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 644, Fri May 12 09:18:31 2000 UTC revision 915, Fri Aug 24 18:21:00 2001 UTC
# Line 11  Line 11 
11  Date:  Date:
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    
15    ----------------------------------------------------------------------
16    Name: Lal George
17    Date: 2001/08/24 13:47:18 EDT 2001
18    Tag: george-20010824-MLRISC-graphs
19    Description:
20    
21     removed clusters from MLRISC completely and replaced with graphs.
22    
23    ----------------------------------------------------------------------
24    Name: Matthias Blume
25    Date: 2001/08/23 17:50:00 EDT
26    Tag: blume-20010823-toplevel
27    Description:
28    
29    - some reorganization of the code that implements various kinds of
30      environments in the compiler (static, dynamic, symbolic, combined)
31    - re-implemented the EnvRef module so that evalStream works properly
32      (if the stream contains references to "use", "CM.make", etc.)
33    - cleaned up evalloop.sml and interact.sml (but they need more cleaning)
34    
35    ----------------------------------------------------------------------
36    Name: Matthias Blume
37    Date: 2001/08/20 15:50 EDT
38    Tag: blume20010820-slipup
39    Description:
40    
41    I forgot to commit a few files.  Here they are...
42    
43    ----------------------------------------------------------------------
44    Name: Matthias Blume
45    Date: 2001/08/20 15:35:00 EDT
46    Tag: blume-20010820-debugprof
47    Description:
48    
49    !!!! NEW BOOTFILES !!!!
50    
51    This is another round of reorganizing the compiler sources.  This
52    time the main goal was to factor out all the "instrumentation"
53    passes (for profiling and backtracing) into their own library.
54    The difficulty was to do it in such a way that it does not depend
55    on elaborate.cm but only on elabdata.cm.
56    
57    Therefore there have been further changes to both elaborate.cm and
58    elabdata.cm -- more "generic" things have been moved from the former
59    to the latter.  As a result, I was forced to split the assignment
60    of numbers indicating "primtyc"s into two portions: SML-generic and
61    SML/NJ-specific.  Since it would have been awkward to maintain,
62    I bit the bullet and actually _changed_ the mapping between these
63    numbers and primtycs.  The bottom line of this is that you need
64    a new set of bin- and bootfiles.
65    
66    I have built new bootfiles for all architectures, so doing a fresh
67    checkout and config/install.sh should be all you need.
68    
69    The newly created library's name is
70    
71        $smlnj/viscomp/debugprof.cm
72    
73    and its sources live under
74    
75        src/compiler/DebugProf
76    
77    ----------------------------------------------------------------------
78    Name: Matthias Blume
79    Date: 2001/08/15 17:15:00 EDT
80    Tag: blume-20010815-compreorg
81    Description:
82    
83    This is a first cut at reorganizing the CM libraries that make up the
84    core of the compiler.  The idea is to separate out pieces that could
85    be used independently by tools, e.g., the parser, the typechecker, etc.
86    
87    The current status is a step in this direction, but it is not quite
88    satisfactory yet.  Expect more changes in the future.
89    
90    Here is the current (new) organization...
91    
92        What used to be $smlnj/viscomp/core.cm is now divided into
93        six CM libraries:
94    
95             $smlnj/viscomp/basics.cm
96                           /parser.cm
97                           /elabdata.cm
98                           /elaborate.cm
99                           /execute.cm
100                           /core.cm
101    
102        The CM files for these libraries live under src/system/smlnj/viscomp.
103        All these libraries are proxy libraries that contain precisely
104        one CM library component.  Here are the locations of the components
105        (all within the src/compiler tree):
106    
107             Basics/basics.cm
108             Parse/parser.cm
109             ElabData/elabdata.cm
110             Elaborator/elaborate.cm
111             Execution/execute.cm
112             core.cm
113    
114         [This organization is the same that has been used already
115         for a while for the architecture-specific parts of the visible
116         compiler and for the old version of core.cm.]
117    
118         As you will notice, many source files have been moved from their
119         respective original locations to a new home in one of the above
120         subtrees.
121    
122         The division of labor between the new libraries is the following:
123    
124             basics.cm:
125                - Simple, basic definitions that pertain to many (or all) of
126                  the other libraries.
127             parser.cm:
128                - The SML parser, producing output of type Ast.dec.
129                - The type family for Ast is also defined and exported here.
130             elabdata.cm:
131                - The datatypes that describe input and output of the elaborator.
132                  This includes types, absyn, and static environments.
133             elaborator.cm:
134                - The SML/NJ type checker and elaborator.
135                  This maps an Ast.dec (with a given static environment) to
136                  an Absyn.dec (with a new static environment).
137                - This libraries implements certain modules that used to be
138                  structures as functors (to remove dependencies on FLINT).
139             execute.cm:
140                - Everything having to do with executing binary code objects.
141                - Dynamic environments.
142             core.cm:
143                - SML/NJ-specific instantiations of the elaborator and MLRISC.
144                - Top-level modules.
145                - FLINT (this should eventually become its own library)
146    
147    Notes:
148    
149    I am not 100% happy with the way I separated the elaborator (and its
150    data structures) from FLINT.  Two instances of the same problem:
151    
152        1. Data structures contain certain fields that carry FLINT-specific
153           information.  I hacked around this using exn and the property list
154           module from smlnj-lib.  But the fact that there are middle-end
155           specific fields around at all is a bit annoying.
156    
157        2. The elaborator calculates certain FLINT-related information.  I tried
158           to make this as abstract as I could using functorization, but, again,
159           the fact that the elaborator has to perform calculations on behalf
160           of the middle-end at all is not nice.
161    
162        3. Having to used exn and property lists is unfortunate because it
163           weakens type checking.  The other alternative (parameterizing
164           nearly *everything*) is not appealing, though.
165    
166    I removed the "rebinding =" warning hack because due to the new organization
167    it was awkward to maintain it.  As a result, the compiler now issues some of
168    these warnings when compiling init.cmi during bootstrap compilation. On
169    the plus side, you also get a warning when you do, for example:
170       val op = = Int32.+
171    which was not the case up to now.
172    
173    I placed "assign" and "deref" into the _Core structure so that the
174    code that deals with the "lazy" keyword can find them there.  This
175    removes the need for having access to the primitive environment
176    during elaboration.
177    
178    ----------------------------------------------------------------------
179    Name: Matthias Blume
180    Date: 2001/08/13
181    Tag: blume-20010813-closures
182    Description:
183    
184    This fix was sent to us by Zhong Shao.  It is supposed to improve the
185    performance of certain loops by avoiding needless closure allocation.
186    
187    ----------------------------------------------------------------------
188    Name: Lal George
189    Date: 2001/07/31 10:03:23 EDT 2001
190    Tag: george-20010731-x86-fmalloc
191    Description: Fixed bug in x86 calls
192    
193        There was a bug where call instructions would mysteriously
194        vanish. The call instruction had to be one that returned
195        a floating point value.
196    
197    ----------------------------------------------------------------------
198    Name: Lal George
199    Date: 2001/07/19 16:36:29 EDT 2001
200    Tag: george-20010719-simple-cells
201    Description:
202    
203    I have dramatically simplified the interface for CELLS in MLRISC.
204    
205    In summary, the cells interface is broken up into three parts:
206    
207      1. CellsBasis : CELLS_BASIS
208    
209            CellsBasis is a top level structure and common for all
210            architectures.  it contains the definitions of basic datatypes
211            and utility  functions over these types.
212    
213      2. functor Cells() : CELLS
214    
215            Cells generates an interface for CELLS that incorporates the
216            specific resources on the target architecture, such as the
217            presence of special register classes, their number and size,
218            and various useful substructures.
219    
220      3. <ARCH>CELLS
221    
222            e.g. SparcCells: SPARCCELLS
223    
224            <ARCH>CELLS usually contains additional bindings for special
225            registers  on the architecture, such as:
226    
227                    val r0 : cell           (* register zero *)
228                    val y : cell            (* Y register *)
229                    val psr : cell          (* processor status register *)
230                    ...
231    
232            The structure returned by applying the Cells functor is opened
233            in this interface.
234    
235    The main implication of all this is that the datatypes for cells is
236    split between CellsBasis and CELLS -- a fairly simple change for user
237    code.
238    
239    In the old scheme the CELLS interface had a definitional binding of
240    the form:
241    
242            signature CELLS = sig
243    
244               structure CellsBasis = CellsBasis
245    
246               ...
247    
248            end
249    
250    With all the sharing constraints that goes on in MLRISC, this old
251    design  quickly leads to errors such as:
252    
253            "structure definition spec inside of sharing ... "
254    
255    
256    and appears to require an unacceptable amount of sharing and where
257    constraint hackery.
258    
259    I think this error message (the interaction of definitional specs and
260    sharing) requires more explanation on our web page.
261    
262    ----------------------------------------------------------------------
263    Name: Matthias Blume
264    Date: 2001/07/19 15:00:00 EDT
265    Tag: blume-20010719-libreorg
266    Description:
267    
268    This update puts together a fairly extensive but straightforward change
269    to the way the libraries that implement the interactive system are
270    organized:
271    
272       The biggest change is the elimination of structure Compiler.  As a
273       replacement for this structure, there is now a CM library
274       (known as $smlnj/compiler.cm or $smlnj/compiler/current.cm)
275       that exports all the substructures of the original structure Compiler
276       directly.  So instead of saying Compiler.Foo.bar one now simply
277       says Foo.bar.  (The CM libraries actually export a collection of
278       structures that is richer than the collection of substructures of
279       structure Compiler.)
280    
281       To make the transition smooth, there is a separate library called
282       $smlnj/compiler/compiler.cm which puts together and exports the
283       original structure Compiler (or at least something very close to it).
284    
285       There are five members of the original structure Compiler
286       that are not exported directly but which instead became members
287       of a new structure Backend (described by signature BACKEND).  These are:
288       structure Profile (: PROFILE), structure Compile (: COMPILE), structure
289       Interact (: INTERACT), structure Machine (: MACHINE), and val
290       architecture (: string).
291    
292       Structure Compiler.Version has become structure CompilerVersion.
293    
294       Cross-compilers for alpha32, hppa, ppc, sparc, and x86 are provided
295       by $smlnj/compiler/<arch>.cm where <arch> is alpha32, hppa, ppc, sparc,
296       or x86, respectively.
297       Each of these exports the same frontend structures that
298       $smlnj/compiler.cm exports.  But they do not have a structure Backend
299       and instead export some structure <Arch>Backend where <Arch> is Alpha32,
300       Hppa, PPC, Sparc, or X86, respectively.
301    
302       Library $smlnj/compiler/all.cm exports the union of the exports of
303       $smlnj/compiler/<arch>.cm
304    
305       There are no structures <Arch>Compiler anymore, use
306       $smlnj/compiler/<arch>.cm instead.
307    
308       Library host-compiler-0.cm is gone.  Instead, the internal library
309       that instantiates CM is now called cm0.cm.  Selection of the host
310       compiler (backend) is no longer done here but. (Responsibility for it
311       now lies with $smlnj/compiler/current.cm.  This seems to be more
312       logical.)
313    
314       Many individual files have been moved or renamed.  Some files have
315       been split into multiple files, and some "dead" files have been deleted.
316    
317    Aside from these changes to library organization, there are also changes
318    to the way the code itself is organized:
319    
320       Structure Binfile has been re-implemented in such a way that it no
321       longer needs any knowledge of the compiler.  It exclusively deals
322       with the details of binfile layout.  It no longer invokes the
323       compiler (for the purpose of creating new prospective binfile
324       content), and it no longer has any knowledge of how to interpret
325       pickles.
