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 832, Tue May 22 20:30:36 2001 UTC revision 1078, Tue Feb 19 21:26:48 2002 UTC
# Line 8  Line 8 
8  The form of an entry should be:  The form of an entry should be:
9    
10  Name:  Name:
11  Date:  Date: yyyy/mm/dd
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    
15  ----------------------------------------------------------------------  ----------------------------------------------------------------------
16  Name: Matthias Blume  Name: Matthias Blume
17    Date: 2002/02/19 16:20:00 EST
18    Tag: blume-20020219-cmffi
19    Description:
20    
21    A wild mix of changes, some minor, some major:
22    
23    * All C FFI-related libraries are now anchored under $c:
24        $/c.cm      --> $c/c.cm
25        $/c-int.cm  --> $c/internals/c-int.cm
26        $/memory.cm --> $c/memory/memory.cm
27    
28    * "make" tool (in CM) now treats its argument pathname slightly
29      differently:
30        1. If the native expansion is an absolute name, then before invoking
31           the "make" command on it, CM will apply OS.Path.mkRelative
32           (with relativeTo = OS.FileSys.getDir()) to it.
33        2. The argument will be passed through to subsequent phases of CM
34           processing without "going native".  In particular, if the argument
35           was an anchored path, then "make" will not lose track of that anchor.
36    
37    * Compiler backends now "know" their respective C calling conventions
38      instead of having to be told about it by ml-nlffigen.  This relieves
39      ml-nlffigen from one of its burdens.
40    
41    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
42    
43    * Export C_DEBUG and C_Debug from $c/c.cm.
44    
45    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
46      subtyping relationship between incomplete pointers and their complete
47      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
48      with the convention of instantiating 'o with ('t, 'c) obj whenever
49      the pointer target type is complete.  In the incomplete case, 'o
50      will be instantiated with some "'c iobj" -- a type obtained by
51      using one of the functors PointerToIncompleteType or PointerToCompleteType.
52    
53      Operations that work on both incomplete and complete pointer types are
54      typed as taking an 'o ptr while operations that require the target to
55      be known are typed as taking some ('t, 'c) obj ptr.
56    
57      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
58      where void is an eqtype without any values.  This makes it possible
59      to work on voidptr values using functions meant to operate on light
60      incomplete pointers.
61    
62    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
63      been vastly simplified.
64    
65    ----------------------------------------------------------------------
66    Name: Matthias Blume
67    Date: 2002/02/19 10:48:00 EST
68    Tag: blume-20020219-pqfix
69    Description:
70    
71    Applied Chris Okasaki's bug fix for priority queues.
72    
73    ----------------------------------------------------------------------
74    Name: Matthias Blume
75    Date: 2002/02/15 17:05:00
76    Tag: Release_110_39
77    Description:
78    
79    Last-minute retagging is becoming a tradition... :-(
80    
81    This is the working release 110.39.
82    
83    ----------------------------------------------------------------------
84    Name: Matthias Blume
85    Date: 2002/02/15 16:00:00 EST
86    Tag: Release_110_39-orig
87    Description:
88    
89    Working release 110.39.  New bootfiles.
90    
91    (Update: There was a small bug in the installer so it wouldn't work
92    with all shells.  So I retagged. -Matthias)
93    
94    ----------------------------------------------------------------------
95    Name: Matthias Blume
96    Date: 2002/02/15 14:17:00 EST
97    Tag: blume-20020215-showbindings
98    Description:
99    
100    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
101    the latter can be useful for exploring what bindings are available at
102    the interactive prompt.  (The first function returns only the list
103    of symbols that are really bound, the second prints those but also the
104    ones that CM's autoloading mechanism knows about.)
105    
106    ----------------------------------------------------------------------
107    Name: Matthias Blume
108    Date: 2002/02/15 12:08:00 EST
109    Tag: blume-20020215-iptrs
110    Description:
111    
112    Two improvements to ml-nlffigen:
113    
114      1. Write files only if they do not exist or if their current contents
115         do not coincide with what's being written.  (That is, avoid messing
116         with the time stamps unless absolutely necessary.)
117    
118      2. Implement a "repository" mechanism for generated files related
119         to "incomplete pointer types".   See the README file for details.
120    
121    ----------------------------------------------------------------------
122    Name: Matthias Blume
123    Date: 2002/02/14 11:50:00 EST
124    Tag: blume-20020214-quote
125    Description:
126    
127    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
128    because of the new and improved tag generation scheme.  (Thanks to Allen
129    Leung for pointing it out.)
130    
131    ----------------------------------------------------------------------
132    Name: Lal George
133    Date: 2002/02/14 09:55:27 EST 2002
134    Tag: george-20020214-isabelle-bug
135    Description:
136    
137    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
138    of Isabelle on the x86.
139    
140    From Allen:
141    -----------
142     I've found the problem:
143    
144         in ra-core.sml, I use the counter "blocked" to keep track of the
145         true number of elements in the freeze queue.  When the counter goes
146         to zero, I skip examining the queue.  But I've messed up the
147         bookkeeping in combine():
148    
149             else ();
150             case !ucol of
151               PSEUDO => (if !cntv > 0 then
152                     (if !cntu > 0 then blocked := !blocked - 1 else ();
153                                        ^^^^^^^^^^^^^^^^^^^^^^^
154                      moveu := mergeMoveList(!movev, !moveu)
155                     )
156                  else ();
157    
158         combine() is called to coalesce two nodes u and v.
159         I think I was thinking that if the move counts of u and v are both
160         greater than zero then after they are coalesced then one node is
161         removed from the freeze queue.  Apparently I was thinking that
162         both u and v are of low degree, but that's clearly not necessarily true.
163    
164    
165    02/12/2002:
166        Here's the patch.  HOL now compiles.
167    
168        I don't know how this impact on performance (compile
169        time or runtime).  This bug caused the RA (especially on the x86)
170        to go thru the potential spill phase when there are still nodes on the
171        freeze queue.
172    
173    
174    
175    
176    ----------------------------------------------------------------------
177    Name: Matthias Blume
178    Date: 2002/02/13 22:40:00 EST
179    Tag: blume-20020213-fptr-rtti
180    Description:
181    
182    Fixed a bug in ml-nlffigen that was introduced with one of the previous
183    updates.
184    
185    ----------------------------------------------------------------------
186    Name: Matthias Blume
187    Date: 2002/02/13 16:41:00 EST
188    Tag: blume-20020213-cmlpq
189    Description:
190    
191    Added new priority queue export symbols (which have just been added to
192    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
193    not compile and the installer would choke.)
194    
195    ----------------------------------------------------------------------
196    Name: Matthias Blume
197    Date: 2002/02/13 16:15:00 EST
198    Tag: blume-20020213-various
199    Description:
200    
201    1. More tweaks to ml-nlffigen:
202    
203       - better internal datastructures (resulting in slight speedup)
204       - "-match" option requires exact match
205       - "localized" gensym counters (untagged structs/unions nested within
206         other structs/unions or within typedefs get a fresh counter; their
207         tag will be prefixed by a concatenation of their parents' tags)
208       - bug fixes (related to calculation of transitive closure of types
209         to be included in the output)
210    
211    2. Minor Basis updates:
212    
213       - added implementations for List.collate and Option.app
214    
215    ----------------------------------------------------------------------
216    Name: Matthias Blume
217    Date: 2002/02/11 15:55:00 EST
218    Tag: blume-20020211-gensym
219    Description:
220    
221    Added a "-gensym" option to command line of ml-nlffigen.  This can be
222    used to specify a "stem" -- a string that is inserted in all "gensym'd"
223    names (ML structure names that correspond to unnamed C structs, unions,
224    and enums), so that separate runs of ml-nlffigen do not clash.
225    
226    ----------------------------------------------------------------------
227    Name: Matthias Blume
228    Date: 2002/02/11 12:05:00 EST
229    Tag: blume-20020211-gensml
230    Description:
231    
232    A quick fix for a problem with GenSML (in the pgraph-util library):
233    Make generation of toplevel "local" optional.  (Strictly speaking,
234    signature definitions within "local" are not legal SML.)
