)
26    
27    2. CM now implements a simple "set calculus" for specifying export lists.
28       In particular, it is now possible to refer to the export lists of
29       other libraries/groups/sources and form unions as well as differences.
30       See the latest CM manual for details.
31    
32    3. An separate notion of "proxy" libraries has again be eliminated from
33       CM's model.  (Proxy libraries are now simply a special case of using
34       the export list calculus.)
35    
36    4. Some of the existing libraries now take advantage of the new set
37       calculus.
38       (Notice that not all libraries have been converted because some
39       of the existing .cm-files are supposed to be backward compatible
40       with 110.0.x.)
41    
42    5. Some cleanup in stand-alone programs.  (Don't use "exnMessage" -- use
43       "General.exnMessage"!  The former relies on a certain hook to be
44       initialized, and that often does not happen in the stand-alone case.)
45    
46    ----------------------------------------------------------------------
47    Name: Lal George
48    Date: 2001/11/21  13:56:18 EST
49    Tag: george-2001121-pseudo-ops
50    Description:
51    
52      Implemented a complete redesign of MLRISC pseudo-ops. Now there
53      ought to never be any question of incompatabilities with
54      pseudo-op syntax expected by host assemblers.
55    
56      For now, only modules supporting GAS syntax are implemented
57      but more should follow, such as MASM, and vendor assembler
58      syntax, e.g. IBM as, Sun as, etc.
59    
60    ----------------------------------------------------------------------
61    Name: Matthias Blume
62    Date: 2001/11/14 11:52:00 EST
63    Tag: blume-20011114-srcname
64    Description:
65    
66    1. Routed the name of the current source file to mlriscgen where it
67       should be directly emitted into the code object.  (This last part
68       is yet to be done.)
69    
70    2. Some cleanup of the pgraph code to make it match the proposal that
71       I put out the other day.  (The proposal notwithstanding, things are
72       still in flux here.)
73    
74    ----------------------------------------------------------------------
75    Name: Lal George
76    Date: 2001/11/14 09:44:04 EST
77    Tag:
78    Description:
79    
80      Fix for a backpatching bug reported by Allen.
81    
82      Because the boundary between short and long span-dependent
83      instructions is +/- 128, there are an astounding number of
84      span-dependent instructions whose size is over estimated.
85    
86      Allen came up with the idea of letting the size of span
87      dependent instructions be non-monotonic, for a maxIter
88      number of times, after which the size must be monotonically
89      increasing.
90    
91      This table shows the number of span-dependent instructions
92      whose size was over-estimated as a function of maxIter, for the
93      file Parse/parse/ml.grm.sml:
94    
95         maxIter            # of instructions:
96            10                      687
97            20                      438
98            30                      198
99            40                        0
100    
101      In compiling the compiler, there is no significant difference in
102      compilation speed between maxIter=10 and maxIter=40. Actually,
103      my measurements showed that maxIter=40 was a tad faster than
104      maxIter=10! Also 96% of the  files in the compiler reach a fix
105      point within 13 iterations, so fixing maxIter at 40, while high,
106      is okay.
107    
108    ----------------------------------------------------------------------
109    Name: Matthias Blume
110    Date: 2001/10/31 15:25:00 EST
111    Tag: blume-20011031-pgraph
112    Description:
113    
114    CKIT:
115    * Changed the "Function" constructor of type Ast.ctype to carry optional
116      argument identifiers.
117    * Changed the return type of TypeUtil.getFunction accordingly.
118    * Type equality ignores the argument names.
119    * TypeUtil.composite tries to preserve argument names but gives up quickly
120      if there is a mismatch.
121    
122    installation script:
123    * attempts to use "curl" if available (unless "wget" is available as well)
124    
125    CM:
126    * has an experimental implementation of "portable graphs" which I will
127      soon propose as an implementation-independent library format
128    * there are also new libraries $/pgraph.cm and $/pgraph-util.cm
129    
130    NLFFI-LIB:
131    * some cleanup (all cosmetic)
132    
133    NLFFIGEN:
134    * temporarily disabled the mechanism that suppresses ML output for
135      C definitions whose identifiers start with an underscore character
136    * generate val bindings for enum constants
137    * user can request that only one style (light or heavy) is being used;
138      default is to use both (command-line arguments: -heavy and -light)
139    * fixed bug in handling of function types involving incomplete pointers
140    * generate ML entry points that take record arguments (i.e., using
141      named arguments) for C functions that have a prototype with named
142      arguments
143      (see changes to CKIT)
144    
145    ----------------------------------------------------------------------
146    Name: Allen Leung
147    Date: 2001/10/27 20:34:00 EDT
148    Tag: leunga-20011027-x86-fast-fp-call
149    Description:
150    
151       Fixed the bug described in blume-20010920-slowfp.
152    
153       The fix involves
154          1. generating FCOPYs in FSTP in ia32-svid
155          2. marking a CALL with the appropriate annotation
156    
157    ----------------------------------------------------------------------
158    Name: Matthias Blume
159    Date: 2001/10/16 11:32:00 EDT
160    Tag: blume-20011016-netbsd
161    Description:
162    
163    Underscore patch from Chris Richards (fixing problem with compiling
164    runtime system under recent NetBSD).
