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 653, Wed Jun 7 07:59:06 2000 UTC revision 829, Mon May 14 15:30:56 2001 UTC
# Line 11  Line 11 
11  Date:  Date:
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    
15    ----------------------------------------------------------------------
16    Name: Matthias Blume
17    Date: 2001/05/14 11:30:00 EDT
18    Tag: blume-20010514-script
19    Description:
20    
21    Some bugs in install script fixed.
22    
23    In addition to that I also made a slight change to the NLFFI API:
24    Functors generated by ml-nlffigen now take the dynamic library as a
25    straight functor argument, not as a suspended one.  (The original
26    functor code used to force the suspension right away anyway, so there
27    was nothing gained by this complication of the interface.)
28    
29    ----------------------------------------------------------------------
30    Name: Matthias Blume
31    Date: 2001/05/11 14:35:00 EDT
32    Tag: blume-20010511-ml-nlffi
33    Description:
34    
35    I finally took the plunge and added my new FFI code to the main
36    repository.  For x86-linux it is now ready for prime-time.
37    
38    There are two new subdirectories of "src":
39    
40      - ml-nlffi-lib:
41           The utility library for programs using the FFI interface.
42           Here is the implementation of $/c.cm and its associated low-level
43           partners $/c-int.cm and $/memory.cm.
44      - ml-nlffigen:
45           A stand-alone program for generating ML glue code from C source
46           code.
47    
48    Building ml-nlffigen requires $/ckit-lib.cm.
49    
50    The config/install.sh script has been updates to do the Right Thing
51    (hopefully).
52    
53    Notice that the source tree for the C-Kit will not be put under "src"
54    but directly under the installation root directory.  (This is the
55    structure that currently exists on the CVS server when you check out
56    module "sml".)  Fortunately, config/install.sh knows about this oddity.
57    
58    Bugs: No documentation yet.
59    
60    ----------------------------------------------------------------------
61    Name: Matthias Blume
62    Date: 2001/05/09 16:35:00 EDT
63    Tag: blume-20010509-cpscontract
64    Description:
65    
66    Fixed a bug in the accounting code in cpsopt/contract.sml.  (The
67    wrapper/unwrapper elimination did not decrement usage counts and some
68    dead variables got overlooked by the dead-up logic.)
69    
70    ----------------------------------------------------------------------
71    Name: Lal George
72    Date: 2001/05/08  17:26:09 EDT
73    Tag: george-20010508-omit-frameptr
74    Description:
75    
76    Changes to implement the omit-frame-pointer optimization to support
77    raw C calls. For now, there is only support on the Intel x86, but
78    other architectures will follow as more experience is gained with this.
79    
80    
81    ----------------------------------------------------------------------
82    Name: Matthias Blume
83    Date: 2001/05/07 14:40:00 EDT
84    Tag: blume-20010507-proxies
85    Description:
86    
87    I made into "proxy libraries" all libraries that qualify for such a
88    change.  (A qualifying library is a library that has another library or
89    groups as its sole member and repeats that member's export list
90    verbatim.  A proxy library avoids this repetition by omitting its export
91    list, effectively inheriting the list that its (only) member exports.
92    See the CM manual for more explanation.)
93    The main effect is that explicit export lists for these libraries
94    do not have to be kepts in sync, making maintenance a bit easier.
95    
96    I also added copyright notices to many .cm-files.
97    
98    Last but not least, I made a new set of bootfiles.
99    
100    ----------------------------------------------------------------------
101    Name: Matthias Blume
102    Date: 2001/05/04 17:00:00 EDT
103    Tag: blume-20010504-cm-lsplit
104    Description:
105    
106    0. John merged pending changes to $/smlnj-lib.cm
107    
108    1. Allen's previous change accidentally backed out of one of Lal's
109       earlier changes.  I undid this mistake (re-introducing Lal's change).
110    
111    2. I used the new topOrder' function from graph-scc.sml (from $/smlnj-lib.cm)
112       within the compiler where applicable.  There is some code simplification
113       because of that.
114    
115    3. The "split" phase (in FLINT) is now part of the default list of phases.
116       Compiler.Control.LambdaSplitting.* can be used to globally control the
117       lambda-splitting (cross-module-inlining) engine.  In addition to that,
118       it can now also be controlled on a per-source basis: CM has been taught
119       a new tool parameter applicable to ML source files.