235    
236    Other than that: updates to INSTALL and cm/TODO.
237    
238    ----------------------------------------------------------------------
239    Name: Matthias Blume
240    Date: 2002/02/08 15:00:00 EST
241    Tag: blume-20020208-uniquepid
242    Description:
243    
244    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
245    
246    1. The installer (config/install.sh) has gotten smarter:
247    
248         - Configuration options are a bit easier to specify now
249           (in config/targets).
250         - Bug in recognizing .tar.bz2 files fixed.
251         - Installer automatically resolves dependencies between
252           configuration options (e.g., if you ask for eXene, you will
253           also get cml -- regardless whether you asked for it or not).
254         - Installer can run in "quieter mode" by setting the environment
255           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
256           "completely silent", though.
257         - Build HashCons library as part of smlnj-lib.
258    
259    2. A new scheme for assigning persistent identifiers to compilation
260       units (and, by extension, to types etc.) has been put into place.
261       This fixes a long-standing bug where types and even dynamic values
262       can get internally confused, thereby compromising type safety
263       (abstraction) and dynamic correctness.  See
264    
265         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
266    
267       for an example of how things could go wrong until now.
268    
269       The downside of the new scheme is that pids are not quite as
270       persistent as they used to be: CM will generate a fresh pid
271       for every compilation unit that it thinks it sees for the first
272       time.  That means that if you compile starting from a clean, fresh
273       source tree at two different times, you end up with different
274       binaries.
275    
276       Cutoff recompilation, however, has not been compromised because
277       CM keeps pid information in special caches between runs.
278    
279    ----------------------------------------------------------------------
280    Name: Lal George
281    Date: 2002/02/07 15:34:13 EST 2002
282    Tag: <none>
283    Description:
284    
285    Compilers that generate assembly code may produce  global labels
286    whose value is resolved at link time. The various peephole optimization
287    modules did not take this in account.
288    
289    TODO. The Labels.addrOf function should really return an option
290    type so that clients are forced to deal with this issue, rather
291    than an exception being raised.
292    
293    ----------------------------------------------------------------------
294    Name: Lal George
295    Date: 2002/02/06 13:55:02 EST
296    Tag: george-20020206-ra-breakup
297    Description:
298    
299    1. A bug fix from Allen.
300    
301        A typo causes extra fstp %st(0)'s to be generated at compensation
302        edges, which might cause stack underflow traps at runtime.  This
303        occurs in fft where there are extraneous fstps right before the 'into'
304        trap instruction (in this case they are harmless since none of the
305        integers overflow.)
306    
307    2. Pulled out various utility modules that were embedded in the modules
308       of the register allocator. I need these modules for other purposes, but
309       they are not complete enough to put into a library (just yet).
310    ----------------------------------------------------------------------
311    Name: Matthias Blume
312    Date: 2002/01/31 16:05:00 EST
313    Tag: blume-20020131-sparc-ccalls
314    Description:
315    
316    1. C-calls on Sparc needlessly allocated a huge chunk (96 bytes)
317       of extra stack space by mistake.  Fixed.
318    
319    2. Bug in logic of handling of command-line options in ml-nlffigen fixed.
320    
321    ----------------------------------------------------------------------
322    Name: Allen Leung
323    Date: 2002/01/30
324    Tag: leunga-20020130-nowhere-bug-fix
325    Description:
326    
327       MLRISC bug fixes:
328       1. Fixed a bindings computation bug in the 'nowhere' program generator tool.
329       2. MachineInt.fromString was negating its value.
330    
331    ----------------------------------------------------------------------
332    Name: Matthias Blume
333    Date: 2002/01/29
334    Tag: blume-20020129-INSTALL
335    Description:
336    
337    - Added somewhat detailed installation instructions (file INSTALL).
338    - Fixed curl-detection bug in config/install.sh.
339    - It is now possible to select the URL getter using the URLGETTER
340      environment variable:
341    
342          not set / "unknown"      --> automatic detection (script tries wget,
343                                       curl, and lynx)
344          "wget" / "curl" / "lynx" --> use the specified program (script "knows"
345                                       how to properly invoke them)
346          other                    --> use $URLGETTER directly, it must take
347                                       precisely two command-line arguments
348                                       (source URL and destination file name)
349    
350    ----------------------------------------------------------------------
351    Name: Matthias Blume
352    Date: 2002/01/28
353    Tag: blume-20020128-sparc-ccalls
354    Description:
355    
356    - Fixed problem with calculation of "used" registers in sparc-c-calls.
357    - Make use of the allocParam argument in sparc-c-calls.
358    
359    ----------------------------------------------------------------------
360    Name: Matthias Blume
361    Date: 2002/01/28
362    Tag: blume-20020128-allocParam
363    Description:
364    
365    John Reppy:  Changes c-calls API to accept client-callback for
366    allocating extra stack space.
367    me: Corresponding changes to mlriscGen (using a dummy argument that
368        does not change the current behavior).
369    
370    ----------------------------------------------------------------------
371    Name: Matthias Blume
372    Date: 2002/01/28 12:00:00
373    Tag: Release_110_38
374    Description:
375    
376    This time for real!!!
377    
378    ----------------------------------------------------------------------
379    Name: Matthias Blume
380    Date: 2002/01/28 10:56:00 EST
381    Tag: blume-20020128-retraction
382    Description:
383    
384    0. Retracted earlier 110.38.  (The Release_110_38 tag has been replaced
385       with blume-Release_110_38-retracted.)
386    
387    1. Fixed a problem with incorrect rounding modes in real64.sml.
388       (Thanks to Andrew Mccreight <andrew.mccreight@yale.edu>.)
389    
390    2. A bug in ml-nlffigen related to the handling of unnamed structs, unions,
391       and enums fixed.  The naming of corresponding ML identifiers should
392       now be consistent again.
393    
394    ----------------------------------------------------------------------
395    Name: Allen Leung
396    Date: 2002/01/27
397    Tag: leunga-20020127-nowhere
398    Description:
399    
400       Added a target called nowhere in the configuration scripts.
401       Enabling this will build the MLRISC 'nowhere' tool (for translating
402       programs with where-clauses into legal SML code) during installation.
403    
404    ----------------------------------------------------------------------
405    Name: Matthias Blume
406    Date: 2002/01/25 21:27:00 EST
407    Tag: blume-Release_110_38-retracted
408    Description:
409    
410    Call it a (working) release!  Version is 110.38. Bootfiles are ready.
411    
412    README will be added later.
413    
414    !!! NOTE:  Re-tagged as blume-Release_110_38-retracted. Original tag
415    (Release_110_38) removed.  Reason: Last-minute bug fixes.
416    
417    ----------------------------------------------------------------------
418    Name: Matthias Blume
419    Date: 2002/01/25
420    Tag: blume-20020125-ffi
421    Description:
422    
423    A large number of tweaks and improvements to ml-nlffi-lib and
424    ml-nlffigen:
425    
426       - ML represenation types have been streamlined
427       - getter and setter functions work with concrete values, not abstract
428         ones where possible
429       - ml-nlffigen command line more flexible (see README file there)
430       - some bugs have been fixed (hopefully)
431    
432    ----------------------------------------------------------------------
433    Name: Lal George
434    Date: 2002/01/24
435    Tag: george-20020124-risc-ra-interface
436    Description:
437    
438       There is a dramatic simplification in the interface to the
439       register allocator for RISC architectures as a result of making
440       parallel copy instructions explicit.
441    
442    ----------------------------------------------------------------------
443    Name: Matthias Blume
444    Date: 2002/01/22
445    Tag: blume-20020122-x86-ccalls
446    Description:
447    
448    Bug fix for c-calls on x86 (having to do with how char- and
449    short-arguments are being handled).