326    
327       Structure Compile (: COMPILE) has been stripped down to the bare
328       essentials of compilation.  It no longer deals with linking/execution.
329       The interface has been cleaned up considerably.
330    
331       Utility routines for dealing with linking and execution have been
332       moved into their own substructures.
333    
334       (The ultimate goal of these changes is to provide a light-weight
335       binfile loader/linker (at least for, e.g., stable libraries) that
336       does not require CM or the compiler to be present.)
337    
338    CM documentation has been updated to reflect the changes to library
339    organization.
340    
341    ----------------------------------------------------------------------
342    Name: Matthias Blume
343    Date: 2001/07/10 17:30:00 EDT
344    Tag: Release_110_34
345    Description:
346    
347    Minor tweak to 110.34 (re-tagged):
348    
349      - README.html file added to CVS repository
350      - runtime compiles properly under FreeBSD 3.X and 4.X
351    
352    ----------------------------------------------------------------------
353    Name: Matthias Blume
354    Date: 2001/07/10 17:30:00 EDT
355    Tag: Release_110_34
356    Description:
357    
358    New version number (110.34). New bootfiles.
359    
360    ----------------------------------------------------------------------
361    Name: Matthias Blume
362    Date: 2001/07/09 16:00:00 EDT
363    Tag: blume-20010709-more-varargs
364    Description:
365    
366    I changed the handling of varargs in ml-nlffigen again:
367    The ellipsis ... will now simply be ignored (with an accompanying warning).
368    
369    The immediate effect is that you can actually call a varargs function
370    from ML -- but you can't actually supply any arguments beyond the ones
371    specified explicitly.  (For example, you can call printf with its format
372    string, but you cannot pass additional arguments.)
373    
374    This behavior is only marginally more useful than the one before, but
375    it has the advantage that a function or, more importantly, a function
376    type never gets dropped on the floor, thus avoiding follow-up problems with
377    other types that refer to the offending one.
378    
379    ----------------------------------------------------------------------
380    Name: Matthias Blume
381    Date: 2001/07/09 11:25:00 EDT
382    Tag: blume-20010709-varargs
383    Description:
384    
385    1. ckit-lib.cm now exports structure Error
386    2. ml-nlffigen reports occurences of "..." (i.e., varargs function types)
387       with a warning accompanied by a source location.  Moreover, it
388       merely skips the offending function or type and proceeds with the
389       rest of its work.u  As a result, one can safely feed C code containing
390       "..." to ml-nlffigen.
391    3. There are some internal improvements to CM, providing slightly
392       more general string substitutions in the tools subsystem.
393    
394    ----------------------------------------------------------------------
395    Name: Matthias Blume
396    Date: 2001/06/27 15:10:00 EDT
397    Tag: blume-20010627-concur
398    Description:
399    
400    Fixed a small bug in CM's handling of parallel compilation.
401    (You could observe the bug by Control-C-interrupting an ordinary
402    CMB.make or CM.stabilize and then attaching some compile servers.
403    The result was that all of a sudden the previously interrupted
404    compilation would continue on its own.  This was because of
405    an over-optimization: CM did not bother to clean out certain queues
406    when no servers were attached "anyway", resulting in the contents
407    of these queues to grab control when new servers did get attached.)
408    
409    There is also another minor update to the CM manual.
410    
411    ----------------------------------------------------------------------
412    Name: Matthias Blume
413    Date: 2001/06/26 16:15:00 EDT
414    Tag: blume-20010626-cmdoc
415    Description:
416    
417    Minor typo fixed in CM manual (syntax diagram for libraries).
418    
419    ----------------------------------------------------------------------
420    Name: Matthias Blume
421    Date: 2001/06/25 22:55:00 EDT
422    Tag: blume-20010625-x86pc
423    Description:
424    
425    Fixed a nasty bug in the X86 assembly code that caused signal
426    handlers to fail (crash) randomly.
427    
428    ----------------------------------------------------------------------
429    Name: Matthias Blume
430    Date: 2001/06/25 12:05:00 EDT
431    Tag: blume-20010625-nlffigen
432    Description:
433    
434    This update fixes a number of minor bugs in ml-nlffigen as reported by
435    Nick Carter <nbc@andrew.cmu.edu>.
436    
437      1. Silly but ok typedefs of the form "typedef void myvoid;" are now accepted.
438      2. Default names for generated files are now derived from the name of
439         the C file *without its directory*.  In particular, this causes generated
440         files to be placed locally even if the C file is in some system directory.
441      3. Default names for generated signatures and structures are also derived
442         from the C file name without its directory.  This avoids silly things
443         like "structure GL/GL".
444         (Other silly names are still possible because ml-nlffigen does not do
445          a thorough check of whether generated names are legal ML identifiers.
446          When in doubt, use command line arguments to force particular names.)
447    
448    ----------------------------------------------------------------------
449    Name: Matthias Blume
450    Date: 2001/06/21 12:25:00 EDT
451    Tag: blume-20010621-eXene
452    Description:
453    
454    eXene now compiles and (sort of) works again.
455    
456    The library name (for version > 110.33) is $/eXene.cm.
457    
458    I also added an new example in src/eXene/examples/nbody.  See the
459    README file there for details.
460    
461    ----------------------------------------------------------------------
462    Name: Matthias Blume
463    Date: 2001/06/20 16:40:00 EDT
464    Tag: blume-20010620-cml
465    Description:
466    
467    CML now compiles and works again.
468    
469    Libraries (for version > 110.33):
470    
471      $cml/cml.cm            Main CML library.
472      $cml/basis.cm          CML's version of $/basis.cm.
473      $cml/cml-internal.cm   Internal helper library.
474      $cml/core-cml.cm       Internal helper library.
475      $cml-lib/trace-cml.cm  Tracing facility.
476      $cml-lib/smlnj-lib.cm  CML's version of $/smlnj-lib.cm
477    
478    The installer (config/install.sh) has been taught how to properly
479    install this stuff.
480    
481    ----------------------------------------------------------------------
482    Name: Matthias Blume
483    Date: 2001/06/19 17:55:00 EDT
484    Tag: blume-20010619-instantiate
485    Description:
486    
487    This un-breaks the fix for bug 1432.
488    (The bug was originally fixed in 110.9 but I broke it again some
489    time after that.)
490    
491    ----------------------------------------------------------------------
492    Name: Matthias Blume
493    Date: 2001/06/19 17:25:00 EDT
494    Tag: blume-20010619-signals
495    Description:
496    
497    This should (hopefully) fix the long-standing signal handling bug.
498    (The runtime system was constructing a continuation record with an
499    incorrect descriptor which would cause the GC to drop data on the floor...)
500    
501    ----------------------------------------------------------------------
502    Name: Matthias Blume
503    Date: 2001/06/15 15:05:00 EDT
504    Tag: blume-20010615-moresparc
505    Description:
506    
507    Here is a short late-hour update related to Sparc c-calls:
508    
509     -- made handling of double-word arguments a bit smarter
510    
511     -- instruction selection phase tries to collapse certain clumsily
512        constructed ML-Trees; typical example:
513    
514            ADD(ty,ADD(_,e,LI d1),LI d2)  ->  ADD(ty,e,LI(d1+d2))
515    
516        This currently has no further impact on SML/NJ since mlriscGen does
517        not seem to generate such patterns in the first place, and c-calls
518        (which did generate them in the beginning) has meanwhile been fixed
519        so as to avoid them as well.
520    
521    ----------------------------------------------------------------------
522    Name: Matthias Blume
523    Date: 2001/06/15 15:05:00 EDT
524    Tag: blume-20010615-sparc
525    Description:
526    
527    The purpose of this update is to provide an implementation of NLFFI
528    on Sparc machines.
529    
530    Here are the changes in detail:
531    
532       * src/MLRISC/sparc/c-calls/sparc-c-calls.sml is a new file containing
533       the Sparc implementation of the c-calls API.
534       * The Sparc backend of SML/NJ has been modified to uniformely use %fp
535       for accessing the ML frame.  Thus, we have a real frame pointer and
536       can freely modify %sp without need for an omit-frame-ptr phase.
537       The vfp logic in src/compiler/CodeGen/* has been changed to accomodate
538       this case.
539       * ml-nlffigen has been taught to produce code for different architectures
540       and calling conventions.
541       * In a way similar to what was done in the x86 case, the Sparc
542       backend uses its own specific extension to mltree.  (For example,
543       it needs to be able to generate UNIMP instructions which are part
544       of the calling convention.)
545       * ml-nlffi-lib was reorganized to make it more modular (in particular,
546       to make it easier to plug in new machine- and os-dependent parts).
547    
548    There are some other fairly unrelated bug fixes and cleanups as well:
549    
550       * I further hacked the .cm files for MLRISC tools (like MDLGen) so
551       that they properly share their libraries with existing SML/NJ libraries.
552       * I fixed a minor cosmetic bug in CM, supressing certain spurious
553       follow-up error messages.
554       * Updates to CM/CMB documentation.
555    
556    TODO items:
557    
558       * MLRISC should use a different register as its asmTemp on the Sparc.
559         (The current %o2 is a really bad choice because it is part of the
560         calling conventions, so things might interfere in unexpected ways.)
561    
562    ----------------------------------------------------------------------
563    Name: Matthias Blume
564    Date: 2001/06/07
565    Tag: blume-20010607-calls
566    Description:
567    
568    A number of internal changes related to C calls and calling conventions:
569    
570    1. ML-Tree CALL statements now carry a "pops" field.  It indicates the
571       number of bytes popped implicitly (by the callee).  In most cases
572       this field is 0 but on x86/win32 it is some non-zero value.  This
573       is information provided for the benefit of the "omit-frameptr" pass.
574    2. The CALL instruction on the x86 carries a similar "pops" field.
575       The instruction selection phase copies its value from the ML-Tree
576       CALL statement.
577    3. On all other architectures, the instruction selection phase checks
578       whether "pops=0" and complains if not.
579    4. The c-calls implementation for x86 now accepts two calling conventions:
580       "ccall" and "stdcall".  When "ccall" is selected, the caller cleans
581       up after the call and pops is set to 0.  For "stdcall", the caller
582       does nothing, leaving the cleanup to the callee; pops is set to
583       the number of bytes that were pushed onto the stack.
584    5. The cproto decoder (compiler/Semant/types/cproto.sml) now can
585       distinguish between "ccall" and "stdcall".
586    6. The UNIMP instruction has been added to the supported Sparc instruction
587       set. (This is needed for implementing the official C calling convention
588       on this architecture.)
589    7. I fixed some of the .cm files under src/MLRISC/Tools to make them
590       work with the latest CM.
591    
592    ----------------------------------------------------------------------
593    Name: Matthias Blume
594    Date: 2001/06/05 15:10:00 EDT
595    Tag: blume-20010605-cm-index
596    Description:
597    
598    0. The "lambdasplit" parameter for class "sml" in CM has been documented.
599    
600    1. CM can now generate "index files".  These are human-readable files
601       that list on a per-.cm-file basis each toplevel symbol defined or
602       imported.  The location of the index file for
603       <p>/<d>.cm is <p>/CM/INDEX/<d>.cm.
604       To enable index-file generation, set CM.Control.generate_index to true
605       or export an environment-symbol: export CM_GENERATE_INDEX=true.
606    
607       The CM manual has been updated accordingly.
608    
609    2. I made some slight modifications to the c-calls API in MLRISC.
610    
611         a) There is now a callback to support saving/restoring of
612            dedicated but caller-save registers around the actual call
613            instruction.
614         b) One can optionally specify a comment-annotation for the
615            call instruction.
616    
617    3. SML/NJ (mlriscGen.sml) uses this new API for the rawccall primop.