165    
166    ----------------------------------------------------------------------
167    Name: Allen Leung
168    Date: 2001/10/12 17:18:32 EDT 2001
169    Tag: leung-20011012-x86-printflowgraph
170    Description:
171    
172    X86RA now uses a valid (instead of dummy) PrintFlowgraph module.
173    
174    ----------------------------------------------------------------------
175    Name: Lal George
176    Date: 2001/10/11 23:51:34 EDT
177    Tag: george-20011011-too-many-instrs
178    Description:
179    
180    The representation of a program point never expected to see more
181    than 65536 instructions in a basic block!
182    
183    ----------------------------------------------------------------------
184    Name: Lal George
185    Date: 2001/10/09 09:41:37 EDT
186    Tag: george-20011008-mlrisc-labels
187    Description:
188    
189    Changed the machine description files to support printing of
190    local and global labels in assembly code, based on host assembler
191    conventions.
192    
193    ----------------------------------------------------------------------
194    Name: Matthias Blume
195    Date: 2001/09/25 15:25:00 EDT
196    Tag: blume-20010925-exninfo
197    Description:
198    
199    I provided a non-hook implementation of exnName (at the toplevel) and
200    made the "dummy" implementation of exnMessage (at the toplevel) more
201    useful: if nothing gets "hooked in", then at least you are going to
202    see the exception name and a message indicating why you don't see more.
203    
204    [For the time being, programs that need exnMessage and want to use
205    ml-build should either use General.exnMessage (strongly recommended) or
206    refer to structure General at some other point so that CM sees a
207    static dependency.]
208    
209    [Similar remarks go for "print" and "use":  If you want to use their
210    functionality in stand-alone programs generated by ml-build, then use
211    TextIO.output and Backend.Interact.useFile (from $smlnj/compiler.cm).]
212    
213    ----------------------------------------------------------------------
214    Name: Matthias Blume
215    Date: 2001/09/20 17:28:00 EDT
216    Tag: blume-20010920-slowfp
217    Description:
218    
219    Allen says that x86-fast-fp is not safe yet, so I turned it off again...
220    
221    ----------------------------------------------------------------------
222    Name: Matthias Blume
223    Date: 2001/09/20 17:20:00 EDT
224    Tag: blume-20010920-canonicalpaths
225    Description:
226    
227    0. Updated the BOOT file (something that I forgot to do earlier).
228    
229    1. Small internal change to CM so that it avoids "/../" in filenames
230       as much as possible (but only where it is safe).
231    
232    2. Changed config/_run-sml (resulting in a changed bin/.run-sml) so
233       that arguments that contain delimiters are passed through correctly.
234       This change also means that all "special" arguments of the form
235       @SMLxxx... must come first.
236    
237    3. Changed install script to put relative anchor names for tool commands
238       into pathconfig.
239    
240    ----------------------------------------------------------------------
241    Name: Matthias Blume
242    >>>>>>> 1.169
243    Date: 2001/09/18 15:35:00 EDT
244    Tag: blume-20010918-readme11036
245    Description:
246    
247    Added README files.
248    
249    ----------------------------------------------------------------------
250    Name: Matthias Blume
251    Date: 2001/09/18 11:45:00 EDT
252    Tag: Release_110_36 (retag)
253    Description:
254    
255    Fixed mistake in config/preloads. Retagged as 110.36.
256    
257    ----------------------------------------------------------------------
258    Name: Matthias Blume
259    Date: 2001/09/18 09:40:00 EDT
260    Tag: Release_110_36_orig (tag changed)
261    Description:
262    
263    New version (110.36).  New bootfiles.
264    
265    ----------------------------------------------------------------------
266    Name: Matthias Blume
267    Date: 2001/09/14 16:15:00 EDT
268    Tag: blume-20010914-x86fastfp
269    Description:
270    
271    John committed some changes that Allen made, in particular a (hopefully)
272    correctly working version of the x86-fp module.
273    
274    I changed the default setting of the Control.MLRISC.getFlag "x86-fast-fp"
275    flag to "true".  Everything seems to compile to a fixpoint ok, and
276    "mandelbrot" speeds up by about 15%.
277    
278    ----------------------------------------------------------------------
279    Name: Matthias Blume
280    Date: 2001/09/13 11:20:00 EDT
281    Tag: blume-20010913-minimal
282    Description:
283    
284    1. Stefan Monnier's patch to fix a miscompilation problem that
285       was brought to light by John Reppy's work on Moby.
286    
287    2. Implemented a minimal "structure Compiler" that contains just
288       "version" and "architecture".  The minimal version will be
289       available when the full version is not.  This is for backward-
290       compatibility with code that wants to test Compiler.version.