450    
451    ----------------------------------------------------------------------
452    Name: Matthias Blume
453    Date: 2002/01/21
454    Tag: blume-20020121-ff
455    Description:
456    
457    Another day of fiddling with the FFI...
458    
459    1. Bug fix/workaround:  CKIT does not complain about negative array
460       dimensions, so ml-nlffigen has to guard itself against this possibility.
461       (Otherwise a negative dimension would send it into an infinite loop.)
462    
463    2. Some of the abstract types (light objects, light pointers, most "base"
464       types) in structure C are now eqtypes.
465    
466    3. Added constructors and test functions for NULL function pointers.
467    
468    ----------------------------------------------------------------------
469    Name: Matthias Blume
470    Date: 2002/01/18
471    Tag: blume-20020118-ready-for-new-release
472    Description:
473    
474    Made config/srcarchiveurl point to a new place.  (Will provide boot
475    files shortly.)
476    
477    Maybe we christen this to be 110.38?
478    
479    ----------------------------------------------------------------------
480    Name: Matthias Blume
481    Date: 2002/01/18
482    Tag: blume-20020118-more-ffifiddle
483    Description:
484    
485    Today's FFI fiddling:
486    
487      - Provided a structure CGetSet with "convenient" versions of C.Get.* and
488        C.Set.* that use concrete (MLRep.*) arguments and results instead
489        of abstract ones.
490    
491      - Provided word-style bit operations etc. for "int" representation
492        types in MLRep.S<Foo>Bitops where <Foo> ranges over Char, Int, Short,
493        and Long.
494    
495    ----------------------------------------------------------------------
496    Name: Matthias Blume
497    Date: 2002/01/18
498    Tag: blume-20020118-use-x86-fp
499    Description:
500    
501    Now that x86-fast-fp seems to be working, I turned it back on again
502    by default.  (Seems to work fine now, even with the FFI.)
503    
504    Other than that, I added some documentation about the FFI to
505    src/ml-nlffigen/README and updated the FFI test examples in
506    src/ml-nlffi-lib/Tests/*.
507    
508    ----------------------------------------------------------------------
509    Name: Allen Leung
510    Date: 2002/01/17
511    Tag: leunga-20020117-x86-fast-fp-call
512    Description:
513    
514       1. Fixed a problem with handling return fp values when x86's fast fp
515          mode is turned on.
516    
517       2. Minor pretty printing fix for cellset.  Print %st(0) as %st(0) instead
518          of %f32.
519    
520       3. Added a constructor INT32lit to the ast of MLRISC tools.
521    
522    ----------------------------------------------------------------------
523    Name: Matthias Blume
524    Date: 2002/01/16
525    Tag: blume-20020116-ffifiddle
526    Description:
527    
528    More fiddling with the FFI interface:
529    
530     - Make constness 'c instead of rw wherever possible.  This eliminates
531       the need for certain explicit coercions.  (However, due to ML's
532       value polymorphism, there will still be many cases where explicit
533       coercions are necessary.  Phantom types are not the whole answer
534       to modeling a subtyping relationship in ML.)
535    
536     - ro/rw coersions for pointers added.  (Avoids the detour through */&.)
537    
538     - "printf" test example added to src/ml-nlffi-lib/Tests.  (Demonstrates
539       clumsy workaround for varargs problem.)
540    
541    ----------------------------------------------------------------------
542    Name: Lal George
543    Date: 2002/01/15
544    Tag: <none>
545    Description:
546    
547    1. Since COPY instructions are no longer native to the architecture,
548       a generic functor can be used to implement the expandCopies function.
549    
550    2. Allowed EXPORT and IMPORT pseudo-op declarations to appear inside a
551       TEXT segment.
552    
553    ----------------------------------------------------------------------
554    Name: Matthias Blume
555    Date: 2002/01/15
556    Tag: blume-20020115-ffiupdates
557    Description:
558    
559    1. Fix for bug resulting in single-precision float values being returned
560       incorrectly from FFI calls.
561    
562    2. Small modifications to C FFI API:
563    
564        - memory-allocation routines return straight objects (no options)
565          and raise an exception in out-of-memory situations
566        - unsafe extensions to cast between function pointers and pointers
567          from/to ints
568        - added structure C_Debug as an alternative to structure C where
569          pointer-dereferencing (|*| and |*!) always check for null-pointers
570        - added open_lib' to DynLinkage;  open_lib' works like open_lib
571          but also takes a (possibly empty) list of existing library handles
572          that the current library depends on
573    
574    ----------------------------------------------------------------------
575    Name: Matthias Blume
576    Date: 2002/01/10
577    Tag: blume-20020110-newffigen
578    Description:
579    
580    1. Updates to portable graph code.
581    
582    2. Major update to ml-nlffigen and ml-nlffi-lib.  Things are much
583       more scalable now so that even huge interfaces such as the one
584       for GTK compile in finite time and space. :-)
585       See src/ml-nlffigen/README for details on what's new.
586    
587    ----------------------------------------------------------------------
588    Name: Lal George
589    Date: 2001/01/09 14:31:35 EST 2002
590    Tag: george-20011206-rm-native-copy
591    Description:
592    
593            Removed the native COPY and FCOPY instructions
594            from all the architectures and replaced it with the
595            explicit COPY instruction from the previous commit.
596    
597            It is now possible to simplify many of the optimizations
598            modules that manipulate copies. This has not been
599            done in this change.
600    
601    ----------------------------------------------------------------------
602    Name: Lal George
603    Date: 2001/12/06 16:50:13 EST 2001
604    Tag: george-20011206-mlrisc-instruction
605    Description:
606    
607    Changed the representation of instructions from being fully abstract
608    to being partially concrete. That is to say:
609    
610      from
611            type instruction
612    
613      to
614            type instr                              (* machine instruction *)
615    
616            datatype instruction =
617                LIVE of {regs: C.cellset, spilled: C.cellset}
618              | KILL of {regs: C.cellset, spilled: C.cellset}
619              | COPYXXX of {k: CB.cellkind, dst: CB.cell list, src: CB.cell list}
620              | ANNOTATION of {i: instruction, a: Annotations.annotation}
621              | INSTR of instr
622    
623    This makes the handling of certain special instructions that appear on
624    all architectures easier and uniform.
625    
626    LIVE and KILL say that a list of registers are live or killed at the
627    program point where they appear. No spill code is generated when an
628    element of the 'regs' field is spilled, but the register is moved to
629    the 'spilled' (which is present, more for debugging than anything else).
630    
631    LIVE replaces the (now deprecated) DEFFREG instruction on the alpha.
632    We used to generate:
633    
634            DEFFREG f1
635            f1 := f2 + f3
636            trapb
637    
638    but now generate:
639    
640            f1 := f2 + f3
641            trapb
642            LIVE {regs=[f1,f2,f3], spilled=[]}
643    
644    Furthermore, the DEFFREG (hack) required that all floating point instruction
645    use all registers mentioned in the instruction. Therefore f1 := f2 + f3,
646    defines f1 and uses [f1,f2,f3]! This hack is no longer required resulting
647    in a cleaner alpha implementation. (Hopefully, intel will not get rid of
648    this architecture).
649    
650    COPYXXX is intended to replace the parallel COPY and FCOPY  available on
651    all the architectures. This will result in further simplification of the
652    register allocator that must be aware of them for coalescing purposes, and
653    will also simplify certain aspects of the machine description that provides
654    callbacks related to parallel copies.
655    
656    ANNOTATION should be obvious, and now INSTR represents the honest to God
657    machine instruction set!
658    
659    The <arch>/instructions/<arch>Instr.sml files define certain utility
660    functions for making porting easier -- essentially converting upper case
661    to lower case. All machine instructions (of type instr) are in upper case,
662    and the lower case form generates an MLRISC instruction. For example on
663    the alpha we have:
664    
665      datatype instr =
666         LDA of {r:cell, b:cell, d:operand}
667       | ...
668    
669      val lda : {r:cell, b:cell, d:operand} -> instruction
670        ...
671    
672    where lda is just (INSTR o LDA), etc.