120    
121       - To turn lambda-splitting off completely:
122            local open Compiler.Control.LambdaSplitting in
123                val _ = set Off
124            end
125       - To make "no lambda-splitting" the global default (but allow per-source
126         overriding); this is the initial setting:
127            local open Compiler.Control.LambdaSplitting in
128                val _ = set (Default NONE)
129            end
130       - To make "lambda-splitting with aggressiveness a" the global default
131         (and allow per-source overriding):
132            local open Compiler.Control.LambdaSplitting in
133                val _ = set (Default (SOME a))
134            end
135    
136       - To turn lambda-splitting off for a given ML souce file (say: a.sml)
137         write (in the respective .cm-file):
138            a.sml (lambdasplitting:off)
139       - To turn lambda-splitting for a.sml on with minimal aggressiveness:
140            a.sml (lambdasplitting:on)
141       - To turn lambda-splitting for a.sml on with aggressiveness <a> (where
142         <a> is a decimal non-negative integer):
143            a.sml (lambdasplitting:<a>)
144       - To turn lambda-splitting for a.sml on with maximal aggressiveness:
145            a.sml (lambdasplitting:infinity)
146       - To use the global default for a.sml:
147            a.sml (lambdasplitting:default)
148         or simply
149            a.sml
150    
151    ----------------------------------------------------------------------
152    Name: Allen Leung
153    Date: 2001/05/04 01:57:00 EDT
154    Tag: leunga-20010504-sync
155    Description:
156    
157      MLRISC features.
158    
159      1. Fix to CMPXCHG instructions.
160      2. Changed RA interface to allow annotations in callbacks.
161      3. Added a new method to the stream interface to allow annotations updates.
162    
163    ----------------------------------------------------------------------
164    Name: Matthias Blume
165    Date: 2001/05/01 11:45:00 EDT
166    Tag: blume-20010501-pcedittmp
167    Description:
168    
169    Changed install.sh to use the current working directory instead of
170    /usr/tmp for a temporary file (pcedittmp).  The previous choice
171    of /usr/tmp caused trouble with MacOS X because of file premission
172    problems.
173    
174    ----------------------------------------------------------------------
175    Name: Matthias Blume
176    Date: 2001/04/20 11:10:00 EDT
177    Tag: blume-20010420-inMLflag
178    Description:
179    
180     - added vp_limitPtrMask to vproc-state.h
181       (for use by the raw-C-calls mechanism to implement proper interrupt
182        handling)
183     - made the ML compiler aware of various data-structure offsets so it
184       can generate code for accessing the vp_inML flag and vp_limitPtrMask
185     - tweaked mlriscGen.sml to have it emit interrupt-handling code for
186       raw C-calls
187    
188    ----------------------------------------------------------------------
189    Name: Lal George
190    Date: 2001/04/20 09:15:28 EDT
191    Tag: george-20010420-macosX
192    Description:
193    
194     - Changes to port to Mac OS X; Darwin.
195    
196     - In the process I found that sqrt was broken on the PPC, because the
197       fsqrt instruction is not implemented.
198    
199    ----------------------------------------------------------------------
200    Name: Matthias Blume
201    Date: 2001/04/18 12:45:00 EDT
202    Tag: blume-20010418-ccalls
203    Description:
204    
205     - fixed two off-by-4 errors in the x86-specific c-calls implementation
206       (this bug prevented structure arguments containing pointers from being
207        passed correctly)
208     - changed the raw-C-call code in mlriscGen.sml in such a way that
209       structure arguments are represented as a pointer to the beginning
210       of the structure (instead of having a series of synthesized arguments,
211       one for each structure member)
212    
213     - made makeml script's verbosity level configurable via environment
214       variable (MAKEML_VERBOSITY)
215    
216     - eliminated placeholder implementations for f32l, w16s, i16s, and f32s
217       in rawmem-x86.sml; we are now using the real thing
218    
219    ----------------------------------------------------------------------
220    Name: Matthias Blume
221    Date: 2001/03/22 16:25:00 EST
222    Tag: blume-20010322-bootfiles
223    Description:
224    
225    Created a new set of bootfiles (for your automatic installation convenience).
226    
227    ----------------------------------------------------------------------
228    Name: Matthias Blume
229    Date: 2001/03/22 15:10:00 EST
230    Tag: blume-20010322-rawmem-parcm
231    Description:
232    
233    1. All "raw memory access" primitives for the new FFI are implemented now
234       (at least on the x86).
235    2. Some further cleanup of CM's parallel make mechanism.
236    
237    ----------------------------------------------------------------------
238    Name: Matthias Blume
239    Date: 2001/03/19 17:53:00 EST
240    Tag: blume-20010319-parallel
241    Description:
242    
243    Parallel make (using compile servers) now works again.
244    
245    To this end, CM.stabilize and CMB.make have been modified to work in
246    two passes when compile servers are attached:
247       1. Compile everything, do not perform stabilization; this pass
248          uses compile servers
249       2. Stabilize everything; this pass does not use compile servers
250    If there are no compile servers, the two passes are combined into one
251    (as before).  Splitting the passes increases the inherent parallelism
252    in the dependency graph because the entire graph including all
253    libraries is available at the same time.  This, in turn, improves
254    server utilization.  The downside is that the master process will
255    have to do some extra work after compilation is done (because for
256    technical reasons it must re-read all the binfiles during stabilization).