618       (For example, the comment annotation shows the C prototype of
619        the function being called.)
620    
621    ----------------------------------------------------------------------
622    Name: Matthias Blume
623    Date: 2001/06/01 13:30:00 EDT
624    Tag: blume-20010601-nlffi-cleanup
625    Description:
626    
627    This is mostly a cleanup of MLFFI stuff:
628    
629       - some signature files have been put into a more exposed place
630       - the ugly 'f type parameter is gone (simplifies types tremendously!)
631       - ml-nlffigen changed accordingly
632       - tutorial updated
633    
634    Other changes:
635    
636       - author's affiliation in CM manual(s) updated
637       - some more recognized keywords added to Allen's sml.sty
638    
639    ----------------------------------------------------------------------
640    Name: Matthias Blume
641    Date: 2001/05/25 15:30:00 EDT
642    Tag: blume-20010525-iptr
643    Description:
644    
645      - put the official 110.33-README (as it appears on the ftp server) under
646        CVS
647      - fixed a small bug related to incomplete pointer types in
648        ml-nlffigen
649      - small cosmetic change to the ml-nlffi-lib's "arr" type constructor
650        (it does not need the 'f type parameter)
651    
652    ----------------------------------------------------------------------
653    Name: Matthias Blume
654    Date: 2001/05/23 14:30:00 EDT
655    Tag: Release_110_33
656    Description:
657    
658    New version number (110.33).  New bootfiles.
659    
660    ----------------------------------------------------------------------
661    Name: Matthias Blume
662    Date: 2001/05/22 18:06:00 EDT
663    Tag: blume-20010522-targets
664    Description:
665    
666    Made install.sh use file config/targets.customized if it exists, falling
667    back to config/targets if it doesn't.  This way one can have a customized
668    version of the targets file without touching the "real thing", thus
669    eliminating the constant fear of accidentally checking something bogus
670    back into the CVS repository...  (File config/targets.customized must
671    not be added to the repository!)
672    
673    ----------------------------------------------------------------------
674    Name: Matthias Blume
675    Date: 2001/05/22 16:30:00 EDT
676    Tag: blume-20010522-minitut
677    Description:
678    
679    1. Bug fix in ml-nlffigen; now (hopefully) correctly handling
680       struct returns.
681    2. Added src/ml-nlffi-lib/Doc/mini-tutorial.txt.  This is some very
682       incomplete, preliminary documentation for NLFFI.
683    
684    ----------------------------------------------------------------------
685    Name: Matthias Blume
686    Date: 2001/05/14 11:30:00 EDT
687    Tag: blume-20010514-script
688    Description:
689    
690    Some bugs in install script fixed.
691    
692    In addition to that I also made a slight change to the NLFFI API:
693    Functors generated by ml-nlffigen now take the dynamic library as a
694    straight functor argument, not as a suspended one.  (The original
695    functor code used to force the suspension right away anyway, so there
696    was nothing gained by this complication of the interface.)
697    
698    ----------------------------------------------------------------------
699    Name: Matthias Blume
700    Date: 2001/05/11 14:35:00 EDT
701    Tag: blume-20010511-ml-nlffi
702    Description:
703    
704    I finally took the plunge and added my new FFI code to the main
705    repository.  For x86-linux it is now ready for prime-time.
706    
707    There are two new subdirectories of "src":
708    
709      - ml-nlffi-lib:
710           The utility library for programs using the FFI interface.
711           Here is the implementation of $/c.cm and its associated low-level
712           partners $/c-int.cm and $/memory.cm.
713      - ml-nlffigen:
714           A stand-alone program for generating ML glue code from C source
715           code.
716    
717    Building ml-nlffigen requires $/ckit-lib.cm.
718    
719    The config/install.sh script has been updates to do the Right Thing
720    (hopefully).
721    
722    Notice that the source tree for the C-Kit will not be put under "src"
723    but directly under the installation root directory.  (This is the
724    structure that currently exists on the CVS server when you check out
725    module "sml".)  Fortunately, config/install.sh knows about this oddity.
726    
727    Bugs: No documentation yet.
728    
729    ----------------------------------------------------------------------
730    Name: Matthias Blume
731    Date: 2001/05/09 16:35:00 EDT
732    Tag: blume-20010509-cpscontract
733    Description:
734    
735    Fixed a bug in the accounting code in cpsopt/contract.sml.  (The
736    wrapper/unwrapper elimination did not decrement usage counts and some
737    dead variables got overlooked by the dead-up logic.)
738    
739    ----------------------------------------------------------------------
740    Name: Lal George
741    Date: 2001/05/08  17:26:09 EDT
742    Tag: george-20010508-omit-frameptr
743    Description:
744    
745    Changes to implement the omit-frame-pointer optimization to support
746    raw C calls. For now, there is only support on the Intel x86, but
747    other architectures will follow as more experience is gained with this.
748    
749    
750    ----------------------------------------------------------------------
751    Name: Matthias Blume
752    Date: 2001/05/07 14:40:00 EDT
753    Tag: blume-20010507-proxies
754    Description:
755    
756    I made into "proxy libraries" all libraries that qualify for such a
757    change.  (A qualifying library is a library that has another library or
758    groups as its sole member and repeats that member's export list
759    verbatim.  A proxy library avoids this repetition by omitting its export
760    list, effectively inheriting the list that its (only) member exports.
761    See the CM manual for more explanation.)
762    The main effect is that explicit export lists for these libraries
763    do not have to be kepts in sync, making maintenance a bit easier.
764    
765    I also added copyright notices to many .cm-files.
766    
767    Last but not least, I made a new set of bootfiles.
768    
769    ----------------------------------------------------------------------
770    Name: Matthias Blume
771    Date: 2001/05/04 17:00:00 EDT
772    Tag: blume-20010504-cm-lsplit
773    Description:
774    
775    0. John merged pending changes to $/smlnj-lib.cm
776    
777    1. Allen's previous change accidentally backed out of one of Lal's
778       earlier changes.  I undid this mistake (re-introducing Lal's change).
779    
780    2. I used the new topOrder' function from graph-scc.sml (from $/smlnj-lib.cm)
781       within the compiler where applicable.  There is some code simplification
782       because of that.
783    
784    3. The "split" phase (in FLINT) is now part of the default list of phases.
785       Compiler.Control.LambdaSplitting.* can be used to globally control the
786       lambda-splitting (cross-module-inlining) engine.  In addition to that,
787       it can now also be controlled on a per-source basis: CM has been taught
788       a new tool parameter applicable to ML source files.
789    
790       - To turn lambda-splitting off completely:
791            local open Compiler.Control.LambdaSplitting in
792                val _ = set Off
793            end
794       - To make "no lambda-splitting" the global default (but allow per-source
795         overriding); this is the initial setting:
796            local open Compiler.Control.LambdaSplitting in
797                val _ = set (Default NONE)
798            end
799       - To make "lambda-splitting with aggressiveness a" the global default
800         (and allow per-source overriding):
801            local open Compiler.Control.LambdaSplitting in
802                val _ = set (Default (SOME a))
803            end
804    
805       - To turn lambda-splitting off for a given ML souce file (say: a.sml)
806         write (in the respective .cm-file):
807            a.sml (lambdasplitting:off)
808       - To turn lambda-splitting for a.sml on with minimal aggressiveness:
809            a.sml (lambdasplitting:on)
810       - To turn lambda-splitting for a.sml on with aggressiveness <a> (where
811         <a> is a decimal non-negative integer):
812            a.sml (lambdasplitting:<a>)
813       - To turn lambda-splitting for a.sml on with maximal aggressiveness:
814            a.sml (lambdasplitting:infinity)
815       - To use the global default for a.sml:
816            a.sml (lambdasplitting:default)
817         or simply
818            a.sml
819    
820    ----------------------------------------------------------------------
821    Name: Allen Leung
822    Date: 2001/05/04 01:57:00 EDT
823    Tag: leunga-20010504-sync
824    Description:
825    
826      MLRISC features.
827    
828      1. Fix to CMPXCHG instructions.
829      2. Changed RA interface to allow annotations in callbacks.
830      3. Added a new method to the stream interface to allow annotations updates.
831    
832    ----------------------------------------------------------------------
833    Name: Matthias Blume
834    Date: 2001/05/01 11:45:00 EDT
835    Tag: blume-20010501-pcedittmp
836    Description:
837    
838    Changed install.sh to use the current working directory instead of
839    /usr/tmp for a temporary file (pcedittmp).  The previous choice
840    of /usr/tmp caused trouble with MacOS X because of file premission
841    problems.
842    
843    ----------------------------------------------------------------------
844    Name: Matthias Blume
845    Date: 2001/04/20 11:10:00 EDT
846    Tag: blume-20010420-inMLflag
847    Description:
848    
849     - added vp_limitPtrMask to vproc-state.h
850       (for use by the raw-C-calls mechanism to implement proper interrupt
851        handling)
852     - made the ML compiler aware of various data-structure offsets so it
853       can generate code for accessing the vp_inML flag and vp_limitPtrMask
854     - tweaked mlriscGen.sml to have it emit interrupt-handling code for
855       raw C-calls
856    
857    ----------------------------------------------------------------------
858    Name: Lal George
859    Date: 2001/04/20 09:15:28 EDT
860    Tag: george-20010420-macosX
861    Description:
862    
863     - Changes to port to Mac OS X; Darwin.
864    
865     - In the process I found that sqrt was broken on the PPC, because the
866       fsqrt instruction is not implemented.
867    
868    ----------------------------------------------------------------------
869    Name: Matthias Blume
870    Date: 2001/04/18 12:45:00 EDT
871    Tag: blume-20010418-ccalls
872    Description:
873    
874     - fixed two off-by-4 errors in the x86-specific c-calls implementation
875       (this bug prevented structure arguments containing pointers from being
876        passed correctly)
877     - changed the raw-C-call code in mlriscGen.sml in such a way that
878       structure arguments are represented as a pointer to the beginning
879       of the structure (instead of having a series of synthesized arguments,
880       one for each structure member)
881    
882     - made makeml script's verbosity level configurable via environment
883       variable (MAKEML_VERBOSITY)
884    
885     - eliminated placeholder implementations for f32l, w16s, i16s, and f32s
886       in rawmem-x86.sml; we are now using the real thing
887    
888    ----------------------------------------------------------------------
889    Name: Matthias Blume
890    Date: 2001/03/22 16:25:00 EST
891    Tag: blume-20010322-bootfiles
892    Description:
893    
894    Created a new set of bootfiles (for your automatic installation convenience).
895    
896    ----------------------------------------------------------------------
897    Name: Matthias Blume
898    Date: 2001/03/22 15:10:00 EST
899    Tag: blume-20010322-rawmem-parcm
900    Description:
901    
902    1. All "raw memory access" primitives for the new FFI are implemented now
903       (at least on the x86).
904    2. Some further cleanup of CM's parallel make mechanism.
905    
906    ----------------------------------------------------------------------
907    Name: Matthias Blume
908    Date: 2001/03/19 17:53:00 EST
909    Tag: blume-20010319-parallel
910    Description:
911    
912    Parallel make (using compile servers) now works again.
913    
914    To this end, CM.stabilize and CMB.make have been modified to work in
915    two passes when compile servers are attached:
916       1. Compile everything, do not perform stabilization; this pass
917          uses compile servers
918       2. Stabilize everything; this pass does not use compile servers
919    If there are no compile servers, the two passes are combined into one
920    (as before).  Splitting the passes increases the inherent parallelism
921    in the dependency graph because the entire graph including all
922    libraries is available at the same time.  This, in turn, improves
923    server utilization.  The downside is that the master process will
924    have to do some extra work after compilation is done (because for
925    technical reasons it must re-read all the binfiles during stabilization).