291    
292    ----------------------------------------------------------------------
293    Name: Matthias Blume
294    Date: 2001/08/28 14:03:00 EDT
295    Tag: blume-20010828-ml-lex
296    Description:
297    
298    Fix for bug 1581, received from Neophytos Michael.
299    
300    ----------------------------------------------------------------------
301    Name: Matthias Blume
302    Date: 2001/08/27 11:20:00 EDT
303    Tag: blume-20010827-readme11035
304    Description:
305    
306    Fleshed out the README file for 110.35.
307    
308    ----------------------------------------------------------------------
309    Name: Matthias Blume
310    Date: 2001/08/24 17:10:00 EDT
311    Tag: Release_110_35
312    Description:
313    
314    New version number (110.35).  New bootfiles.
315    
316    ----------------------------------------------------------------------
317    Name: Lal George
318    Date: 2001/08/24 13:47:18 EDT 2001
319    Tag: george-20010824-MLRISC-graphs
320    Description:
321    
322     removed clusters from MLRISC completely and replaced with graphs.
323    
324    ----------------------------------------------------------------------
325    Name: Matthias Blume
326    Date: 2001/08/23 17:50:00 EDT
327    Tag: blume-20010823-toplevel
328    Description:
329    
330    - some reorganization of the code that implements various kinds of
331      environments in the compiler (static, dynamic, symbolic, combined)
332    - re-implemented the EnvRef module so that evalStream works properly
333      (if the stream contains references to "use", "CM.make", etc.)
334    - cleaned up evalloop.sml and interact.sml (but they need more cleaning)
335    
336    ----------------------------------------------------------------------
337    Name: Matthias Blume
338    Date: 2001/08/20 15:50 EDT
339    Tag: blume20010820-slipup
340    Description:
341    
342    I forgot to commit a few files.  Here they are...
343    
344    ----------------------------------------------------------------------
345    Name: Matthias Blume
346    Date: 2001/08/20 15:35:00 EDT
347    Tag: blume-20010820-debugprof
348    Description:
349    
350    !!!! NEW BOOTFILES !!!!
351    
352    This is another round of reorganizing the compiler sources.  This
353    time the main goal was to factor out all the "instrumentation"
354    passes (for profiling and backtracing) into their own library.
355    The difficulty was to do it in such a way that it does not depend
356    on elaborate.cm but only on elabdata.cm.
357    
358    Therefore there have been further changes to both elaborate.cm and
359    elabdata.cm -- more "generic" things have been moved from the former
360    to the latter.  As a result, I was forced to split the assignment
361    of numbers indicating "primtyc"s into two portions: SML-generic and
362    SML/NJ-specific.  Since it would have been awkward to maintain,
363    I bit the bullet and actually _changed_ the mapping between these
364    numbers and primtycs.  The bottom line of this is that you need
365    a new set of bin- and bootfiles.
366    
367    I have built new bootfiles for all architectures, so doing a fresh
368    checkout and config/install.sh should be all you need.
369    
370    The newly created library's name is
371    
372        $smlnj/viscomp/debugprof.cm
373    
374    and its sources live under
375    
376        src/compiler/DebugProf
377    
378    ----------------------------------------------------------------------
379    Name: Matthias Blume
380    Date: 2001/08/15 17:15:00 EDT
381    Tag: blume-20010815-compreorg
382    Description:
383    
384    This is a first cut at reorganizing the CM libraries that make up the
385    core of the compiler.  The idea is to separate out pieces that could
386    be used independently by tools, e.g., the parser, the typechecker, etc.
387    
388    The current status is a step in this direction, but it is not quite
389    satisfactory yet.  Expect more changes in the future.
390    
391    Here is the current (new) organization...
392    
393        What used to be $smlnj/viscomp/core.cm is now divided into
394        six CM libraries:
395    
396             $smlnj/viscomp/basics.cm
397                           /parser.cm
398                           /elabdata.cm
399                           /elaborate.cm
400                           /execute.cm
401                           /core.cm
402    
403        The CM files for these libraries live under src/system/smlnj/viscomp.
404        All these libraries are proxy libraries that contain precisely
405        one CM library component.  Here are the locations of the components
406        (all within the src/compiler tree):
407    
408             Basics/basics.cm
409             Parse/parser.cm
410             ElabData/elabdata.cm
411             Elaborator/elaborate.cm
412             Execution/execute.cm
413             core.cm
414    
415         [This organization is the same that has been used already
416         for a while for the architecture-specific parts of the visible
417         compiler and for the old version of core.cm.]
418    
419         As you will notice, many source files have been moved from their
420         respective original locations to a new home in one of the above
421         subtrees.
422    
423         The division of labor between the new libraries is the following:
424    
425             basics.cm:
426                - Simple, basic definitions that pertain to many (or all) of
427                  the other libraries.
428             parser.cm:
429                - The SML parser, producing output of type Ast.dec.