257    
258    ----------------------------------------------------------------------
259    Name: Matthias Blume
260    Date: 2001/03/16 12:22:00 EST
261    Tag: blume-20010316-bootfiles
262    Description:
263    
264    Created a new set of bootfiles (for your automatic installation convenience).
265    
266    ----------------------------------------------------------------------
267    Name: Matthias Blume
268    Date: 2001/03/16 11:00:00 EST
269    Tag: blume-20010316-MLTREE-fixup
270    Description:
271    
272    This is a minor fixup for an (untagged) earlier commit by Allen.
273    (A file was missing).
274    
275    ----------------------------------------------------------------------
276    Name: Allen Leung
277    Date: Mon Mar  5 18:54:57 EST 2001
278    Tag: leunga-20010305-cut-support
279    
280    1. New support for alternative control-flow in MLTREE.
281       Currently we support
282    
283          FLOW_TO(CALL ...., [k1,...,kn])
284    
285       This is needed for 'cuts to' in C-- and try/handle-like constructs
286       in Moby
287    
288       New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
289    
290    2. Register Allocator
291    
292       Changes in interface [from Fermin, John]
293    
294    3. Alpha 8-bit SLL support [Fermin]
295    
296    4. All architectures
297    
298       A new module (ClusterExpandCopies) for expanding parallel copies.
299    
300    ----------------------------------------------------------------------
301    Name: Allen Leung
302    Date: 2001/02/27 23:07:00 EST
303    Tag: leunga-20010227-minor-stuff
304    
305    1. Alpha bug fix for CMOVNE
306    2. Handle mltree COND(..,FCMP ...,...)
307    3. Bug fix in simplifier
308    
309    ----------------------------------------------------------------------
310    Name: Matthias Blume
311    Date: 2001/01/30 17:50:00 EST
312    Tag: blume-20010130-sync
313    Description:
314    
315    This is just a minor update to sync my devel branch with the main brach.
316    The only visible change is the addition of some README files.
317    
318    ----------------------------------------------------------------------
319    Name: Matthias Blume
320    Date: 2001/01/12 23:30:00 JST
321    Tag: blume-20010112-bootfiles
322    Description:
323    
324    Made a new set of bootfiles that goes with the current state of the
325    repository.
326    
327    ----------------------------------------------------------------------
328    Name: Matthias Blume
329    Date: 2001/01/12 21:20:00 JST
330    Tag: blume-20010112-sync
331    Description:
332    
333    I am just flushing out some minor changes that had accumulated in
334    my private branch in order to sync with the main tree.  (This is
335    mainly because I had CVS trouble when trying to merge _into_ my
336    private branch.)
337    
338    Most people should be completely unaffected by this.
339    
340    ----------------------------------------------------------------------
341    Name: Allen Leung
342    Date: Thu Jan 11 21:03:00 EST 2001
343    Tag: leunga-20010111-labexp=mltree
344    Description:
345    
346    1.  Removed the type LabelExp and replace it by MLTree.
347    2.  Rewritten mltree-simplify with the pattern matcher tool.
348    3.  There were some bugs in alpha code generator which would break
349        64-bit code generation.
350    4.  Redo the tools to generate code with the
351    5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
352        are now generated by perl scripts.
353    
354    ----------------------------------------------------------------------
355    Name: Matthias Blume
356    Date: 2001/01/10 21:55:00 JST
357    Tag: blume-20010110-rcc
358    Description:
359    
360    The RCC stuff now seems to work (but only on the x86).
361    This required hacking of the c-calls interface (and -implementation) in
362    MLRISC.
363    
364    Normal compiler users should be unaffected.
365    
366    ----------------------------------------------------------------------
367    Name: Matthias Blume
368    Date: 2001/01/09 01:20:00 JST
369    Tag: blume-20010109-rcc
370    Description:
371    
372    This is a fairly big patch, flushing out a large number of pending
373    changes that I made to my development copy over the last couple of days.
374    
375    Of practical relevance at this moment is a workaround for a pickling
376    bug that Allen ran into the other day.  The cause of the bug itself is
377    still unknown and it might be hard to fix it properly, but the
378    workaround has some merits of its own (namely somewhat reducing pickling
379    overhead for certain libraries).  Therefore, I think this solution should
380    be satisfactory at this time.
381    
382    The rest of the changes (i.e., the vast majority) has to do with my
383    ongoing efforts of providing direct support for C function calls from
384    ML.  At the moment there is a new primop "RAW_CCALL", typing magic
385    in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
386    case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
387    translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
388    handle RCC.