926    
927    ----------------------------------------------------------------------
928    Name: Matthias Blume
929    Date: 2001/03/16 12:22:00 EST
930    Tag: blume-20010316-bootfiles
931    Description:
932    
933    Created a new set of bootfiles (for your automatic installation convenience).
934    
935    ----------------------------------------------------------------------
936    Name: Matthias Blume
937    Date: 2001/03/16 11:00:00 EST
938    Tag: blume-20010316-MLTREE-fixup
939    Description:
940    
941    This is a minor fixup for an (untagged) earlier commit by Allen.
942    (A file was missing).
943    
944    ----------------------------------------------------------------------
945    Name: Allen Leung
946    Date: Mon Mar  5 18:54:57 EST 2001
947    Tag: leunga-20010305-cut-support
948    
949    1. New support for alternative control-flow in MLTREE.
950       Currently we support
951    
952          FLOW_TO(CALL ...., [k1,...,kn])
953    
954       This is needed for 'cuts to' in C-- and try/handle-like constructs
955       in Moby
956    
957       New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
958    
959    2. Register Allocator
960    
961       Changes in interface [from Fermin, John]
962    
963    3. Alpha 8-bit SLL support [Fermin]
964    
965    4. All architectures
966    
967       A new module (ClusterExpandCopies) for expanding parallel copies.
968    
969    ----------------------------------------------------------------------
970    Name: Allen Leung
971    Date: 2001/02/27 23:07:00 EST
972    Tag: leunga-20010227-minor-stuff
973    
974    1. Alpha bug fix for CMOVNE
975    2. Handle mltree COND(..,FCMP ...,...)
976    3. Bug fix in simplifier
977    
978    ----------------------------------------------------------------------
979    Name: Matthias Blume
980    Date: 2001/01/30 17:50:00 EST
981    Tag: blume-20010130-sync
982    Description:
983    
984    This is just a minor update to sync my devel branch with the main brach.
985    The only visible change is the addition of some README files.
986    
987    ----------------------------------------------------------------------
988    Name: Matthias Blume
989    Date: 2001/01/12 23:30:00 JST
990    Tag: blume-20010112-bootfiles
991    Description:
992    
993    Made a new set of bootfiles that goes with the current state of the
994    repository.
995    
996    ----------------------------------------------------------------------
997    Name: Matthias Blume
998    Date: 2001/01/12 21:20:00 JST
999    Tag: blume-20010112-sync
1000    Description:
1001    
1002    I am just flushing out some minor changes that had accumulated in
1003    my private branch in order to sync with the main tree.  (This is
1004    mainly because I had CVS trouble when trying to merge _into_ my
1005    private branch.)
1006    
1007    Most people should be completely unaffected by this.
1008    
1009    ----------------------------------------------------------------------
1010    Name: Allen Leung
1011    Date: Thu Jan 11 21:03:00 EST 2001
1012    Tag: leunga-20010111-labexp=mltree
1013    Description:
1014    
1015    1.  Removed the type LabelExp and replace it by MLTree.
1016    2.  Rewritten mltree-simplify with the pattern matcher tool.
1017    3.  There were some bugs in alpha code generator which would break
1018        64-bit code generation.
1019    4.  Redo the tools to generate code with the
1020    5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
1021        are now generated by perl scripts.
1022    
1023    ----------------------------------------------------------------------
1024    Name: Matthias Blume
1025    Date: 2001/01/10 21:55:00 JST
1026    Tag: blume-20010110-rcc
1027    Description:
1028    
1029    The RCC stuff now seems to work (but only on the x86).
1030    This required hacking of the c-calls interface (and -implementation) in
1031    MLRISC.
1032    
1033    Normal compiler users should be unaffected.
1034    
1035    ----------------------------------------------------------------------
1036    Name: Matthias Blume
1037    Date: 2001/01/09 01:20:00 JST
1038    Tag: blume-20010109-rcc
1039    Description:
1040    
1041    This is a fairly big patch, flushing out a large number of pending
1042    changes that I made to my development copy over the last couple of days.
1043    
1044    Of practical relevance at this moment is a workaround for a pickling
1045    bug that Allen ran into the other day.  The cause of the bug itself is
1046    still unknown and it might be hard to fix it properly, but the
1047    workaround has some merits of its own (namely somewhat reducing pickling
1048    overhead for certain libraries).  Therefore, I think this solution should
1049    be satisfactory at this time.
1050    
1051    The rest of the changes (i.e., the vast majority) has to do with my
1052    ongoing efforts of providing direct support for C function calls from
1053    ML.  At the moment there is a new primop "RAW_CCALL", typing magic
1054    in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
1055    case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
1056    translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
1057    handle RCC.
1058    
1059    The last part (the changes to mlriscGen.sml) are still known to be
1060    wrong on the x86 and not implemented on all other architectures.  But
1061    the infrastructure is in place. I had to change a few functor
1062    signatures in the backend to be able to route the CCalls interface
1063    from MLRISC there, and I had to specialize the mltree type (on the
1064    x86) to include the necessary extensions. (The extensions themselves
1065    were already there and redy to go in MLRISC/x86).
1066    
1067    Everything should be very happy as soon as someone helps me with
1068    mlriscGen.sml...
1069    
1070    In any case, nothing of this should matter to anyone as long as the
1071    new primop is not being used (which is going to be the case unless you
1072    find it where I hid it :). The rest of the compiler is completely
1073    unaffected.
1074    
1075    ----------------------------------------------------------------------
1076    Name: Matthias Blume
1077    Date: 2001/01/05 00:30:00 JST
1078    Tag: blume-20010105-primops
1079    Description:
1080    
1081    Added some experimental support for work that I am doing right now.
1082    These changes mostly concern added primops, but there is also a new
1083    experimental C library in the runtime system (but currently not enabled
1084    anywhere except on Linux/X86).
1085    
1086    In the course of adding primops (and playing with them), I discovered that
1087    Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
1088    broken.  (Zhong was very right he labeled this stuff as "major gross hack".)
1089    To recover, I made type information in INL_PRIM mandatory and changed
1090    prim.sml as well as built-in.sml accordingly.  The InLine structure now
1091    has complete, correct type information (i.e., no bottom types).
1092    
1093    Since all these changes mean that we need new binfiles, I also bumped the
1094    version number to 110.32.1.
1095    
1096    ----------------------------------------------------------------------
1097    Name: Matthias Blume
1098    Date: 2000/12/30 22:10:00 JST
1099    Tag: blume-20001230-various
1100    Description:
1101    
1102    Added proxy libraries for MLRISC and let MLRISC libraries refer
1103    to each other using path anchors.  (See CM manual for explanation.)
1104    
1105    Updated CM documentation.
1106    
1107    Fixed some bugs in CM.
1108    
1109    Implemented "proxy" libraries (= syntactic sugar for CM).
1110    
1111    Added "-quiet" option to makeml and changed runtime system accordingly.
1112    
1113    Added cleanup handler for exportML to reset timers and compiler stats.
1114    
1115    ----------------------------------------------------------------------
1116    Name: Lal George
1117    Date: 2000/12/22 22:22:58 EST 2000
1118    Tag: Release_110_32
1119    Description:
1120    
1121            Infinite precision used throughout MLRISC.
1122            see MLRISC/mltree/machine-int.sig
1123    
1124    ----------------------------------------------------------------------
1125    Name: Matthias Blume
1126    Date: 2000/12/22 23:16:00 JST
1127    Tag: blume-20001222-warn
1128    Description:
1129    
1130    Corrected wording and formatting of some CM warning message which I
1131    broke in my previous patch.
1132    
1133    ----------------------------------------------------------------------
1134    Name: Matthias Blume
1135    Date: 2000/12/22 21:20:00 JST
1136    Tag: blume-20001222-anchorenv
1137    Description:
1138    
1139    Fixed CM's handling of anchor environments in connection with CMB.make.
1140    
1141    ----------------------------------------------------------------------
1142    Name: Matthias Blume
1143    Date: 2000/12/22 13:15:00 JST
1144    Tag: blume-20001222-cleanup
1145    Description:
1146    
1147    Removed src/cm/ffi which does not (and did not) belong here.
1148    
1149    ----------------------------------------------------------------------
1150    Name: Matthias Blume
1151    Date: 2000/12/21 23:55:00 JST
1152    Tag: blume-20001221-exn
1153    Description:
1154    
1155    Probably most important: CM no longer silently swallows all exceptions
1156    in the compiler.
1157    Plus: some other minor CM changes.  For example, CM now reports some
1158    sizes for generated binfiles (code, data, envpickle, lambdapickle).
1159    
1160    ----------------------------------------------------------------------
1161    Name: Matthias Blume
1162    Date: 2000/12/15 00:01:05 JST
1163    Tag: blume-20001215-dirtool
1164    Description:
1165    
1166    - "dir" tool added.
1167    - improvements and cleanup to Tools structure
1168    - documentation updates
1169    
1170    ----------------------------------------------------------------------
1171    Name: Allen Leung
1172    Date: Thu Dec 14 03:45:24 EST 2000
1173    Description:
1174    Tag:  leunga-20001214-int-inf
1175    Description:
1176    
1177       In IntInf, added these standard functions, which are missing from our
1178    implementation:
1179    
1180        andb : int * int -> int
1181        xorb : int * int -> int
1182        orb  : int * int -> int
1183        notb : int -> int
1184         <<   : int * word -> int
1185        ~>>  : int * word -> int
1186    
1187       Not tested, I hope they are correct.
1188    
1189    ----------------------------------------------------------------------
1190    Name: Allen Leung
1191    Date: Fri Dec  8 19:23:26 EST 2000
1192    Description:
1193    Tag:  leunga-20001208-nowhere
1194    Description:
1195    
1196      Slight improvements to the 'nowhere' tool to handle OR-patterns,
1197    to generate better error messages etc.  Plus a brief manual.
1198    
1199    ----------------------------------------------------------------------
1200    Name: Lal George
1201    Date: 2000/12/08 09:54:02 EST 2000
1202    Tag: Release_110_31
1203    Description:
1204    
1205    - Version 110.31
1206    ----------------------------------------------------------------------
1207    Name: Allen Leung
1208    Date: Thu Dec  7 22:01:04 EST 2000
1209    Tag:  leunga-20001207-cell-monster-hack
1210    Description:
1211    
1212    Major MLRISC internal changes.  Affect all clients.
1213    Summary:
1214    
1215    1.  Type CELLS.cell = int is now replaced by a datatype.
1216        As a result, the old regmap is now gone.  Almost all interfaces
1217        in MLRISC change as a consequence.
1218    
1219    2.  A new brand version of machine description tool (v3.0) that generates
1220        modules expecting the new interface.  The old version is removed.
1221    
1222    3.  The RA interface has been further abstracted into two new functors.
1223        RISC_RA and X86RA.  These functors have much simpler interfaces.
1224        [See also directory MLRISC/demo.]
1225    
1226    4.  Some other new source->source code generation tools are available:
1227    
1228        a. MLRISC/Tools/RewriteGen -- generate rewriters from rules.
1229        b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules.
1230           I use this tool to generate the peephole optimizers---with the new
1231           cell type changes, peephole rules are becoming difficult to write
1232           without conditional pattern matching.
1233    
1234    5.  More Intmap -> IntHashTable change.  Previous changes by Matthias didn't
1235        cover the entire MLRISC source tree so many things broke.
1236    
1237    6.  CM files have been moved to the subdirectory MLRISC/cm.
1238        They are moved because there are a lot of them and they clutter up the
1239        root dir.
1240    
1241    7.  More detailed documentation to come...