430                - The type family for Ast is also defined and exported here.
431             elabdata.cm:
432                - The datatypes that describe input and output of the elaborator.
433                  This includes types, absyn, and static environments.
434             elaborator.cm:
435                - The SML/NJ type checker and elaborator.
436                  This maps an Ast.dec (with a given static environment) to
437                  an Absyn.dec (with a new static environment).
438                - This libraries implements certain modules that used to be
439                  structures as functors (to remove dependencies on FLINT).
440             execute.cm:
441                - Everything having to do with executing binary code objects.
442                - Dynamic environments.
443             core.cm:
444                - SML/NJ-specific instantiations of the elaborator and MLRISC.
445                - Top-level modules.
446                - FLINT (this should eventually become its own library)
447    
448    Notes:
449    
450    I am not 100% happy with the way I separated the elaborator (and its
451    data structures) from FLINT.  Two instances of the same problem:
452    
453        1. Data structures contain certain fields that carry FLINT-specific
454           information.  I hacked around this using exn and the property list
455           module from smlnj-lib.  But the fact that there are middle-end
456           specific fields around at all is a bit annoying.
457    
458        2. The elaborator calculates certain FLINT-related information.  I tried
459           to make this as abstract as I could using functorization, but, again,
460           the fact that the elaborator has to perform calculations on behalf
461           of the middle-end at all is not nice.
462    
463        3. Having to used exn and property lists is unfortunate because it
464           weakens type checking.  The other alternative (parameterizing
465           nearly *everything*) is not appealing, though.
466    
467    I removed the "rebinding =" warning hack because due to the new organization
468    it was awkward to maintain it.  As a result, the compiler now issues some of
469    these warnings when compiling init.cmi during bootstrap compilation. On
470    the plus side, you also get a warning when you do, for example:
471       val op = = Int32.+
472    which was not the case up to now.
473    
474    I placed "assign" and "deref" into the _Core structure so that the
475    code that deals with the "lazy" keyword can find them there.  This
476    removes the need for having access to the primitive environment
477    during elaboration.
478    
479    ----------------------------------------------------------------------
480    Name: Matthias Blume
481    Date: 2001/08/13
482    Tag: blume-20010813-closures
483    Description:
484    
485    This fix was sent to us by Zhong Shao.  It is supposed to improve the
486    performance of certain loops by avoiding needless closure allocation.
487    
488    ----------------------------------------------------------------------
489    Name: Lal George
490    Date: 2001/07/31 10:03:23 EDT 2001
491    Tag: george-20010731-x86-fmalloc
492    Description: Fixed bug in x86 calls
493    
494        There was a bug where call instructions would mysteriously
495        vanish. The call instruction had to be one that returned
496        a floating point value.
497    
498    ----------------------------------------------------------------------
499    Name: Lal George
500    Date: 2001/07/19 16:36:29 EDT 2001
501    Tag: george-20010719-simple-cells
502    Description:
503    
504    I have dramatically simplified the interface for CELLS in MLRISC.
505    
506    In summary, the cells interface is broken up into three parts:
507    
508      1. CellsBasis : CELLS_BASIS
509    
510            CellsBasis is a top level structure and common for all
511            architectures.  it contains the definitions of basic datatypes
512            and utility  functions over these types.
513    
514      2. functor Cells() : CELLS
515    
516            Cells generates an interface for CELLS that incorporates the
517            specific resources on the target architecture, such as the
518            presence of special register classes, their number and size,
519            and various useful substructures.
520    
521      3. <ARCH>CELLS
522    
523            e.g. SparcCells: SPARCCELLS
524    
525            <ARCH>CELLS usually contains additional bindings for special
526            registers  on the architecture, such as:
527    
528                    val r0 : cell           (* register zero *)
529                    val y : cell            (* Y register *)
530                    val psr : cell          (* processor status register *)
531                    ...
532    
533            The structure returned by applying the Cells functor is opened
534            in this interface.
535    
536    The main implication of all this is that the datatypes for cells is
537    split between CellsBasis and CELLS -- a fairly simple change for user
538    code.
539    
540    In the old scheme the CELLS interface had a definitional binding of
541    the form:
542    
543            signature CELLS = sig
544    
545               structure CellsBasis = CellsBasis
546    
547               ...
548    
549            end
550    
551    With all the sharing constraints that goes on in MLRISC, this old
552    design  quickly leads to errors such as:
553    
554            "structure definition spec inside of sharing ... "
555    
556    
557    and appears to require an unacceptable amount of sharing and where
558    constraint hackery.
559    
560    I think this error message (the interaction of definitional specs and
561    sharing) requires more explanation on our web page.