673    
674    ----------------------------------------------------------------------
675    Name: Matthias Blume
676    Date: 2001/11/22 21:40:00 EST
677    Tag: Release_110_37
678    Description:
679    
680    Release 110.37.  This time for real.
681    
682    ----------------------------------------------------------------------
683    Name: Matthias Blume
684    Date: 2001/11/21 16:35:00 EST
685    Tag: blume-20011121-foot-in-mouth
686    Description:
687    
688    Removed the "Release_110_37" tag because of a serious bug.
689    This will be re-tagged once the bug is fixed.
690    
691    ----------------------------------------------------------------------
692    Name: Matthias Blume
693    Date: 2001/11/21 16:14:00 EST
694    Tag: blume-20011121-forgottenfile
695    Description:
696    
697    Forgot to add a file.  (Just a .tex-file -- part of
698    the CM manual source.)
699    
700    ----------------------------------------------------------------------
701    Name: Matthias Blume
702    Date: 2001/11/21 16:10:00 EST
703    Tag: blume-20011121-invalid_110_37
704    Description:
705    
706    Note: I removed the original tag "Release_110_37" from this commit
707          because we found a serious bug in all non-x86 backends.
708          - Matthias
709    
710    1. Modifications to the SML/NJ code generator and to the runtime system
711       so that code object name strings are directly inserted into code
712       objects at code generation time.  The only business the runtime system
713       has with this is now to read the name strings on occasions.
714       (The encoding of the name string has also changed somewhat.)
715    
716    2. CM now implements a simple "set calculus" for specifying export lists.
717       In particular, it is now possible to refer to the export lists of
718       other libraries/groups/sources and form unions as well as differences.
719       See the latest CM manual for details.
720    
721    3. An separate notion of "proxy" libraries has again be eliminated from
722       CM's model.  (Proxy libraries are now simply a special case of using
723       the export list calculus.)
724    
725    4. Some of the existing libraries now take advantage of the new set
726       calculus.
727       (Notice that not all libraries have been converted because some
728       of the existing .cm-files are supposed to be backward compatible
729       with 110.0.x.)
730    
731    5. Some cleanup in stand-alone programs.  (Don't use "exnMessage" -- use
732       "General.exnMessage"!  The former relies on a certain hook to be
733       initialized, and that often does not happen in the stand-alone case.)
734    
735    ----------------------------------------------------------------------
736    Name: Lal George
737    Date: 2001/11/21  13:56:18 EST
738    Tag: george-2001121-pseudo-ops
739    Description:
740    
741      Implemented a complete redesign of MLRISC pseudo-ops. Now there
742      ought to never be any question of incompatabilities with
743      pseudo-op syntax expected by host assemblers.
744    
745      For now, only modules supporting GAS syntax are implemented
746      but more should follow, such as MASM, and vendor assembler
747      syntax, e.g. IBM as, Sun as, etc.
748    
749    ----------------------------------------------------------------------
750    Name: Matthias Blume
751    Date: 2001/11/14 11:52:00 EST
752    Tag: blume-20011114-srcname
753    Description:
754    
755    1. Routed the name of the current source file to mlriscgen where it
756       should be directly emitted into the code object.  (This last part
757       is yet to be done.)
758    
759    2. Some cleanup of the pgraph code to make it match the proposal that
760       I put out the other day.  (The proposal notwithstanding, things are
761       still in flux here.)
762    
763    ----------------------------------------------------------------------
764    Name: Lal George
765    Date: 2001/11/14 09:44:04 EST
766    Tag:
767    Description:
768    
769      Fix for a backpatching bug reported by Allen.
770    
771      Because the boundary between short and long span-dependent
772      instructions is +/- 128, there are an astounding number of
773      span-dependent instructions whose size is over estimated.
774    
775      Allen came up with the idea of letting the size of span
776      dependent instructions be non-monotonic, for a maxIter
777      number of times, after which the size must be monotonically
778      increasing.
779    
780      This table shows the number of span-dependent instructions
781      whose size was over-estimated as a function of maxIter, for the
782      file Parse/parse/ml.grm.sml:
783    
784         maxIter            # of instructions:
785            10                      687
786            20                      438
787            30                      198
788            40                        0
789    
790      In compiling the compiler, there is no significant difference in
791      compilation speed between maxIter=10 and maxIter=40. Actually,
792      my measurements showed that maxIter=40 was a tad faster than
793      maxIter=10! Also 96% of the  files in the compiler reach a fix
794      point within 13 iterations, so fixing maxIter at 40, while high,
795      is okay.
796    
797    ----------------------------------------------------------------------
798    Name: Matthias Blume
799    Date: 2001/10/31 15:25:00 EST
800    Tag: blume-20011031-pgraph
801    Description:
802    
803    CKIT:
804    * Changed the "Function" constructor of type Ast.ctype to carry optional
805      argument identifiers.
806    * Changed the return type of TypeUtil.getFunction accordingly.
807    * Type equality ignores the argument names.
808    * TypeUtil.composite tries to preserve argument names but gives up quickly
809      if there is a mismatch.
810    
811    installation script:
812    * attempts to use "curl" if available (unless "wget" is available as well)
813    
814    CM:
815    * has an experimental implementation of "portable graphs" which I will
816      soon propose as an implementation-independent library format
817    * there are also new libraries $/pgraph.cm and $/pgraph-util.cm
818    
819    NLFFI-LIB:
820    * some cleanup (all cosmetic)
821    
822    NLFFIGEN:
823    * temporarily disabled the mechanism that suppresses ML output for
824      C definitions whose identifiers start with an underscore character
825    * generate val bindings for enum constants
826    * user can request that only one style (light or heavy) is being used;
827      default is to use both (command-line arguments: -heavy and -light)
828    * fixed bug in handling of function types involving incomplete pointers
829    * generate ML entry points that take record arguments (i.e., using
830      named arguments) for C functions that have a prototype with named
831      arguments
832      (see changes to CKIT)
833    
834    ----------------------------------------------------------------------
835    Name: Allen Leung
836    Date: 2001/10/27 20:34:00 EDT
837    Tag: leunga-20011027-x86-fast-fp-call
838    Description:
839    
840       Fixed the bug described in blume-20010920-slowfp.
841    
842       The fix involves
843          1. generating FCOPYs in FSTP in ia32-svid
844          2. marking a CALL with the appropriate annotation
845    
846    ----------------------------------------------------------------------
847    Name: Matthias Blume
848    Date: 2001/10/16 11:32:00 EDT
849    Tag: blume-20011016-netbsd
850    Description:
851    
852    Underscore patch from Chris Richards (fixing problem with compiling
853    runtime system under recent NetBSD).
854    
855    ----------------------------------------------------------------------
856    Name: Allen Leung
857    Date: 2001/10/12 17:18:32 EDT 2001
858    Tag: leung-20011012-x86-printflowgraph
859    Description:
860    
861    X86RA now uses a valid (instead of dummy) PrintFlowgraph module.
862    
863    ----------------------------------------------------------------------
864    Name: Lal George
865    Date: 2001/10/11 23:51:34 EDT
866    Tag: george-20011011-too-many-instrs
867    Description:
868    
869    The representation of a program point never expected to see more
870    than 65536 instructions in a basic block!
871    
872    ----------------------------------------------------------------------
873    Name: Lal George
874    Date: 2001/10/09 09:41:37 EDT
875    Tag: george-20011008-mlrisc-labels
876    Description:
877    
878    Changed the machine description files to support printing of
879    local and global labels in assembly code, based on host assembler
880    conventions.
881    
882    ----------------------------------------------------------------------
883    Name: Matthias Blume
884    Date: 2001/09/25 15:25:00 EDT
885    Tag: blume-20010925-exninfo
886    Description:
887    
888    I provided a non-hook implementation of exnName (at the toplevel) and
889    made the "dummy" implementation of exnMessage (at the toplevel) more
890    useful: if nothing gets "hooked in", then at least you are going to
891    see the exception name and a message indicating why you don't see more.