1242    
1243        NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do
1244        a makeml -rebuild first.  This is because of other other
1245        changes that Matthias has made (see below).
1246    
1247    
1248    ----------------------------------------------------------------------
1249    Name: Matthias Blume
1250    Date: 2000/11/30 23:12:00 JST
1251    Tag: blume-20001130-filereorg
1252    Description:
1253    
1254    Some manual updates and some file reorganizations in CM.
1255    
1256    ----------------------------------------------------------------------
1257    Name: Matthias Blume
1258    Date: 2000/11/24 17:45:00 JST
1259    Tag: blume-20001124-link
1260    Description:
1261    
1262    Drastically improved link traversal code for the case that the dynamic
1263    value was already loaded at bootstrap time.  As a result, CM and CMB
1264    now both load blazingly fast -- even on a very slow machine.  Also,
1265    memory consumption has been further reduced by this.
1266    
1267    Warning: The format of the PIDMAP file has changed.  THerefore, to
1268    bootstrap you have to do this:
1269    
1270    1. Run CMB.make
1271    2. Make a symbolic link for the boot directory:
1272         ln -s sml.boot.ARCH-OS xxx
1273    3. "Rebuild" the boot directory:
1274         ./makeml -boot xxx -rebuild sml ; rm xxx
1275    4. Boot normally:
1276          ./makeml
1277    
1278    ----------------------------------------------------------------------
1279    Name: Matthias Blume
1280    Date: 2000/11/21 21:20:00 JST
1281    Tag: blume-20001121-tools
1282    Description:
1283    
1284    Continued hacking on autoloading problem -- with success this time.
1285    Also changed tool-plugin mechanism.  See new CM manual.
1286    
1287    ----------------------------------------------------------------------
1288    Name: Matthias Blume
1289    Date: 2000/11/19 14:30:00 JST
1290    Tag:  blume-20001119-autoload
1291    Description:
1292    
1293    Some hacking to make autoloading faster.  Success for CMB, no success
1294    so far for CM.  There is a reduced structure CM' that autoloads faster.
1295    (This is a temporary, non-documented hack to be eliminated again when
1296    the general problem is solved.)
1297    
1298    ----------------------------------------------------------------------
1299    Name: Matthias Blume
1300    Date: 2000/11/17 14:10:00 JST
1301    Tag: blume-20001117-pickle-lib
1302    Description:
1303    
1304    1. Eliminated comp-lib.cm
1305    2. Made pickle-lib.cm
1306    3. Eliminated all uses of intset.sml (from comp-lib.cm)
1307    4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
1308       equivalent constructs from smlnj-lib.cm (INtHashTable).
1309    5. Point 4. also goes for those uses of intmap.* in MLRISC.
1310       Duplicated intmap modules thrown out.
1311    6. Hunted down all duplicated SCC code and replaced it with
1312       equivalent stuff (GraphSCCFn from smlnj-lib.cm).
1313    7. Rewrote Feedback module.
1314    8. Moved sortedlist.sml into viscomp-lib.cm.  Eventually it
1315       should be thrown out and equivalent modules from smlnj-lib.cm
1316       should be used (IntRedBlackSet, IntListSet, ...).
1317    
1318    Confirmed that compiler compiles to fixpoint.
1319    
1320    ----------------------------------------------------------------------
1321    Name: Allen Leung
1322    Date: 2000/11/10 18:00:00
1323    Tag: leunga-20001110-new-x86-fp
1324    
1325    A new x86 floating point code generator has been added.
1326    By default this is turned off.  To turn this on, do:
1327    
1328        CM.autoload "$smlnj/compiler.cm";
1329        Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
1330    
1331    Changes:
1332    
1333    1.  Changed FTAN to FPTAN so that the assembly output is correct.
1334    2.  Changed the extension callback for FTANGENT to generate:
1335    
1336              fptan
1337              fstp  %st(0)
1338        instead of
1339              fptan
1340              fstpl ftempmem
1341    
1342    3.  Numerous assembly fixes for x86.
1343    
1344    5.  Cleaned up the machine code output module x86/x86MC.sml and added
1345        support for a whole bunch of instructions and addressing modes:
1346    
1347          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st, %st(n)
1348          faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp  %st, %st(n)
1349          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st(n), %st
1350          fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
1351          fxch %st(n)
1352          fld %st(n)
1353          fst %st(n)
1354          fst mem
1355          fstp %st(n)
1356          fucom %st(n)
1357          fucomp %st(n)
1358    
1359        All these are now generated when the fast fp mode is turned on.
1360    
1361    6.  Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
1362    
1363    ----------------------------------------------------------------------
1364    Name: Matthias Blume
1365    Date: 2000/11/09 11:20:00 JST
1366    Tag: blume-20001109-scc
1367    Description:
1368    
1369    Eliminated some code duplication:
1370    
1371    1. Added "where" clause to GraphSCCFn in SML/NJ Library.
1372       (Otherwise the functor is useless.)
1373    2. Used GraphSCCFn where SCCUtilFun was used previously.
1374    3. Got rid of SCCUtilFun (in comp-lib.cm).
1375    
1376    ----------------------------------------------------------------------
1377    Name: Lal George
1378    Date: 2000/11/06 09:02:21 EST 2000
1379    Tag: Release_110_30
1380    Description:
1381    
1382    - Version 110.30
1383    ----------------------------------------------------------------------
1384    Name: Matthias Blume
1385    Date: 2000/11/04 14:45:00
1386    Tag: blume-20001104-mlbuild
1387    Description:
1388    
1389    - Made ml-build faster on startup.
1390    - Documentation fixes.
1391    
1392    ----------------------------------------------------------------------
1393    Name: Matthias Blume
1394    Date: 2000/11/02 17:00:00 JST
1395    Tag: blume-20001102-condcomp
1396    Description:
1397    
1398    - Small tweaks to pickler -- new BOOTFILES!
1399    - Version bumped to 110.29.2.
1400    - Added conditional compilation facility to init.cmi (see comment there).
1401    ----------------------------------------------------------------------
1402    Name: Allen Leung
1403    Date: 2000/10/23 19:31:00
1404    Tag: leunga-20001023-demo-ra
1405    
1406    1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
1407    2. Test programs for the graph library updated
1408    3. Some new MLRISC demo programs added
1409    
1410    ----------------------------------------------------------------------
1411    Name: Matthias Blume
1412    Date: 2000/08/31 22:15:00 JST
1413    Tag: blume-20001017-errmsg
1414    Description:
1415    
1416    More error message grief: Where there used to be no messages, there
1417    now were some that had bogus error regions.  Fixed.
1418    
1419    ----------------------------------------------------------------------
1420    Name: Matthias Blume
1421    Date: 2000/08/31 17:30:00 JST
1422    Tag: blume-20001017-v110p29p1
1423    Description:
1424    
1425    I made a version 110.29.1 with new bootfiles.
1426    
1427    Changes:  Modified pickler/unpickler for faster and leaner unpickling.
1428              CM documentation changes and a small bugfix in CM's error reporting.
1429    
1430    ----------------------------------------------------------------------
1431    Name: Lal George
1432    Date: 2000/09/27 14:42:35 EDT
1433    Tag: george-20000927-nodestatus
1434    Description:
1435    
1436    Changed the type of the nodestatus, so that:
1437    
1438            SPILLED(~1)             is now SPILLED
1439            SPILLED(m) where m>=0   is now MEMREG(m)
1440            SPILLED(s) where s<~1   is now SPILL_LOC(~s)
1441    
1442    ----------------------------------------------------------------------
1443    Name: Matthias Blume
1444    Date: 2000/09/07 14:45:00 JST
1445    Tag: blume-20000907-cmerrmsg
1446    Description:
1447    
1448    Small tweak to CM to avoid getting ML syntax error messages twice.
1449    
1450    ----------------------------------------------------------------------
1451    Name: Matthias Blume
1452    Date: 2000/08/31 18:00:00 JST
1453    Tag: blume-20000831-cvsbootfiles
1454    Description:
1455    
1456    New URL for boot files (because the 110.29 files on the BL server do
1457    now work correctly with my updated install scripts for yacc and lex).
1458    
1459    ----------------------------------------------------------------------
1460    Name: Matthias Blume
1461    Date: 2000/08/08 12:33:00 JST
1462    Tag: blume-20000808-manual
1463    Description:
1464    
1465    Tiny update to CM manual.
1466    
1467    ----------------------------------------------------------------------
1468    Name: Allen Leung
1469    Date: 2000/08/7 19:31:00
1470    Tag: leunga-20000807-a-whole-bunch-of-stuff
1471    
1472      Moby, C--, SSA, x86, machine descriptions etc.  Should only affect C--
1473    and Mobdy.
1474    
1475    1.  x86
1476    
1477       a.  Fixes to peephole module by John and Dan.
1478       b.  Assembly fix to SETcc by Allen.
1479       c.  Fix to c-call by John.
1480       d.  Fix to spilling by John.  (This one deals with the missing FSTPT case)
1481       e.  Instruction selection optimization to SETcc as suggested by John.
1482    
1483           For example,
1484    
1485            MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
1486    
1487           should generate:
1488    
1489            MOVL a, x
1490            SUBL b, x
1491            SHRL 31, x
1492    
1493    2.  IR stuff
1494    
1495         A bunch of new DJ-graph related algorithms added.  These
1496         speed up SSA construction.
1497    
1498    3.  SSA + Scheduling
1499    
1500         Added code for SSA and scheduling to the repository
1501    
1502    ----------------------------------------------------------------------
1503    Name: Lal George
1504    Date: 2000/07/27 11:53:14 EDT
1505    
1506    Tag: lal-20000727-linux-ppc
1507    Description:
1508    
1509     Made changes to support Linux PPC.
1510     p.s. I have confirmation that the 110.29 boot files work fine.
1511    
1512    ----------------------------------------------------------------------
1513    Name: Matthias Blume
1514    Date: 2000/07/27 17:40:00 JST
1515    Tag: blume-20000727-scripts
1516    Description:
1517    
1518    !!!! WARNING !!!!
1519    You must recompile the runtime system!
1520    !!!! WARNING !!!!
1521    
1522    This is basically another round of script-enhancements:
1523    
1524    1. sml, ml-build, and ml-makedepend accept options -D and -U to define
1525       and undefine CM preprocessor symbols.
1526    
1527    2. ml-build avoids generating a new heap image if it finds that the
1528       existing one is still ok.  (The condition is that no ML file had to
1529       be recompiled and all ML files are found to be older that the heap
1530       file.)
1531    
1532       To make this work smoothly, I also hacked the runtime system as
1533       well as SMLofNJ.SysInfo to get access to the heap image suffix
1534       (.sparc-solaris, ...) that is currently being used.
1535    
1536       Moreover, the signature of CM.mk_standalone has changed.  See the
1537       CM manual.
1538    
1539    3. ml-makedepend accepts additional options -n, -a, and -o.  (See the
1540       CM manual for details.)
1541    
1542    4. More CM manual updates:
1543        - all of the above has been documented.
1544        - there is now a section describing the (CM-related) command line
1545          arguments that are accepted by the "sml" command
1546    
1547    ----------------------------------------------------------------------
1548    Name: Matthias Blume
1549    Date: 2000/07/25 16:20:00 JST
1550    Tag: blume-20000725-makedepend
1551    Description:
1552    
1553    Added a script called ml-makedepend.  This can be used in makefiles
1554    for Unix' make in a way very similar to the "makedepend" command for
1555    C.
1556    
1557    The script internally uses function CM.sources.
1558    
1559    Synopsis:
1560    
1561        ml-makedepend [-f makefile] cmfile targetname
1562    
1563    The default for the makefile is "makefile" (or "Makefile" should
1564    "makefile" not exist).