562    
563    ----------------------------------------------------------------------
564    Name: Matthias Blume
565    Date: 2001/07/19 15:00:00 EDT
566    Tag: blume-20010719-libreorg
567    Description:
568    
569    This update puts together a fairly extensive but straightforward change
570    to the way the libraries that implement the interactive system are
571    organized:
572    
573       The biggest change is the elimination of structure Compiler.  As a
574       replacement for this structure, there is now a CM library
575       (known as $smlnj/compiler.cm or $smlnj/compiler/current.cm)
576       that exports all the substructures of the original structure Compiler
577       directly.  So instead of saying Compiler.Foo.bar one now simply
578       says Foo.bar.  (The CM libraries actually export a collection of
579       structures that is richer than the collection of substructures of
580       structure Compiler.)
581    
582       To make the transition smooth, there is a separate library called
583       $smlnj/compiler/compiler.cm which puts together and exports the
584       original structure Compiler (or at least something very close to it).
585    
586       There are five members of the original structure Compiler
587       that are not exported directly but which instead became members
588       of a new structure Backend (described by signature BACKEND).  These are:
589       structure Profile (: PROFILE), structure Compile (: COMPILE), structure
590       Interact (: INTERACT), structure Machine (: MACHINE), and val
591       architecture (: string).
592    
593       Structure Compiler.Version has become structure CompilerVersion.
594    
595       Cross-compilers for alpha32, hppa, ppc, sparc, and x86 are provided
596       by $smlnj/compiler/<arch>.cm where <arch> is alpha32, hppa, ppc, sparc,
597       or x86, respectively.
598       Each of these exports the same frontend structures that
599       $smlnj/compiler.cm exports.  But they do not have a structure Backend
600       and instead export some structure <Arch>Backend where <Arch> is Alpha32,
601       Hppa, PPC, Sparc, or X86, respectively.
602    
603       Library $smlnj/compiler/all.cm exports the union of the exports of
604       $smlnj/compiler/<arch>.cm
605    
606       There are no structures <Arch>Compiler anymore, use
607       $smlnj/compiler/<arch>.cm instead.
608    
609       Library host-compiler-0.cm is gone.  Instead, the internal library
610       that instantiates CM is now called cm0.cm.  Selection of the host
611       compiler (backend) is no longer done here but. (Responsibility for it
612       now lies with $smlnj/compiler/current.cm.  This seems to be more
613       logical.)
614    
615       Many individual files have been moved or renamed.  Some files have
616       been split into multiple files, and some "dead" files have been deleted.
617    
618    Aside from these changes to library organization, there are also changes
619    to the way the code itself is organized:
620    
621       Structure Binfile has been re-implemented in such a way that it no
622       longer needs any knowledge of the compiler.  It exclusively deals
623       with the details of binfile layout.  It no longer invokes the
624       compiler (for the purpose of creating new prospective binfile
625       content), and it no longer has any knowledge of how to interpret
626       pickles.
627    
628       Structure Compile (: COMPILE) has been stripped down to the bare
629       essentials of compilation.  It no longer deals with linking/execution.
630       The interface has been cleaned up considerably.
631    
632       Utility routines for dealing with linking and execution have been
633       moved into their own substructures.
634    
635       (The ultimate goal of these changes is to provide a light-weight
636       binfile loader/linker (at least for, e.g., stable libraries) that
637       does not require CM or the compiler to be present.)
638    
639    CM documentation has been updated to reflect the changes to library
640    organization.
641    
642    ----------------------------------------------------------------------
643    Name: Matthias Blume
644    Date: 2001/07/10 17:30:00 EDT
645    Tag: Release_110_34
646    Description:
647    
648    Minor tweak to 110.34 (re-tagged):
649    
650      - README.html file added to CVS repository
651      - runtime compiles properly under FreeBSD 3.X and 4.X
652    
653    ----------------------------------------------------------------------
654    Name: Matthias Blume
655    Date: 2001/07/10 17:30:00 EDT
656    Tag: Release_110_34
657    Description:
658    
659    New version number (110.34). New bootfiles.
660    
661    ----------------------------------------------------------------------
662    Name: Matthias Blume
663    Date: 2001/07/09 16:00:00 EDT
664    Tag: blume-20010709-more-varargs
665    Description:
666    
667    I changed the handling of varargs in ml-nlffigen again:
668    The ellipsis ... will now simply be ignored (with an accompanying warning).
669    
670    The immediate effect is that you can actually call a varargs function
671    from ML -- but you can't actually supply any arguments beyond the ones
672    specified explicitly.  (For example, you can call printf with its format
673    string, but you cannot pass additional arguments.)
674    
675    This behavior is only marginally more useful than the one before, but
676    it has the advantage that a function or, more importantly, a function
677    type never gets dropped on the floor, thus avoiding follow-up problems with
678    other types that refer to the offending one.
679    
680    ----------------------------------------------------------------------
681    Name: Matthias Blume
682    Date: 2001/07/09 11:25:00 EDT
683    Tag: blume-20010709-varargs
684    Description:
685    
686    1. ckit-lib.cm now exports structure Error
687    2. ml-nlffigen reports occurences of "..." (i.e., varargs function types)
688       with a warning accompanied by a source location.  Moreover, it
689       merely skips the offending function or type and proceeds with the
690       rest of its work.u  As a result, one can safely feed C code containing
691       "..." to ml-nlffigen.