892    
893    [For the time being, programs that need exnMessage and want to use
894    ml-build should either use General.exnMessage (strongly recommended) or
895    refer to structure General at some other point so that CM sees a
896    static dependency.]
897    
898    [Similar remarks go for "print" and "use":  If you want to use their
899    functionality in stand-alone programs generated by ml-build, then use
900    TextIO.output and Backend.Interact.useFile (from $smlnj/compiler.cm).]
901    
902    ----------------------------------------------------------------------
903    Name: Matthias Blume
904    Date: 2001/09/20 17:28:00 EDT
905    Tag: blume-20010920-slowfp
906    Description:
907    
908    Allen says that x86-fast-fp is not safe yet, so I turned it off again...
909    
910    ----------------------------------------------------------------------
911    Name: Matthias Blume
912    Date: 2001/09/20 17:20:00 EDT
913    Tag: blume-20010920-canonicalpaths
914    Description:
915    
916    0. Updated the BOOT file (something that I forgot to do earlier).
917    
918    1. Small internal change to CM so that it avoids "/../" in filenames
919       as much as possible (but only where it is safe).
920    
921    2. Changed config/_run-sml (resulting in a changed bin/.run-sml) so
922       that arguments that contain delimiters are passed through correctly.
923       This change also means that all "special" arguments of the form
924       @SMLxxx... must come first.
925    
926    3. Changed install script to put relative anchor names for tool commands
927       into pathconfig.
928    
929    ----------------------------------------------------------------------
930    Name: Matthias Blume
931    Date: 2001/09/18 15:35:00 EDT
932    Tag: blume-20010918-readme11036
933    Description:
934    
935    Added README files.
936    
937    ----------------------------------------------------------------------
938    Name: Matthias Blume
939    Date: 2001/09/18 11:45:00 EDT
940    Tag: Release_110_36 (retag)
941    Description:
942    
943    Fixed mistake in config/preloads. Retagged as 110.36.
944    
945    ----------------------------------------------------------------------
946    Name: Matthias Blume
947    Date: 2001/09/18 09:40:00 EDT
948    Tag: Release_110_36_orig (tag changed)
949    Description:
950    
951    New version (110.36).  New bootfiles.
952    
953    ----------------------------------------------------------------------
954    Name: Matthias Blume
955    Date: 2001/09/14 16:15:00 EDT
956    Tag: blume-20010914-x86fastfp
957    Description:
958    
959    John committed some changes that Allen made, in particular a (hopefully)
960    correctly working version of the x86-fp module.
961    
962    I changed the default setting of the Control.MLRISC.getFlag "x86-fast-fp"
963    flag to "true".  Everything seems to compile to a fixpoint ok, and
964    "mandelbrot" speeds up by about 15%.
965    
966    ----------------------------------------------------------------------
967    Name: Matthias Blume
968    Date: 2001/09/13 11:20:00 EDT
969    Tag: blume-20010913-minimal
970    Description:
971    
972    1. Stefan Monnier's patch to fix a miscompilation problem that
973       was brought to light by John Reppy's work on Moby.
974    
975    2. Implemented a minimal "structure Compiler" that contains just
976       "version" and "architecture".  The minimal version will be
977       available when the full version is not.  This is for backward-
978       compatibility with code that wants to test Compiler.version.
979    
980    ----------------------------------------------------------------------
981    Name: Matthias Blume
982    Date: 2001/08/28 14:03:00 EDT
983    Tag: blume-20010828-ml-lex
984    Description:
985    
986    Fix for bug 1581, received from Neophytos Michael.
987    
988    ----------------------------------------------------------------------
989    Name: Matthias Blume
990    Date: 2001/08/27 11:20:00 EDT
991    Tag: blume-20010827-readme11035
992    Description:
993    
994    Fleshed out the README file for 110.35.
995    
996    ----------------------------------------------------------------------
997    Name: Matthias Blume
998    Date: 2001/08/24 17:10:00 EDT
999    Tag: Release_110_35
1000    Description:
1001    
1002    New version number (110.35).  New bootfiles.
1003    
1004    ----------------------------------------------------------------------
1005    Name: Lal George
1006    Date: 2001/08/24 13:47:18 EDT 2001
1007    Tag: george-20010824-MLRISC-graphs
1008    Description:
1009    
1010     removed clusters from MLRISC completely and replaced with graphs.
1011    
1012    ----------------------------------------------------------------------
1013    Name: Matthias Blume
1014    Date: 2001/08/23 17:50:00 EDT
1015    Tag: blume-20010823-toplevel
1016    Description:
1017    
1018    - some reorganization of the code that implements various kinds of
1019      environments in the compiler (static, dynamic, symbolic, combined)
1020    - re-implemented the EnvRef module so that evalStream works properly
1021      (if the stream contains references to "use", "CM.make", etc.)
1022    - cleaned up evalloop.sml and interact.sml (but they need more cleaning)
1023    
1024    ----------------------------------------------------------------------
1025    Name: Matthias Blume
1026    Date: 2001/08/20 15:50 EDT
1027    Tag: blume20010820-slipup
1028    Description:
1029    
1030    I forgot to commit a few files.  Here they are...
1031    
1032    ----------------------------------------------------------------------
1033    Name: Matthias Blume
1034    Date: 2001/08/20 15:35:00 EDT
1035    Tag: blume-20010820-debugprof
1036    Description:
1037    
1038    !!!! NEW BOOTFILES !!!!
1039    
1040    This is another round of reorganizing the compiler sources.  This
1041    time the main goal was to factor out all the "instrumentation"
1042    passes (for profiling and backtracing) into their own library.
1043    The difficulty was to do it in such a way that it does not depend
1044    on elaborate.cm but only on elabdata.cm.
1045    
1046    Therefore there have been further changes to both elaborate.cm and
1047    elabdata.cm -- more "generic" things have been moved from the former
1048    to the latter.  As a result, I was forced to split the assignment
1049    of numbers indicating "primtyc"s into two portions: SML-generic and
1050    SML/NJ-specific.  Since it would have been awkward to maintain,
1051    I bit the bullet and actually _changed_ the mapping between these
1052    numbers and primtycs.  The bottom line of this is that you need
1053    a new set of bin- and bootfiles.
1054    
1055    I have built new bootfiles for all architectures, so doing a fresh
1056    checkout and config/install.sh should be all you need.
1057    
1058    The newly created library's name is
1059    
1060        $smlnj/viscomp/debugprof.cm
1061    
1062    and its sources live under
1063    
1064        src/compiler/DebugProf
1065    
1066    ----------------------------------------------------------------------
1067    Name: Matthias Blume
1068    Date: 2001/08/15 17:15:00 EDT
1069    Tag: blume-20010815-compreorg
1070    Description:
1071    
1072    This is a first cut at reorganizing the CM libraries that make up the
1073    core of the compiler.  The idea is to separate out pieces that could
1074    be used independently by tools, e.g., the parser, the typechecker, etc.
1075    
1076    The current status is a step in this direction, but it is not quite
1077    satisfactory yet.  Expect more changes in the future.
1078    
1079    Here is the current (new) organization...
1080    
1081        What used to be $smlnj/viscomp/core.cm is now divided into
1082        six CM libraries:
1083    
1084             $smlnj/viscomp/basics.cm
1085                           /parser.cm
1086                           /elabdata.cm
1087                           /elaborate.cm
1088                           /execute.cm
1089                           /core.cm
1090    
1091        The CM files for these libraries live under src/system/smlnj/viscomp.
1092        All these libraries are proxy libraries that contain precisely
1093        one CM library component.  Here are the locations of the components
1094        (all within the src/compiler tree):
1095    
1096             Basics/basics.cm
1097             Parse/parser.cm
1098             ElabData/elabdata.cm
1099             Elaborator/elaborate.cm
1100             Execution/execute.cm
1101             core.cm
1102    
1103         [This organization is the same that has been used already
1104         for a while for the architecture-specific parts of the visible
1105         compiler and for the old version of core.cm.]