1565    
1566    ml-makedepend adds a cmfile/targetname-specific section to this
1567    makefile (after removing the previous version of this section).  The
1568    section contains a single dependency specification with targetname on
1569    the LHS (targetname is an arbitrary name), and a list of files derived
1570    from the cmfile on the RHS.  Some of the files on the RHS are
1571    ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
1572    "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
1573    path names.  The makefile writer is responsible for making sure that
1574    these variables have correct at the time "make" is invoked.
1575    
1576    ----------------------------------------------------------------------
1577    Name: Matthias Blume
1578    Date: 2000/07/22 23:30:00 JST
1579    Tag: blume-20000722-urlupdate
1580    Description:
1581    
1582    Changed BOOT and config/srcarchiveurl to point to BL server:
1583    
1584        ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
1585    
1586    ----------------------------------------------------------------------
1587    Name: Matthias Blume
1588    Date: 2000/07/18 18:00:00 JST
1589    Tag: blume-20000718-Version_110_29
1590    Description:
1591    
1592    1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
1593    
1594    2. Updated config/version to 110.29
1595    
1596    3. Updated config/srcarchiveurl
1597    
1598    3. New boot files!
1599       ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
1600    
1601    ----------------------------------------------------------------------
1602    Name: Matthias Blume
1603    Date: 2000/07/11 13:58:00 JST
1604    Tag: blume-20000711-doctypo
1605    Description:
1606    
1607    Fixed a few typos in CM manual.
1608    
1609    ----------------------------------------------------------------------
1610    Name: Allen Leung
1611    Date: 2000/06/15 00:38:00
1612    Tag: leunga-20000704-sparc-x86
1613    
1614    1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
1615    2. fix to x86 RET bug [found by Dan Grossman]
1616    3. sparc assembly bug fix for ticc instructions [found by Fermin]
1617    
1618       Affects c-- and moby only
1619    
1620    ----------------------------------------------------------------------
1621    Name: Matthias Blume
1622    Date: 2000/07/04 15:26:00
1623    Tag: blume-20000704-trigger
1624    Description:
1625    
1626    1. Improvements to CM manual.
1627    2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
1628       of getting a back-trace.  The function, when called, raises an
1629       internal exception which explicitly carries the full back-trace history,
1630       so it is unaffected by any intervening handle-raise pairs ("trivial"
1631       or not).  The interactive loop will print that history once it arrives
1632       at top level.
1633       Short of having all exceptions implicitly carry the full history, the
1634       recommended way of using this facility is:
1635         - compile your program with instrumentation "on"
1636         - run it, when it raises an exception, look at the history
1637         - if the history is "cut off" because of some handler, go and modify
1638           your program so that it explicitly calls BTrace.trigger
1639         - recompile (still instrumented), and rerun; look at the full history
1640    
1641    ----------------------------------------------------------------------
1642    Name: Matthias Blume
1643    Date: 2000/07/03 15:36:00 JST
1644    Tag: blume-20000702-manual
1645    Description:
1646    
1647    Small corrections and updates to CM manual.
1648    
1649    ----------------------------------------------------------------------
1650    Name: Matthias Blume
1651    Date: 2000/06/29 16:04:00 JST
1652    Tag: blume-20000629-yacctool
1653    Description:
1654    
1655    Changes:
1656    
1657    1. Class "mlyacc" now takes separate arguments to pass options to
1658       generated .sml- and .sig-files independently.
1659    2. Corresponding CM manual updates.
1660    3. BTrace module now also reports call sites.  (However, for loop clusters
1661       it only shows from where the cluster was entered.)  There are associated
1662       modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
1663    
1664    ----------------------------------------------------------------------
1665    Name: Matthias Blume
1666    Date: 2000/06/27 16:51:00 JST
1667    Tag: blume-20000627-noweb
1668    Description:
1669    
1670    Changes:
1671    
1672     1. Implemented "subdir" and "witness" options for noweb tool.
1673        This caused some slight internal changes in CM's tool implementation.
1674     2. Fixed bug in "tool plugin" mechanism.  This is essentially cleaning
1675        some remaining issues from earlier path anchor changes.
1676     3. Updated CM manual accordingly.
1677    
1678     4. Changed implementation of back-tracing so that I now consider it
1679        ready for prime-time.
1680    
1681        In particular, you don't have to explicitly trigger the back-trace
1682        anymore.  Instead, if you are running BTrace-instrumented code and
1683        there is an uncaught exception (regardless of whether or not it was
1684        raised in instrumented code), the top-level evalloop will print
1685        the back-trace.
1686    
1687        Features:
1688    
1689          - Instrumented and uninstrumented code work together seemlessly.
1690            (Of course, uninstrumented code is never mentioned in actual
1691             back-traces.)
1692    
1693          - Asymptotic time- and space-complexity of instrumented code is
1694            equal to that of uninstrumented code.  (This means that
1695            tail-recursion is preserved by the instrumentation phase.)
1696    
1697          - Modules whose code has been instrumented in different sessions
1698            work together without problem.
1699    
1700          - There is no penalty whatsoever on uninstrumented code.
1701    
1702          - There is no penalty on "raise" expressions, even in
1703            instrumented code.
1704    
1705        A potential bug (or perhaps it is a feature, too):
1706    
1707          A back-trace reaches no further than the outermost instrumented
1708          non-trivial "raise".  Here, a "trivial" raise is one that is the
1709          sole RHS of a "handle" rule.  Thus, back-traces reach trough
1710    
1711               <exp> handle e => raise e
1712    
1713          and even
1714    
1715               <exp> handle Foo => raise Bar
1716    
1717          and, of course, through
1718    
1719               <exp> handle Foo => ...
1720    
1721         if the exception was not Foo.
1722    
1723         Back-traces always reach right through any un-instrumented code
1724         including any of its "handle" expressions, trivial or not.
1725    
1726       To try this out, do the following:
1727    
1728         - Erase all existing binfiles for your program.
1729           (You may keep binfiles for those modules where you think you
1730            definitely don't need back-tracing.)
1731         - Turn on back-trace instrumentation:
1732              SMLofNJ.Internals.BTrace.mode (SOME true);
1733         - Recompile your program.  (I.e., run "CM.make" or "use".)
1734         - You may now turn instrumentation off again (if you want):
1735              SMLofNJ.Internals.BTrace.mode (SOME false);
1736         - Run your program as usual.  If it raises an exception that
1737           reaches the interactive toplevel, then a back-trace will
1738           automatically be printed.  After that, the toplevel loop
1739           will print the exception history as usual.
1740    
1741    ----------------------------------------------------------------------
1742    Name: Matthias Blume
1743    Date: 2000/06/26 09:56:46 JST
1744    Tag: blume-20000626-setup
1745    Description:
1746    
1747    CM: - setup-parameter to "sml" added; this can be used to run arbitrary
1748          ML code before and after compiling a file (e.g., to set compiler
1749          flags)
1750    
1751    Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
1752              - associated changes to btrace.sml (BTrace instrumentation pass)
1753              - cleaner implementation of btimp.sml (BTrace tracing and report
1754                module)
1755    
1756    CM manual: * new path encoding documented
1757               * description of setup-parameter to "sml" added
1758    
1759    The biggest user-visible change to back-tracing is that it is no
1760    longer necessary to compile all traced modules within the same
1761    session.  (This was a real limitation.)
1762    
1763    ----------------------------------------------------------------------
1764    Name: Matthias Blume
1765    Date: 2000/06/24 12:40:00 JST
1766    Tag: blume-20000624-startup
1767    Description:
1768    
1769    Fixes startup slowdown problem.  (I was calling SrcPath.sync a _tad_
1770    bit too often -- to put it mildly. :)
1771    
1772    ----------------------------------------------------------------------
1773    Name: Matthias Blume
1774    Date: 2000/06/23 18:20:00 JST
1775    Tag: blume-20000623-btrace
1776    Description:
1777    
1778    This updates adds a backtrace facility to aid programmers in debugging
1779    their programs.  This involves the following changes:
1780    
1781    1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
1782       keeping track of the current call stack.  When programs are compiled
1783       in a special mode, the compiler will insert calls to these hooks
1784       into the user program.
1785       "Hook" means that it is possible for different implementations of
1786       back-tracing to register themselves (at different times).
1787    
1788    2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
1789       as an Absyn.dec->Absyn.dec rewrite.  Normally this phase is turned off.
1790       It can be turned on using this call:
1791         SMLofNJ.Internals.BTrace.mode (SOME true);
1792       Turning it off again:
1793         SMLofNJ.Internals.BTrace.mode (SOME false);
1794       Querying the current status:
1795         SMLofNJ.Internals.BTrace.mode NONE;
1796       Annotated programs are about twice as big as normal ones, and they
1797       run a factor of 2 to 4 slower with a dummy back-trace plugin (one
1798       where all hooks do nothing).  The slowdown with a plugin that is
1799       actually useful (such as the one supplied by default) is even greater,
1800       but in the case of the default plugin it is still only an constant
1801       factor (amortized).
1802    
1803    3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
1804       with a sub-structure BTrace for controlling back-tracing.  In particular,
1805       the above-mentioned function "mode" controls whether the annotation
1806       phase is invoked by the compiler.  Another important function is
1807       "trigger": when called it aborts the current execution and causes
1808       the top-level loop to print a full back-trace.
1809    
1810    4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
1811       for back-tracing.  It keeps track of the dynamic call stack and in
1812       addition to that it keeps a partial history at each "level" of that
1813       stack.  For example, if a tail-calls b, b tail-calls c, and c tail-calls
1814       d and b (at separate times, dynamically), then the report will show:
1815    
1816       GOTO   d
1817             /c
1818       GOTO  \b
1819       CALL   a
1820    
1821       This shows that there was an initial non-tail call of a, then a
1822       tail-call to b or c, looping behavior in a cluster of functions that
1823       consist of b and c, and then a goto from that cluster (i.e., either from
1824       b or from c) to d.
1825    
1826       Note that (depending on the user program) the amount of information
1827       that the back-trace module has to keep track of at each level is bounded
1828       by a constant.  Thus, the whole implementation has the same asymptotical
1829       complexity as the original program (both in space and in time).
1830    
1831    5. compiler/TopLevel/interact/evalloop.sml has been modified to
1832       handle the special exception SMLofNJ.Internals.BTrace.BTrace
1833       which is raised by the "trigger" function mentioned above.
1834    
1835    Notes on usage:
1836    
1837    - Annotated code works well together with unannotated code:
1838    Unannotated calls simply do not show up at all in the backtrace.
1839    
1840    - It is not a good idea to let modules that were annotated during
1841    different sessions run at the same time.  This is because the compiler
1842    chooses small integers to identify individual functions, and there
1843    will be clashes if different modules were compiled in separate sessions.
1844    (Nothing will crash, and you will even be told about the clashes, but
1845    back-trace information will in general not be useful.)
1846    
1847    - Back-tracing can be confused by callcc and capture.
1848    
1849    - The only way of getting a back-trace right now is to explicitly
1850    invoke the "trigger" function from your user program.  Eventually, we
1851    should make every exception carry back-trace information (if
1852    available).  But since this creates more overhead at "raise"-time
1853    (similar to the current exnHistory overhead), I have not yet
1854    implemented this.  (The implementation will be rather easy.)  With
1855    exceptions carrying back-trace information, this facility will be even
1856    more useful because users don't need to modify their programs...
1857    
1858    - While it is possible to compile the compiler with back-trace
1859    annotations turned on (I did it to get some confidence in
1860    correctness), you must make absolutely sure that core.sml and
1861    btimp.sml are compiled WITHOUT annotation!  (core.sml cannot actually
1862    be compiled with annotation because there is no core access yet, but
1863    if you compile btimp.sml with annotation, then the system will go into
1864    an infinite recursion and crash.)