692    3. There are some internal improvements to CM, providing slightly
693       more general string substitutions in the tools subsystem.
694    
695    ----------------------------------------------------------------------
696    Name: Matthias Blume
697    Date: 2001/06/27 15:10:00 EDT
698    Tag: blume-20010627-concur
699    Description:
700    
701    Fixed a small bug in CM's handling of parallel compilation.
702    (You could observe the bug by Control-C-interrupting an ordinary
703    CMB.make or CM.stabilize and then attaching some compile servers.
704    The result was that all of a sudden the previously interrupted
705    compilation would continue on its own.  This was because of
706    an over-optimization: CM did not bother to clean out certain queues
707    when no servers were attached "anyway", resulting in the contents
708    of these queues to grab control when new servers did get attached.)
709    
710    There is also another minor update to the CM manual.
711    
712    ----------------------------------------------------------------------
713    Name: Matthias Blume
714    Date: 2001/06/26 16:15:00 EDT
715    Tag: blume-20010626-cmdoc
716    Description:
717    
718    Minor typo fixed in CM manual (syntax diagram for libraries).
719    
720    ----------------------------------------------------------------------
721    Name: Matthias Blume
722    Date: 2001/06/25 22:55:00 EDT
723    Tag: blume-20010625-x86pc
724    Description:
725    
726    Fixed a nasty bug in the X86 assembly code that caused signal
727    handlers to fail (crash) randomly.
728    
729    ----------------------------------------------------------------------
730    Name: Matthias Blume
731    Date: 2001/06/25 12:05:00 EDT
732    Tag: blume-20010625-nlffigen
733    Description:
734    
735    This update fixes a number of minor bugs in ml-nlffigen as reported by
736    Nick Carter <nbc@andrew.cmu.edu>.
737    
738      1. Silly but ok typedefs of the form "typedef void myvoid;" are now accepted.
739      2. Default names for generated files are now derived from the name of
740         the C file *without its directory*.  In particular, this causes generated
741         files to be placed locally even if the C file is in some system directory.
742      3. Default names for generated signatures and structures are also derived
743         from the C file name without its directory.  This avoids silly things
744         like "structure GL/GL".
745         (Other silly names are still possible because ml-nlffigen does not do
746          a thorough check of whether generated names are legal ML identifiers.
747          When in doubt, use command line arguments to force particular names.)
748    
749    ----------------------------------------------------------------------
750    Name: Matthias Blume
751    Date: 2001/06/21 12:25:00 EDT
752    Tag: blume-20010621-eXene
753    Description:
754    
755    eXene now compiles and (sort of) works again.
756    
757    The library name (for version > 110.33) is $/eXene.cm.
758    
759    I also added an new example in src/eXene/examples/nbody.  See the
760    README file there for details.
761    
762    ----------------------------------------------------------------------
763    Name: Matthias Blume
764    Date: 2001/06/20 16:40:00 EDT
765    Tag: blume-20010620-cml
766    Description:
767    
768    CML now compiles and works again.
769    
770    Libraries (for version > 110.33):
771    
772      $cml/cml.cm            Main CML library.
773      $cml/basis.cm          CML's version of $/basis.cm.
774      $cml/cml-internal.cm   Internal helper library.
775      $cml/core-cml.cm       Internal helper library.
776      $cml-lib/trace-cml.cm  Tracing facility.
777      $cml-lib/smlnj-lib.cm  CML's version of $/smlnj-lib.cm
778    
779    The installer (config/install.sh) has been taught how to properly
780    install this stuff.
781    
782    ----------------------------------------------------------------------
783    Name: Matthias Blume
784    Date: 2001/06/19 17:55:00 EDT
785    Tag: blume-20010619-instantiate
786    Description:
787    
788    This un-breaks the fix for bug 1432.
789    (The bug was originally fixed in 110.9 but I broke it again some
790    time after that.)
791    
792    ----------------------------------------------------------------------
793    Name: Matthias Blume
794    Date: 2001/06/19 17:25:00 EDT
795    Tag: blume-20010619-signals
796    Description:
797    
798    This should (hopefully) fix the long-standing signal handling bug.
799    (The runtime system was constructing a continuation record with an
800    incorrect descriptor which would cause the GC to drop data on the floor...)
801    
802    ----------------------------------------------------------------------
803    Name: Matthias Blume
804    Date: 2001/06/15 15:05:00 EDT
805    Tag: blume-20010615-moresparc
806    Description:
807    
808    Here is a short late-hour update related to Sparc c-calls:
809    
810     -- made handling of double-word arguments a bit smarter
811    
812     -- instruction selection phase tries to collapse certain clumsily
813        constructed ML-Trees; typical example:
814    
815            ADD(ty,ADD(_,e,LI d1),LI d2)  ->  ADD(ty,e,LI(d1+d2))
816    
817        This currently has no further impact on SML/NJ since mlriscGen does
818        not seem to generate such patterns in the first place, and c-calls
819        (which did generate them in the beginning) has meanwhile been fixed
820        so as to avoid them as well.