1106    
1107         As you will notice, many source files have been moved from their
1108         respective original locations to a new home in one of the above
1109         subtrees.
1110    
1111         The division of labor between the new libraries is the following:
1112    
1113             basics.cm:
1114                - Simple, basic definitions that pertain to many (or all) of
1115                  the other libraries.
1116             parser.cm:
1117                - The SML parser, producing output of type Ast.dec.
1118                - The type family for Ast is also defined and exported here.
1119             elabdata.cm:
1120                - The datatypes that describe input and output of the elaborator.
1121                  This includes types, absyn, and static environments.
1122             elaborator.cm:
1123                - The SML/NJ type checker and elaborator.
1124                  This maps an Ast.dec (with a given static environment) to
1125                  an Absyn.dec (with a new static environment).
1126                - This libraries implements certain modules that used to be
1127                  structures as functors (to remove dependencies on FLINT).
1128             execute.cm:
1129                - Everything having to do with executing binary code objects.
1130                - Dynamic environments.
1131             core.cm:
1132                - SML/NJ-specific instantiations of the elaborator and MLRISC.
1133                - Top-level modules.
1134                - FLINT (this should eventually become its own library)
1135    
1136    Notes:
1137    
1138    I am not 100% happy with the way I separated the elaborator (and its
1139    data structures) from FLINT.  Two instances of the same problem:
1140    
1141        1. Data structures contain certain fields that carry FLINT-specific
1142           information.  I hacked around this using exn and the property list
1143           module from smlnj-lib.  But the fact that there are middle-end
1144           specific fields around at all is a bit annoying.
1145    
1146        2. The elaborator calculates certain FLINT-related information.  I tried
1147           to make this as abstract as I could using functorization, but, again,
1148           the fact that the elaborator has to perform calculations on behalf
1149           of the middle-end at all is not nice.
1150    
1151        3. Having to used exn and property lists is unfortunate because it
1152           weakens type checking.  The other alternative (parameterizing
1153           nearly *everything*) is not appealing, though.
1154    
1155    I removed the "rebinding =" warning hack because due to the new organization
1156    it was awkward to maintain it.  As a result, the compiler now issues some of
1157    these warnings when compiling init.cmi during bootstrap compilation. On
1158    the plus side, you also get a warning when you do, for example:
1159       val op = = Int32.+
1160    which was not the case up to now.
1161    
1162    I placed "assign" and "deref" into the _Core structure so that the
1163    code that deals with the "lazy" keyword can find them there.  This
1164    removes the need for having access to the primitive environment
1165    during elaboration.
1166    
1167    ----------------------------------------------------------------------
1168    Name: Matthias Blume
1169    Date: 2001/08/13
1170    Tag: blume-20010813-closures
1171    Description:
1172    
1173    This fix was sent to us by Zhong Shao.  It is supposed to improve the
1174    performance of certain loops by avoiding needless closure allocation.
1175    
1176    ----------------------------------------------------------------------
1177    Name: Lal George
1178    Date: 2001/07/31 10:03:23 EDT 2001
1179    Tag: george-20010731-x86-fmalloc
1180    Description: Fixed bug in x86 calls
1181    
1182        There was a bug where call instructions would mysteriously
1183        vanish. The call instruction had to be one that returned
1184        a floating point value.
1185    
1186    ----------------------------------------------------------------------
1187    Name: Lal George
1188    Date: 2001/07/19 16:36:29 EDT 2001
1189    Tag: george-20010719-simple-cells
1190    Description:
1191    
1192    I have dramatically simplified the interface for CELLS in MLRISC.
1193    
1194    In summary, the cells interface is broken up into three parts:
1195    
1196      1. CellsBasis : CELLS_BASIS
1197    
1198            CellsBasis is a top level structure and common for all
1199            architectures.  it contains the definitions of basic datatypes
1200            and utility  functions over these types.
1201    
1202      2. functor Cells() : CELLS
1203    
1204            Cells generates an interface for CELLS that incorporates the
1205            specific resources on the target architecture, such as the
1206            presence of special register classes, their number and size,
1207            and various useful substructures.
1208    
1209      3. <ARCH>CELLS
1210    
1211            e.g. SparcCells: SPARCCELLS
1212    
1213            <ARCH>CELLS usually contains additional bindings for special
1214            registers  on the architecture, such as:
1215    
1216                    val r0 : cell           (* register zero *)
1217                    val y : cell            (* Y register *)
1218                    val psr : cell          (* processor status register *)
1219                    ...
1220    
1221            The structure returned by applying the Cells functor is opened
1222            in this interface.
1223    
1224    The main implication of all this is that the datatypes for cells is
1225    split between CellsBasis and CELLS -- a fairly simple change for user
1226    code.
1227    
1228    In the old scheme the CELLS interface had a definitional binding of
1229    the form:
1230    
1231            signature CELLS = sig
1232    
1233               structure CellsBasis = CellsBasis
1234    
1235               ...
1236    
1237            end
1238    
1239    With all the sharing constraints that goes on in MLRISC, this old
1240    design  quickly leads to errors such as:
1241    
1242            "structure definition spec inside of sharing ... "
1243    
1244    
1245    and appears to require an unacceptable amount of sharing and where
1246    constraint hackery.
1247    
1248    I think this error message (the interaction of definitional specs and
1249    sharing) requires more explanation on our web page.
1250    
1251    ----------------------------------------------------------------------
1252    Name: Matthias Blume
1253    Date: 2001/07/19 15:00:00 EDT
1254    Tag: blume-20010719-libreorg
1255    Description:
1256    
1257    This update puts together a fairly extensive but straightforward change
1258    to the way the libraries that implement the interactive system are
1259    organized:
1260    
1261       The biggest change is the elimination of structure Compiler.  As a
1262       replacement for this structure, there is now a CM library
1263       (known as $smlnj/compiler.cm or $smlnj/compiler/current.cm)
1264       that exports all the substructures of the original structure Compiler
1265       directly.  So instead of saying Compiler.Foo.bar one now simply
1266       says Foo.bar.  (The CM libraries actually export a collection of
1267       structures that is richer than the collection of substructures of
1268       structure Compiler.)
1269    
1270       To make the transition smooth, there is a separate library called
1271       $smlnj/compiler/compiler.cm which puts together and exports the
1272       original structure Compiler (or at least something very close to it).
1273    
1274       There are five members of the original structure Compiler
1275       that are not exported directly but which instead became members
1276       of a new structure Backend (described by signature BACKEND).  These are:
1277       structure Profile (: PROFILE), structure Compile (: COMPILE), structure
1278       Interact (: INTERACT), structure Machine (: MACHINE), and val
1279       architecture (: string).
1280    
1281       Structure Compiler.Version has become structure CompilerVersion.
1282    
1283       Cross-compilers for alpha32, hppa, ppc, sparc, and x86 are provided
1284       by $smlnj/compiler/<arch>.cm where <arch> is alpha32, hppa, ppc, sparc,
1285       or x86, respectively.
1286       Each of these exports the same frontend structures that
1287       $smlnj/compiler.cm exports.  But they do not have a structure Backend
1288       and instead export some structure <Arch>Backend where <Arch> is Alpha32,
1289       Hppa, PPC, Sparc, or X86, respectively.
1290    
1291       Library $smlnj/compiler/all.cm exports the union of the exports of
1292       $smlnj/compiler/<arch>.cm
1293    
1294       There are no structures <Arch>Compiler anymore, use
1295       $smlnj/compiler/<arch>.cm instead.
1296    
1297       Library host-compiler-0.cm is gone.  Instead, the internal library
1298       that instantiates CM is now called cm0.cm.  Selection of the host
1299       compiler (backend) is no longer done here but. (Responsibility for it
1300       now lies with $smlnj/compiler/current.cm.  This seems to be more
1301       logical.)
1302    
1303       Many individual files have been moved or renamed.  Some files have
1304       been split into multiple files, and some "dead" files have been deleted.