1865    Since CM currently does not know about BTrace, the only way to turn
1866    annotations on and off for different modules of the compiler is to
1867    interrupt CMB.make, change the settings, and re-invoke it.  Of course,
1868    this is awkward and clumsy.
1869    
1870    Sample sessions:
1871    
1872    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
1873    - SMLofNJ.Internals.BTrace.mode (SOME true);
1874    [autoloading]
1875    [autoloading done]
1876    val it = false : bool
1877    - structure X = struct
1878    -     fun main n = let
1879    -         fun a (x, 0) = d x
1880    -           | a (x, n) = b (x, n - 1)
1881    -         and b (x, n) = c (x, n)
1882    -         and c (x, n) = a (x, n)
1883    -         and d x = e (x, 3)
1884    -         and e (x, 0) = f x
1885    -           | e (x, n) = e (x, n - 1)
1886    -         and f 0 = SMLofNJ.Internals.BTrace.trigger ()
1887    -           | f n = n * g (n - 1)
1888    -         and g n = a (n, 3)
1889    -     in
1890    -         f n
1891    -     end
1892    - end;
1893    structure X : sig val main : int -> int end
1894    - X.main 3;
1895    *** BACK-TRACE ***
1896    GOTO   stdIn:4.2-13.20: X.main[2].f
1897    GOTO-( stdIn:4.2-13.20: X.main[2].e
1898    GOTO   stdIn:4.2-13.20: X.main[2].d
1899         / stdIn:4.2-13.20: X.main[2].a
1900         | stdIn:4.2-13.20: X.main[2].b
1901    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1902    CALL   stdIn:4.2-13.20: X.main[2].g
1903    GOTO   stdIn:4.2-13.20: X.main[2].f
1904    GOTO-( stdIn:4.2-13.20: X.main[2].e
1905    GOTO   stdIn:4.2-13.20: X.main[2].d
1906         / stdIn:4.2-13.20: X.main[2].a
1907         | stdIn:4.2-13.20: X.main[2].b
1908    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1909    CALL   stdIn:4.2-13.20: X.main[2].g
1910    GOTO   stdIn:4.2-13.20: X.main[2].f
1911    GOTO-( stdIn:4.2-13.20: X.main[2].e
1912    GOTO   stdIn:4.2-13.20: X.main[2].d
1913         / stdIn:4.2-13.20: X.main[2].a
1914         | stdIn:4.2-13.20: X.main[2].b
1915    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1916    CALL   stdIn:4.2-13.20: X.main[2].g
1917    GOTO   stdIn:4.2-13.20: X.main[2].f
1918    CALL   stdIn:2.15-17.4: X.main[2]
1919    -
1920    
1921    (Note that because of a FLINt bug the above code currently does not
1922    compile without BTrace turned on.)
1923    
1924    Here is another example, using my modified Tiger compiler:
1925    
1926    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
1927    - SMLofNJ.Internals.BTrace.mode (SOME true);
1928    [autoloading]
1929    [autoloading done]
1930    val it = false : bool
1931    - CM.make "sources.cm";
1932    [autoloading]
1933    ...
1934    [autoloading done]
1935    [scanning sources.cm]
1936    [parsing (sources.cm):parse.sml]
1937    [creating directory CM/SKEL ...]
1938    [parsing (sources.cm):tiger.lex.sml]
1939    ...
1940    [wrote CM/sparc-unix/semant.sml]
1941    [compiling (sources.cm):main.sml]
1942    [wrote CM/sparc-unix/main.sml]
1943    [New bindings added.]
1944    val it = true : bool
1945    - Main.compile ("../testcases/merge.tig", "foo.out");
1946    *** BACK-TRACE ***
1947    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
1948    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1949    CALL   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1950    GOTO   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1951    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1952    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1953    CALL   lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
1954         / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
1955    CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
1956    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1957    CALL   lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
1958    CALL   main.sml:1.18-118.4: Main.compile[2]
1959    -
1960    
1961    ----------------------------------------------------------------------
1962    Name: Matthias Blumen
1963    Date: 2000/06/21 18:00:00 JST
1964    Tag: blume-20000621-manual
1965    Description:
1966    
1967    CM manual update: Path environments documented.
1968    
1969    ----------------------------------------------------------------------
1970    Name: Matthias Blume
1971    Date: 2000/06/19 13:40:00
1972    Tag: blume-20000619-manual
1973    Description:
1974    
1975    CM manual and system/README update.  This only covers the fact that
1976    there are no more implicit anchors.  (Path environments and the "bind"
1977    option to "cm" have yet to be documented.)
1978    
1979    ----------------------------------------------------------------------
1980    Name: Matthias Blume
1981    Date: 2000/06/19 11:05:00 JST
1982    Tag: blume-20000619-chdir-bugfix
1983    Description:
1984    
1985    Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
1986    
1987    ----------------------------------------------------------------------
1988    Name: Matthias Blume
1989    Date: 2000/06/18 22:00:10 JST
1990    Tag: blume-20000618-implicit-anchors-really-gone
1991    Description:
1992    
1993    I updates the previous HISTORY entry where I forgot to mention that
1994    implicit anchors are no longer with us.
1995    
1996    The current update also gets rid of the (now useless) controller
1997    CM.Control.implicit_anchors.
1998    
1999    ----------------------------------------------------------------------
2000    Name: Matthias Blume
2001    Date: 2000/06/16 17:30:00 JST
2002    Tag: blume-20000616-anchorenv
2003    Description:
2004    
2005    This patch implements the long anticipated (just kidding :) "anchor
2006    environment" mechanism.  In the course of doing this, I also
2007    re-implemented CM's internal "SrcPath" module from scratch.  The new
2008    one should be more robust in certain boundary cases.  In any case, it
2009    is a lot cleaner than its predecessor (IMHO).
2010    
2011    This time, although there is yet another boot file format change, I
2012    kept the unpickler backward-compatible.  As a result, no new bootfiles
2013    are necessary and bootstrapping is straightforward.  (You cannot read
2014    new bootfiles into an old system, but the other way around is no
2015    problem.)
2016    
2017    Visible changes:
2018    
2019    ** 0. Implicit path anchors (without the leading $-symbol) are no
2020    longer recognized at all. This means that such path names are not
2021    illegal either.  For example, the name basis.cm simply refers to a
2022    local file called "basis.cm" (i.e, the name is an ordinary path
2023    relative to .cm-files directory).  Or, to put it differently, only
2024    names that start with $ are anchored paths.
2025    
2026    ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
2027    vanished.
2028    
2029    John (Reppy) had critizised this as soon as I originally proposed and
2030    implemented it, but at that time I did not really deeply believe
2031    him. :) Now I came full-circle because I need the $<singlearc> syntax
2032    in another place where it cannot be seen as an abbreviation for
2033    $/<singlearc>.  To avoid the confusion, $<singlearc> now means what it
2034    seems to mean (i.e., it "expands" into the corresponding anchor
2035    value).
2036    
2037    However, when paths are used as members in CM description files, it
2038    continues to be true that there must be at least another arc after the
2039    anchor.  This is now enforced separately during semantic analysis
2040    (i.e., from a lexical/syntactical point of view, the notation is ok.)
2041    
2042    ** 2. The "cm" class now accepts an option "bind".  The option's value
2043    is a sub-option list of precisely two items -- one labeled "anchor"
2044    and the other one labeled "value".  As you might expect, "anchor" is
2045    used to specify an anchor name to be bound, and "value" specifies what
2046    the anchor is being bound to.
2047    
2048    The value must be a directory name and can be given in either standard
2049    syntax (including the possibility that it is itself an anchored path)
2050    or native syntax.
2051    
2052    Examples:
2053    
2054       foo.cm (bind:(anchor:bar value:$mystuff/bar))
2055       lib.cm (bind:(anchor:a value:"H:\\x\\y\\z"))  (* only works under windows *)
2056    
2057    and so on.
2058    
2059    The meaning of this is that the .cm-file will be processed with an
2060    augmented anchor environment where the given anchor(s) is/are bound to
2061    the given values(s).
2062    
2063    The rationale for having this feature is this: Suppose you are trying
2064    to use two different (already stable) libraries a.cm and b.cm (that
2065    you perhaps didn't write yourself).  Further, suppose each of these
2066    two libraries internally uses its own auxiliary library $aux/lib.cm.
2067    Normally you would now have a problem because the anchor "lib" can not
2068    be bound to more than one value globally.  Therefore, the project that
2069    uses both a.cm and b.cm must locally redirect the anchor to some other
2070    place:
2071    
2072       a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
2073       b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
2074    
2075    This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
2076    /usr/lib/smlnj/b-stuff/aux.cm, respectively.
2077    
2078    Hard-wiring path names is a bit inflexible (and CM will verbosely warn
2079    you when you do so at the time of CM.stabilize).  Therefore, you can
2080    also use an anchored path as the value:
2081    
2082      a.cm (bind:(anchor:lib value:$a-lib))
2083      b.cm (bind:(anchor:lib value:$b-lib))
2084    
2085    Now you can globally configure (using the usual CM.Anchor.anchor or
2086    pathconfig machinery) bindings for "a-lib" and "b-lib".  Since "lib"
2087    itself is always locally bound, setting it globally is no longer
2088    meaningful or necessary (but it does not hurt either).  In fact, "lib"
2089    can still be used as a global anchor for separate purposes.  As a
2090    matter of fact, one can locally define "lib" in terms of a global
2091    "lib":
2092    
2093      a.cm (bind:(anchor:lib value:$lib/a))
2094      b.cm (bind:(anchor:lib value:$lib/b))
2095    
2096    ** 3: The encoding of path names has changed.  This affects the way
2097    path names are shown in CM's progress report and also the internal
2098    protocol encoding used for parallel make.
2099    
2100    The encoding now uses one or more ':'-separated segments.  Each
2101    segments corresponds to a file that has been specified relative to the
2102    file given by its preceding segment.  The first segment is either
2103    relative to the CWD, absolute, or anchored.  Each segment itself is
2104    basically a Unix pathname; all segments but the first are relative.
2105    
2106    Example:
2107    
2108       $foo/bar/baz.cm:a/b/c.sml
2109    
2110    This path denotes the file bar/a/b/c.sml relative to the directory
2111    denoted by anchor "foo".  Notice that the encoding also includes
2112    baz.cm which is the .cm-file that listed a/b/c.sml.  As usual, such
2113    paths are resolved relative to the .cm-files directory, so baz.cm must
2114    be ignored to get the "real" pathname.
2115    
2116    To make this fact more obvious, CM puts the names of such "virtual
2117    arcs" into parentheses when they appear in progress reports. (No
2118    parentheses will appear in the internal protocol encoding.)  Thus,
2119    what you really see is:
2120    
2121      $foo/bar/(baz.cm):a/b/c.sml
2122    
2123    I find this notation to be much more informative than before.
2124    
2125    Another new feature of the encoding is that special characters
2126    including parentheses, colons, (back)slashes, and white space are
2127    written as \ddd (where ddd is the decimal encoding of the character).
2128    
2129    *** The CM manual still needs to be updated.
2130    
2131    ----------------------------------------------------------------------
2132    Name: Allen Leung
2133    Date: 2000/06/15 00:38:00
2134    Tag: leunga-20000615-x86-peephole
2135    
2136    x86 Peephole fix by Fermin.  Affects c-- and moby only.
2137    
2138    ----------------------------------------------------------------------
2139    Name: Matthias Blume
2140    Date: 2000/06/12 11:40:00
2141    Tag: blume-20000612-parmakefix
2142    Description:
2143    
2144    More cleanup after changing the file naming scheme: This time I
2145    repaired the parallel make mechanism for CMB.make which I broke earlier.