821    
822    ----------------------------------------------------------------------
823    Name: Matthias Blume
824    Date: 2001/06/15 15:05:00 EDT
825    Tag: blume-20010615-sparc
826    Description:
827    
828    The purpose of this update is to provide an implementation of NLFFI
829    on Sparc machines.
830    
831    Here are the changes in detail:
832    
833       * src/MLRISC/sparc/c-calls/sparc-c-calls.sml is a new file containing
834       the Sparc implementation of the c-calls API.
835       * The Sparc backend of SML/NJ has been modified to uniformely use %fp
836       for accessing the ML frame.  Thus, we have a real frame pointer and
837       can freely modify %sp without need for an omit-frame-ptr phase.
838       The vfp logic in src/compiler/CodeGen/* has been changed to accomodate
839       this case.
840       * ml-nlffigen has been taught to produce code for different architectures
841       and calling conventions.
842       * In a way similar to what was done in the x86 case, the Sparc
843       backend uses its own specific extension to mltree.  (For example,
844       it needs to be able to generate UNIMP instructions which are part
845       of the calling convention.)
846       * ml-nlffi-lib was reorganized to make it more modular (in particular,
847       to make it easier to plug in new machine- and os-dependent parts).
848    
849    There are some other fairly unrelated bug fixes and cleanups as well:
850    
851       * I further hacked the .cm files for MLRISC tools (like MDLGen) so
852       that they properly share their libraries with existing SML/NJ libraries.
853       * I fixed a minor cosmetic bug in CM, supressing certain spurious
854       follow-up error messages.
855       * Updates to CM/CMB documentation.
856    
857    TODO items:
858    
859       * MLRISC should use a different register as its asmTemp on the Sparc.
860         (The current %o2 is a really bad choice because it is part of the
861         calling conventions, so things might interfere in unexpected ways.)
862    
863    ----------------------------------------------------------------------
864    Name: Matthias Blume
865    Date: 2001/06/07
866    Tag: blume-20010607-calls
867    Description:
868    
869    A number of internal changes related to C calls and calling conventions:
870    
871    1. ML-Tree CALL statements now carry a "pops" field.  It indicates the
872       number of bytes popped implicitly (by the callee).  In most cases
873       this field is 0 but on x86/win32 it is some non-zero value.  This
874       is information provided for the benefit of the "omit-frameptr" pass.
875    2. The CALL instruction on the x86 carries a similar "pops" field.
876       The instruction selection phase copies its value from the ML-Tree
877       CALL statement.
878    3. On all other architectures, the instruction selection phase checks
879       whether "pops=0" and complains if not.
880    4. The c-calls implementation for x86 now accepts two calling conventions:
881       "ccall" and "stdcall".  When "ccall" is selected, the caller cleans
882       up after the call and pops is set to 0.  For "stdcall", the caller
883       does nothing, leaving the cleanup to the callee; pops is set to
884       the number of bytes that were pushed onto the stack.
885    5. The cproto decoder (compiler/Semant/types/cproto.sml) now can
886       distinguish between "ccall" and "stdcall".
887    6. The UNIMP instruction has been added to the supported Sparc instruction
888       set. (This is needed for implementing the official C calling convention
889       on this architecture.)
890    7. I fixed some of the .cm files under src/MLRISC/Tools to make them
891       work with the latest CM.
892    
893    ----------------------------------------------------------------------
894    Name: Matthias Blume
895    Date: 2001/06/05 15:10:00 EDT
896    Tag: blume-20010605-cm-index
897    Description:
898    
899    0. The "lambdasplit" parameter for class "sml" in CM has been documented.
900    
901    1. CM can now generate "index files".  These are human-readable files
902       that list on a per-.cm-file basis each toplevel symbol defined or
903       imported.  The location of the index file for
904       <p>/<d>.cm is <p>/CM/INDEX/<d>.cm.
905       To enable index-file generation, set CM.Control.generate_index to true
906       or export an environment-symbol: export CM_GENERATE_INDEX=true.
907    
908       The CM manual has been updated accordingly.
909    
910    2. I made some slight modifications to the c-calls API in MLRISC.
911    
912         a) There is now a callback to support saving/restoring of
913            dedicated but caller-save registers around the actual call
914            instruction.
915         b) One can optionally specify a comment-annotation for the
916            call instruction.
917    
918    3. SML/NJ (mlriscGen.sml) uses this new API for the rawccall primop.
919       (For example, the comment annotation shows the C prototype of
920        the function being called.)