1305    
1306    Aside from these changes to library organization, there are also changes
1307    to the way the code itself is organized:
1308    
1309       Structure Binfile has been re-implemented in such a way that it no
1310       longer needs any knowledge of the compiler.  It exclusively deals
1311       with the details of binfile layout.  It no longer invokes the
1312       compiler (for the purpose of creating new prospective binfile
1313       content), and it no longer has any knowledge of how to interpret
1314       pickles.
1315    
1316       Structure Compile (: COMPILE) has been stripped down to the bare
1317       essentials of compilation.  It no longer deals with linking/execution.
1318       The interface has been cleaned up considerably.
1319    
1320       Utility routines for dealing with linking and execution have been
1321       moved into their own substructures.
1322    
1323       (The ultimate goal of these changes is to provide a light-weight
1324       binfile loader/linker (at least for, e.g., stable libraries) that
1325       does not require CM or the compiler to be present.)
1326    
1327    CM documentation has been updated to reflect the changes to library
1328    organization.
1329    
1330    ----------------------------------------------------------------------
1331    Name: Matthias Blume
1332    Date: 2001/07/10 17:30:00 EDT
1333    Tag: Release_110_34
1334    Description:
1335    
1336    Minor tweak to 110.34 (re-tagged):
1337    
1338      - README.html file added to CVS repository
1339      - runtime compiles properly under FreeBSD 3.X and 4.X
1340    
1341    ----------------------------------------------------------------------
1342    Name: Matthias Blume
1343    Date: 2001/07/10 17:30:00 EDT
1344    Tag: Release_110_34
1345    Description:
1346    
1347    New version number (110.34). New bootfiles.
1348    
1349    ----------------------------------------------------------------------
1350    Name: Matthias Blume
1351    Date: 2001/07/09 16:00:00 EDT
1352    Tag: blume-20010709-more-varargs
1353    Description:
1354    
1355    I changed the handling of varargs in ml-nlffigen again:
1356    The ellipsis ... will now simply be ignored (with an accompanying warning).
1357    
1358    The immediate effect is that you can actually call a varargs function
1359    from ML -- but you can't actually supply any arguments beyond the ones
1360    specified explicitly.  (For example, you can call printf with its format
1361    string, but you cannot pass additional arguments.)
1362    
1363    This behavior is only marginally more useful than the one before, but
1364    it has the advantage that a function or, more importantly, a function
1365    type never gets dropped on the floor, thus avoiding follow-up problems with
1366    other types that refer to the offending one.
1367    
1368    ----------------------------------------------------------------------
1369    Name: Matthias Blume
1370    Date: 2001/07/09 11:25:00 EDT
1371    Tag: blume-20010709-varargs
1372    Description:
1373    
1374    1. ckit-lib.cm now exports structure Error
1375    2. ml-nlffigen reports occurences of "..." (i.e., varargs function types)
1376       with a warning accompanied by a source location.  Moreover, it
1377       merely skips the offending function or type and proceeds with the
1378       rest of its work.u  As a result, one can safely feed C code containing
1379       "..." to ml-nlffigen.
1380    3. There are some internal improvements to CM, providing slightly
1381       more general string substitutions in the tools subsystem.
1382    
1383    ----------------------------------------------------------------------
1384    Name: Matthias Blume
1385    Date: 2001/06/27 15:10:00 EDT
1386    Tag: blume-20010627-concur
1387    Description:
1388    
1389    Fixed a small bug in CM's handling of parallel compilation.
1390    (You could observe the bug by Control-C-interrupting an ordinary
1391    CMB.make or CM.stabilize and then attaching some compile servers.
1392    The result was that all of a sudden the previously interrupted
1393    compilation would continue on its own.  This was because of
1394    an over-optimization: CM did not bother to clean out certain queues
1395    when no servers were attached "anyway", resulting in the contents
1396    of these queues to grab control when new servers did get attached.)
1397    
1398    There is also another minor update to the CM manual.
1399    
1400    ----------------------------------------------------------------------
1401    Name: Matthias Blume
1402    Date: 2001/06/26 16:15:00 EDT
1403    Tag: blume-20010626-cmdoc
1404    Description:
1405    
1406    Minor typo fixed in CM manual (syntax diagram for libraries).
1407    
1408    ----------------------------------------------------------------------
1409    Name: Matthias Blume
1410    Date: 2001/06/25 22:55:00 EDT
1411    Tag: blume-20010625-x86pc
1412    Description:
1413    
1414    Fixed a nasty bug in the X86 assembly code that caused signal
1415    handlers to fail (crash) randomly.
1416    
1417    ----------------------------------------------------------------------
1418    Name: Matthias Blume
1419    Date: 2001/06/25 12:05:00 EDT
1420    Tag: blume-20010625-nlffigen
1421    Description:
1422    
1423    This update fixes a number of minor bugs in ml-nlffigen as reported by
1424    Nick Carter <nbc@andrew.cmu.edu>.
1425    
1426      1. Silly but ok typedefs of the form "typedef void myvoid;" are now accepted.
1427      2. Default names for generated files are now derived from the name of
1428         the C file *without its directory*.  In particular, this causes generated
1429         files to be placed locally even if the C file is in some system directory.
1430      3. Default names for generated signatures and structures are also derived
1431         from the C file name without its directory.  This avoids silly things
1432         like "structure GL/GL".
1433         (Other silly names are still possible because ml-nlffigen does not do
1434          a thorough check of whether generated names are legal ML identifiers.
1435          When in doubt, use command line arguments to force particular names.)
1436    
1437    ----------------------------------------------------------------------
1438    Name: Matthias Blume
1439    Date: 2001/06/21 12:25:00 EDT
1440    Tag: blume-20010621-eXene
1441    Description:
1442    
1443    eXene now compiles and (sort of) works again.
1444    
1445    The library name (for version > 110.33) is $/eXene.cm.
1446    
1447    I also added an new example in src/eXene/examples/nbody.  See the
1448    README file there for details.
1449    
1450    ----------------------------------------------------------------------
1451    Name: Matthias Blume
1452    Date: 2001/06/20 16:40:00 EDT
1453    Tag: blume-20010620-cml
1454    Description:
1455    
1456    CML now compiles and works again.
1457    
1458    Libraries (for version > 110.33):
1459    
1460      $cml/cml.cm            Main CML library.
1461      $cml/basis.cm          CML's version of $/basis.cm.
1462      $cml/cml-internal.cm   Internal helper library.
1463      $cml/core-cml.cm       Internal helper library.
1464      $cml-lib/trace-cml.cm  Tracing facility.
1465      $cml-lib/smlnj-lib.cm  CML's version of $/smlnj-lib.cm
1466    
1467    The installer (config/install.sh) has been taught how to properly
1468    install this stuff.
1469    
1470    ----------------------------------------------------------------------
1471    Name: Matthias Blume
1472    Date: 2001/06/19 17:55:00 EDT
1473    Tag: blume-20010619-instantiate
1474    Description:
1475    
1476    This un-breaks the fix for bug 1432.
1477    (The bug was originally fixed in 110.9 but I broke it again some
1478    time after that.)
1479    
1480    ----------------------------------------------------------------------
1481    Name: Matthias Blume
1482    Date: 2001/06/19 17:25:00 EDT
1483    Tag: blume-20010619-signals
1484    Description:
1485    
1486    This should (hopefully) fix the long-standing signal handling bug.
1487    (The runtime system was constructing a continuation record with an
1488    incorrect descriptor which would cause the GC to drop data on the floor...)
1489    
1490    ----------------------------------------------------------------------
1491    Name: Matthias Blume
1492    Date: 2001/06/15 15:05:00 EDT
1493    Tag: blume-20010615-moresparc
1494    Description:
1495    
1496    Here is a short late-hour update related to Sparc c-calls:
1497    
1498     -- made handling of double-word arguments a bit smarter
1499    
1500     -- instruction selection phase tries to collapse certain clumsily
1501        constructed ML-Trees; typical example:
1502    
1503            ADD(ty,ADD(_,e,LI d1),LI d2)  ->  ADD(ty,e,LI(d1+d2))
1504    
1505        This currently has no further impact on SML/NJ since mlriscGen does
1506        not seem to generate such patterns in the first place, and c-calls
1507        (which did generate them in the beginning) has meanwhile been fixed
1508        so as to avoid them as well.