2146    
2147    ----------------------------------------------------------------------
2148    Name: Allen Leung
2149    Date: 2000/06/09 01:25:00
2150    Tag: leunga-20000609-various
2151    
2152    None of these things should affect normal SML/NJ operations
2153    
2154    1. Peephole improvements provided by Fermin (c--)
2155    2. New annotation DEFUSE for adding extra dependence (moby)
2156    3. New X86 LOCK instructions (moby)
2157    4. New machine description language for reservation tables (scheduling)
2158    5. Fixes to various optimization/analysis modules (branch chaining, dominator
2159       trees etc.)
2160    6. I've changed the CM files so that they can work with versions
2161       110.0.6, 110.25 and 110.28
2162    
2163    ----------------------------------------------------------------------
2164    Name: Matthias Blume
2165    Date: 2000/06/09 12:40:00
2166    Tag: blume-20000609-log
2167    Description:
2168    
2169    - Removed all(?) remaining RCS Log entries from sources.
2170    
2171    - Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
2172      anchored paths).
2173    
2174    ----------------------------------------------------------------------
2175    Name: Matthias Blume
2176    Date: 2000/06/07 17:00:00 JST
2177    Tag: blume-20000607-no-implicit-anchors
2178    Description:
2179    
2180    1. This update changes the default setting for
2181    CM.Control.implicit_anchors from true to false.  This means that
2182    implicit anchors are no longer permitted by default.  I also tried to
2183    make sure that nothing else still relies on implicit anchors.
2184    (This is the next step on the schedule towards a CM that does not even
2185    have the notion of implicit anchors anymore.)
2186    
2187    2. More CM manual updates.
2188    
2189    3. I managed to track down and fix the pickling bug I mentioned last
2190    time.  Because of the previously existing workaround, this entails no
2191    immediate practical changes.
2192    
2193    ----------------------------------------------------------------------
2194    Name: Matthias Blume
2195    Date: 2000/06/06 11:15:00 JST
2196    Tag: blume-20000606-lazierpickle
2197    Description:
2198    
2199    !!!! NEW BOOT FILES !!!!
2200    
2201    * The main purpose of this update is to make library pickles lazier in
2202    order to reduce the initial space penalty for autoloading a library.
2203    As a result, it is now possible to have $smlnj/compiler.cm
2204    pre-registered.  This should take care of the many complaints or
2205    inquiries about missing structure Compiler.  This required changes to
2206    CM's internal data structures and small tweaks to some algorithms.
2207    
2208    As a neat additional effect, it is no longer necessary (for the sake
2209    of lean heap image files) to distinguish between a "minimal" CM and a
2210    "full" CM.  Now, there is only one CM (i.e., the "full" version:
2211    $smlnj/cm.cm aka $smlnj/cm/full.cm), and it is always available at the
2212    interactive top level. ($smlnj/cm/minimal.cm is gone.)
2213    
2214    To make the life of compiler-hackers easier, "makeml" now also
2215    pre-registers $smlnj/cmb.cm (aka $smlnj/cmb/current.cm).  In other
2216    words, after you bootstrap a new sml for the first time, you will not
2217    have to autoload $smlnj/cmb.cm again afterwards.  (The first time
2218    around you will still have to do it, though.)
2219    
2220    * A second change consists of major updates to the CM manual.  There
2221    are now several appendices with summary information and also a full
2222    specification of the CM description file syntax.
2223    
2224    * In directory src/system I added the script "allcross".  This script
2225    invokes sml and cross-compiles the compiler for all supported
2226    architectures.  (Useful when providing a new set of boot files.)
2227    
2228    * There seems to be a latent bug in my "lazy pickles" mechanism.  I
2229    added a small tweak to pickle-util.sml to work around this problem,
2230    but it is not a proper fix yet.  I will investigate further.  (The
2231    effect of the bug was an inflation of library pickle size.)
2232    
2233    * Version number increased to 110.28.1 (to avoid compatibility problems).
2234    
2235    ----------------------------------------------------------------------
2236    Name: Allen Leung
2237    Date: 2000/05/25 17:28 EDT
2238    Tag: leunga-20000525-ra
2239    Description:
2240    
2241      Fixed a bug in freezing phase of the register allocator.
2242    
2243    ----------------------------------------------------------------------
2244    Name: Allen Leung
2245    Date: 2000/05/15 22:53 EDT
2246    Tag: leunga-20000515-alpha-x86-ra
2247    Description:
2248    
2249      1. Alpha
2250    
2251          Slight cleanup.  Removed the instruction SGNXL
2252    
2253      2. X86
2254    
2255          Added the following instructions to the instruction set:
2256    
2257            ROLx, RORx,
2258            BTx, BTSx, BTLx, BTRx,
2259            XCHGx, and variants with the LOCK prefix
2260    
2261      3. Register Allocation
2262    
2263          The module ra-rewrite-with-renaming has been improved.
2264    
2265      These have no effect on SML/NJ.
2266    
2267    ----------------------------------------------------------------------
2268    Name: Matthias Blume
2269    Date: 2000/05/15 16:20:00 JST
2270    Tag: blume-20000515-lightrebuild
2271    Description:
2272    
2273    1. I added an alternative to "-rebuild" to "makeml".  The difference is
2274       that prior to calling CMB.make' the CM-variable "LIGHT" will be
2275       defined.  In effect, the command will not build any cross-compiler
2276       backends and therefore finish more quickly.
2277    
2278       The "fixpt" script also takes a "-light" switch to be able to use
2279       this new facility while compiling for a fixpoint.
2280    
2281    2. I replaced all mentions of anchored paths in group owner specifications
2282       with simple relative paths (usually starting with "..").
2283       The rationale is that a library's internal workings should not be
2284       compromised by the lack of some anchor.  (An anchor is necessary
2285       for someone who wants to refer to the library by an anchored path,
2286       but it should not be necessary to build the same library in the first
2287       place.)
2288    
2289    3. I changed the way CM's tool mechanism determines the shell command
2290       string used for things like ml-yacc etc. so that it does not break
2291       when CM.Control.implicit_anchors is turned off.
2292    
2293  ----------------------------------------------------------------------  ----------------------------------------------------------------------
2294  Name: Matthias Blume  Name: Matthias Blume
2295  Date: 2000/05/12 18:20:00 JST  Date: 2000/05/12 18:20:00 JST
# Line 444  Line 2723 
2723    
2724  3.  Assembly  3.  Assembly
2725    
2726      When generating assemby, resolve the value of client defined constants,      When generating assembly, resolve the value of client defined constants,
2727      instead of generating symbolic values.  This is controlled by the      instead of generating symbolic values.  This is controlled by the
2728      new flag "asm-resolve-constants", which is default to true.      new flag "asm-resolve-constants", which is default to true.
2729    
# Line 467  Line 2746 
2746    
2747  To this end, I arranged that instead of "structure Core" as "structure  To this end, I arranged that instead of "structure Core" as "structure
2748  _Core" is bound in the pervasive environment.  Core access is done via  _Core" is bound in the pervasive environment.  Core access is done via
2749  _Core (which can never be accidentially rebound because _Core is not a  _Core (which can never be accidentally rebound because _Core is not a
2750  legal surface-syntax symbol).  legal surface-syntax symbol).
2751    
2752  The current solution is much cleaner because the core environment is  The current solution is much cleaner because the core environment is
# Line 477  Line 2756 
2756  with dynamic and symbolic parts of the core environment.  with dynamic and symbolic parts of the core environment.
2757    
2758  Remaining hackery (to bind the "magic" symbol _Core) is localized in the  Remaining hackery (to bind the "magic" symbol _Core) is localized in the
2759  compilation mananger's bootstrap compiler (actually: in the "init group"  compilation manager's bootstrap compiler (actually: in the "init group"
2760  handling).  See the comments in src/system/smlnj/init/init.cmi for  handling).  See the comments in src/system/smlnj/init/init.cmi for
2761  more details.  more details.
2762    
# Line 592  Line 2871 
2871     (specified in the .cm file at each instance where the tool's class is     (specified in the .cm file at each instance where the tool's class is
2872     used).     used).
2873    
2874     This was done to accomodate the new "make" and "shell" tools which     This was done to accommodate the new "make" and "shell" tools which
2875     facilitate fairly seemless hookup to portions of code managed using     facilitate fairly seamless hookup to portions of code managed using
2876     Makefiles or Shell scripts.     Makefiles or Shell scripts.
2877    
2878     There are no classes "shared" or "private" anymore.  Instead, the     There are no classes "shared" or "private" anymore.  Instead, the
# Line 605  Line 2884 
2884    
2885     All existing tools are described in the CM manual.     All existing tools are described in the CM manual.
2886    
2887     - Slightly better error handling.  (CM now surpresses many followup     - Slightly better error handling.  (CM now suppresses many followup
2888     error messages that tended to be more annoying than helpful.)     error messages that tended to be more annoying than helpful.)
2889    
2890  2. Major changes to the compiler's static environment data structures.  2. Major changes to the compiler's static environment data structures.
# Line 739  Line 3018 
3018    
3019       I've changed andl to testl in the floating point test sequence       I've changed andl to testl in the floating point test sequence
3020       whenever appropriate.  The Intel optimization guide states that       whenever appropriate.  The Intel optimization guide states that
3021       testl is perferable to andl.       testl is preferable to andl.
3022    
3023  7. RA (x86 only)  7. RA (x86 only)
3024    
# Line 921  Line 3200 
3200    
3201  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
3202  additional argument called "template" which is an optional string that  additional argument called "template" which is an optional string that
3203  specifiel the layout of the tool command line.  See the CM manual for  specifies the layout of the tool command line.  See the CM manual for
3204  explanation.  explanation.
3205    
3206  2. A special-purpose tool can be "regisitered" by simply dropping the  2. A special-purpose tool can be "registered" by simply dropping the
3207  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
3208  directory where the .cm file lives that uses this tool.  (The  directory where the .cm file lives that uses this tool.  (The
3209  behavior/misfeature until now was to look for the tool description  behavior/misfeature until now was to look for the tool description
# Line 968  Line 3247 
3247  rebuild.  Having sets of the form <base><k>.{bin,boot}.<arch>-unix for  rebuild.  Having sets of the form <base><k>.{bin,boot}.<arch>-unix for
3248  <k>=1,2,... is normally not a good idea when invoking fixpt.  However,  <k>=1,2,... is normally not a good idea when invoking fixpt.  However,
3249  they might be the result of an earlier partial run of fixpt (which  they might be the result of an earlier partial run of fixpt (which
3250  perhaps got accidentially killed).  In this case, fixpt will quickly  perhaps got accidentally killed).  In this case, fixpt will quickly
3251  move through what exists before continuing where it left off earlier,  move through what exists before continuing where it left off earlier,
3252  and, thus, saves a lot of time.  and, thus, saves a lot of time.
3253    
# Line 1018  Line 3297 
3297    it from that remote directory.    it from that remote directory.
3298    This should simplify installation further:  For machines that have    This should simplify installation further:  For machines that have
3299    access to the internet, just fetch <version>-config.tgz, unpack it,    access to the internet, just fetch <version>-config.tgz, unpack it,
3300    edit config/targets, and go (run config/install.sh).  The scipt will    edit config/targets, and go (run config/install.sh).  The script will
3301    fetch everything else that it might need all by itself.    fetch everything else that it might need all by itself.
3302    
3303    For CVS users, this mechanism is not relevant for source archives, but    For CVS users, this mechanism is not relevant for source archives, but

Legend:
Removed from v.644  
changed lines
  Added in v.915

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