921    
922    ----------------------------------------------------------------------
923    Name: Matthias Blume
924    Date: 2001/06/01 13:30:00 EDT
925    Tag: blume-20010601-nlffi-cleanup
926    Description:
927    
928    This is mostly a cleanup of MLFFI stuff:
929    
930       - some signature files have been put into a more exposed place
931       - the ugly 'f type parameter is gone (simplifies types tremendously!)
932       - ml-nlffigen changed accordingly
933       - tutorial updated
934    
935    Other changes:
936    
937       - author's affiliation in CM manual(s) updated
938       - some more recognized keywords added to Allen's sml.sty
939    
940    ----------------------------------------------------------------------
941    Name: Matthias Blume
942    Date: 2001/05/25 15:30:00 EDT
943    Tag: blume-20010525-iptr
944    Description:
945    
946      - put the official 110.33-README (as it appears on the ftp server) under
947        CVS
948      - fixed a small bug related to incomplete pointer types in
949        ml-nlffigen
950      - small cosmetic change to the ml-nlffi-lib's "arr" type constructor
951        (it does not need the 'f type parameter)
952    
953    ----------------------------------------------------------------------
954    Name: Matthias Blume
955    Date: 2001/05/23 14:30:00 EDT
956    Tag: Release_110_33
957    Description:
958    
959    New version number (110.33).  New bootfiles.
960    
961    ----------------------------------------------------------------------
962    Name: Matthias Blume
963    Date: 2001/05/22 18:06:00 EDT
964    Tag: blume-20010522-targets
965    Description:
966    
967    Made install.sh use file config/targets.customized if it exists, falling
968    back to config/targets if it doesn't.  This way one can have a customized
969    version of the targets file without touching the "real thing", thus
970    eliminating the constant fear of accidentally checking something bogus
971    back into the CVS repository...  (File config/targets.customized must
972    not be added to the repository!)
973    
974    ----------------------------------------------------------------------
975    Name: Matthias Blume
976    Date: 2001/05/22 16:30:00 EDT
977    Tag: blume-20010522-minitut
978    Description:
979    
980    1. Bug fix in ml-nlffigen; now (hopefully) correctly handling
981       struct returns.
982    2. Added src/ml-nlffi-lib/Doc/mini-tutorial.txt.  This is some very
983       incomplete, preliminary documentation for NLFFI.
984    
985    ----------------------------------------------------------------------
986    Name: Matthias Blume
987    Date: 2001/05/14 11:30:00 EDT
988    Tag: blume-20010514-script
989    Description:
990    
991    Some bugs in install script fixed.
992    
993    In addition to that I also made a slight change to the NLFFI API:
994    Functors generated by ml-nlffigen now take the dynamic library as a
995    straight functor argument, not as a suspended one.  (The original
996    functor code used to force the suspension right away anyway, so there
997    was nothing gained by this complication of the interface.)
998    
999    ----------------------------------------------------------------------
1000    Name: Matthias Blume
1001    Date: 2001/05/11 14:35:00 EDT
1002    Tag: blume-20010511-ml-nlffi
1003    Description:
1004    
1005    I finally took the plunge and added my new FFI code to the main
1006    repository.  For x86-linux it is now ready for prime-time.
1007    
1008    There are two new subdirectories of "src":
1009    
1010      - ml-nlffi-lib:
1011           The utility library for programs using the FFI interface.
1012           Here is the implementation of $/c.cm and its associated low-level
1013           partners $/c-int.cm and $/memory.cm.
1014      - ml-nlffigen:
1015           A stand-alone program for generating ML glue code from C source
1016           code.
1017    
1018    Building ml-nlffigen requires $/ckit-lib.cm.
1019    
1020    The config/install.sh script has been updates to do the Right Thing
1021    (hopefully).
1022    
1023    Notice that the source tree for the C-Kit will not be put under "src"
1024    but directly under the installation root directory.  (This is the
1025    structure that currently exists on the CVS server when you check out
1026    module "sml".)  Fortunately, config/install.sh knows about this oddity.
1027    
1028    Bugs: No documentation yet.
1029    
1030    ----------------------------------------------------------------------
1031    Name: Matthias Blume
1032  Date: 2001/05/09 16:35:00 EDT  Date: 2001/05/09 16:35:00 EDT
1033  Tag: blume-20010509-cpscontract  Tag: blume-20010509-cpscontract
1034  Description:  Description:
# Line 2201  Line 3216 
3216       elaborator). There were a lot of changes during my "linkpath" trials       elaborator). There were a lot of changes during my "linkpath" trials
3217       that could have been reverted to their original state but weren't.       that could have been reverted to their original state but weren't.
3218       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
3219       than what was strictly necessary...  (I _did_ resist the tempation       than what was strictly necessary...  (I _did_ resist the temptation
3220       of doing any "global reformatting" to avoid an untimely death at       of doing any "global reformatting" to avoid an untimely death at
3221       Dave's hands. :)       Dave's hands. :)
3222    

Legend:
Removed from v.826  
changed lines
  Added in v.986

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