1509    
1510    ----------------------------------------------------------------------
1511    Name: Matthias Blume
1512    Date: 2001/06/15 15:05:00 EDT
1513    Tag: blume-20010615-sparc
1514    Description:
1515    
1516    The purpose of this update is to provide an implementation of NLFFI
1517    on Sparc machines.
1518    
1519    Here are the changes in detail:
1520    
1521       * src/MLRISC/sparc/c-calls/sparc-c-calls.sml is a new file containing
1522       the Sparc implementation of the c-calls API.
1523       * The Sparc backend of SML/NJ has been modified to uniformely use %fp
1524       for accessing the ML frame.  Thus, we have a real frame pointer and
1525       can freely modify %sp without need for an omit-frame-ptr phase.
1526       The vfp logic in src/compiler/CodeGen/* has been changed to accomodate
1527       this case.
1528       * ml-nlffigen has been taught to produce code for different architectures
1529       and calling conventions.
1530       * In a way similar to what was done in the x86 case, the Sparc
1531       backend uses its own specific extension to mltree.  (For example,
1532       it needs to be able to generate UNIMP instructions which are part
1533       of the calling convention.)
1534       * ml-nlffi-lib was reorganized to make it more modular (in particular,
1535       to make it easier to plug in new machine- and os-dependent parts).
1536    
1537    There are some other fairly unrelated bug fixes and cleanups as well:
1538    
1539       * I further hacked the .cm files for MLRISC tools (like MDLGen) so
1540       that they properly share their libraries with existing SML/NJ libraries.
1541       * I fixed a minor cosmetic bug in CM, supressing certain spurious
1542       follow-up error messages.
1543       * Updates to CM/CMB documentation.
1544    
1545    TODO items:
1546    
1547       * MLRISC should use a different register as its asmTemp on the Sparc.
1548         (The current %o2 is a really bad choice because it is part of the
1549         calling conventions, so things might interfere in unexpected ways.)
1550    
1551    ----------------------------------------------------------------------
1552    Name: Matthias Blume
1553    Date: 2001/06/07
1554    Tag: blume-20010607-calls
1555    Description:
1556    
1557    A number of internal changes related to C calls and calling conventions:
1558    
1559    1. ML-Tree CALL statements now carry a "pops" field.  It indicates the
1560       number of bytes popped implicitly (by the callee).  In most cases
1561       this field is 0 but on x86/win32 it is some non-zero value.  This
1562       is information provided for the benefit of the "omit-frameptr" pass.
1563    2. The CALL instruction on the x86 carries a similar "pops" field.
1564       The instruction selection phase copies its value from the ML-Tree
1565       CALL statement.
1566    3. On all other architectures, the instruction selection phase checks
1567       whether "pops=0" and complains if not.
1568    4. The c-calls implementation for x86 now accepts two calling conventions:
1569       "ccall" and "stdcall".  When "ccall" is selected, the caller cleans
1570       up after the call and pops is set to 0.  For "stdcall", the caller
1571       does nothing, leaving the cleanup to the callee; pops is set to
1572       the number of bytes that were pushed onto the stack.
1573    5. The cproto decoder (compiler/Semant/types/cproto.sml) now can
1574       distinguish between "ccall" and "stdcall".
1575    6. The UNIMP instruction has been added to the supported Sparc instruction
1576       set. (This is needed for implementing the official C calling convention
1577       on this architecture.)
1578    7. I fixed some of the .cm files under src/MLRISC/Tools to make them
1579       work with the latest CM.
1580    
1581    ----------------------------------------------------------------------
1582    Name: Matthias Blume
1583    Date: 2001/06/05 15:10:00 EDT
1584    Tag: blume-20010605-cm-index
1585    Description:
1586    
1587    0. The "lambdasplit" parameter for class "sml" in CM has been documented.
1588    
1589    1. CM can now generate "index files".  These are human-readable files
1590       that list on a per-.cm-file basis each toplevel symbol defined or
1591       imported.  The location of the index file for
1592       <p>/<d>.cm is <p>/CM/INDEX/<d>.cm.
1593       To enable index-file generation, set CM.Control.generate_index to true
1594       or export an environment-symbol: export CM_GENERATE_INDEX=true.
1595    
1596       The CM manual has been updated accordingly.
1597    
1598    2. I made some slight modifications to the c-calls API in MLRISC.
1599    
1600         a) There is now a callback to support saving/restoring of
1601            dedicated but caller-save registers around the actual call
1602            instruction.
1603         b) One can optionally specify a comment-annotation for the
1604            call instruction.
1605    
1606    3. SML/NJ (mlriscGen.sml) uses this new API for the rawccall primop.
1607       (For example, the comment annotation shows the C prototype of
1608        the function being called.)
1609    
1610    ----------------------------------------------------------------------
1611    Name: Matthias Blume
1612    Date: 2001/06/01 13:30:00 EDT
1613    Tag: blume-20010601-nlffi-cleanup
1614    Description:
1615    
1616    This is mostly a cleanup of MLFFI stuff:
1617    
1618       - some signature files have been put into a more exposed place
1619       - the ugly 'f type parameter is gone (simplifies types tremendously!)
1620       - ml-nlffigen changed accordingly
1621       - tutorial updated
1622    
1623    Other changes:
1624    
1625       - author's affiliation in CM manual(s) updated
1626       - some more recognized keywords added to Allen's sml.sty
1627    
1628    ----------------------------------------------------------------------
1629    Name: Matthias Blume
1630    Date: 2001/05/25 15:30:00 EDT
1631    Tag: blume-20010525-iptr
1632    Description:
1633    
1634      - put the official 110.33-README (as it appears on the ftp server) under
1635        CVS
1636      - fixed a small bug related to incomplete pointer types in
1637        ml-nlffigen
1638      - small cosmetic change to the ml-nlffi-lib's "arr" type constructor
1639        (it does not need the 'f type parameter)
1640    
1641    ----------------------------------------------------------------------
1642    Name: Matthias Blume
1643    Date: 2001/05/23 14:30:00 EDT
1644    Tag: Release_110_33
1645    Description:
1646    
1647    New version number (110.33).  New bootfiles.
1648    
1649    ----------------------------------------------------------------------
1650    Name: Matthias Blume
1651    Date: 2001/05/22 18:06:00 EDT
1652    Tag: blume-20010522-targets
1653    Description:
1654    
1655    Made install.sh use file config/targets.customized if it exists, falling
1656    back to config/targets if it doesn't.  This way one can have a customized
1657    version of the targets file without touching the "real thing", thus
1658    eliminating the constant fear of accidentally checking something bogus
1659    back into the CVS repository...  (File config/targets.customized must
1660    not be added to the repository!)
1661    
1662    ----------------------------------------------------------------------
1663    Name: Matthias Blume
1664  Date: 2001/05/22 16:30:00 EDT  Date: 2001/05/22 16:30:00 EDT
1665  Tag: blume-20010522-minitut  Tag: blume-20010522-minitut
1666  Description:  Description:
# Line 2257  Line 3904 
3904       elaborator). There were a lot of changes during my "linkpath" trials       elaborator). There were a lot of changes during my "linkpath" trials
3905       that could have been reverted to their original state but weren't.       that could have been reverted to their original state but weren't.
3906       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
3907       than what was strictly necessary...  (I _did_ resist the tempation       than what was strictly necessary...  (I _did_ resist the temptation
3908       of doing any "global reformatting" to avoid an untimely death at       of doing any "global reformatting" to avoid an untimely death at
3909       Dave's hands. :)       Dave's hands. :)
3910    

Legend:
Removed from v.832  
changed lines
  Added in v.1078

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