389    
390    The last part (the changes to mlriscGen.sml) are still known to be
391    wrong on the x86 and not implemented on all other architectures.  But
392    the infrastructure is in place. I had to change a few functor
393    signatures in the backend to be able to route the CCalls interface
394    from MLRISC there, and I had to specialize the mltree type (on the
395    x86) to include the necessary extensions. (The extensions themselves
396    were already there and redy to go in MLRISC/x86).
397    
398    Everything should be very happy as soon as someone helps me with
399    mlriscGen.sml...
400    
401    In any case, nothing of this should matter to anyone as long as the
402    new primop is not being used (which is going to be the case unless you
403    find it where I hid it :). The rest of the compiler is completely
404    unaffected.
405    
406    ----------------------------------------------------------------------
407    Name: Matthias Blume
408    Date: 2001/01/05 00:30:00 JST
409    Tag: blume-20010105-primops
410    Description:
411    
412    Added some experimental support for work that I am doing right now.
413    These changes mostly concern added primops, but there is also a new
414    experimental C library in the runtime system (but currently not enabled
415    anywhere except on Linux/X86).
416    
417    In the course of adding primops (and playing with them), I discovered that
418    Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
419    broken.  (Zhong was very right he labeled this stuff as "major gross hack".)
420    To recover, I made type information in INL_PRIM mandatory and changed
421    prim.sml as well as built-in.sml accordingly.  The InLine structure now
422    has complete, correct type information (i.e., no bottom types).
423    
424    Since all these changes mean that we need new binfiles, I also bumped the
425    version number to 110.32.1.
426    
427    ----------------------------------------------------------------------
428    Name: Matthias Blume
429    Date: 2000/12/30 22:10:00 JST
430    Tag: blume-20001230-various
431    Description:
432    
433    Added proxy libraries for MLRISC and let MLRISC libraries refer
434    to each other using path anchors.  (See CM manual for explanation.)
435    
436    Updated CM documentation.
437    
438    Fixed some bugs in CM.
439    
440    Implemented "proxy" libraries (= syntactic sugar for CM).
441    
442    Added "-quiet" option to makeml and changed runtime system accordingly.
443    
444    Added cleanup handler for exportML to reset timers and compiler stats.
445    
446    ----------------------------------------------------------------------
447    Name: Lal George
448    Date: 2000/12/22 22:22:58 EST 2000
449    Tag: Release_110_32
450    Description:
451    
452            Infinite precision used throughout MLRISC.
453            see MLRISC/mltree/machine-int.sig
454    
455    ----------------------------------------------------------------------
456    Name: Matthias Blume
457    Date: 2000/12/22 23:16:00 JST
458    Tag: blume-20001222-warn
459    Description:
460    
461    Corrected wording and formatting of some CM warning message which I
462    broke in my previous patch.
463    
464    ----------------------------------------------------------------------
465    Name: Matthias Blume
466    Date: 2000/12/22 21:20:00 JST
467    Tag: blume-20001222-anchorenv
468    Description:
469    
470    Fixed CM's handling of anchor environments in connection with CMB.make.
471    
472    ----------------------------------------------------------------------
473    Name: Matthias Blume
474    Date: 2000/12/22 13:15:00 JST
475    Tag: blume-20001222-cleanup
476    Description:
477    
478    Removed src/cm/ffi which does not (and did not) belong here.
479    
480    ----------------------------------------------------------------------
481    Name: Matthias Blume
482    Date: 2000/12/21 23:55:00 JST
483    Tag: blume-20001221-exn
484    Description:
485    
486    Probably most important: CM no longer silently swallows all exceptions
487    in the compiler.
488    Plus: some other minor CM changes.  For example, CM now reports some
489    sizes for generated binfiles (code, data, envpickle, lambdapickle).
490    
491    ----------------------------------------------------------------------
492    Name: Matthias Blume
493    Date: 2000/12/15 00:01:05 JST
494    Tag: blume-20001215-dirtool
495    Description:
496    
497    - "dir" tool added.
498    - improvements and cleanup to Tools structure
499    - documentation updates
500    
501    ----------------------------------------------------------------------
502    Name: Allen Leung
503    Date: Thu Dec 14 03:45:24 EST 2000
504    Description:
505    Tag:  leunga-20001214-int-inf
506    Description:
507    
508       In IntInf, added these standard functions, which are missing from our
509    implementation:
510    
511        andb : int * int -> int
512        xorb : int * int -> int
513        orb  : int * int -> int
514        notb : int -> int
515         <<   : int * word -> int
516        ~>>  : int * word -> int
517    
518       Not tested, I hope they are correct.
519    
520    ----------------------------------------------------------------------
521    Name: Allen Leung
522    Date: Fri Dec  8 19:23:26 EST 2000
523    Description:
524    Tag:  leunga-20001208-nowhere
525    Description:
526    
527      Slight improvements to the 'nowhere' tool to handle OR-patterns,
528    to generate better error messages etc.  Plus a brief manual.
529    
530    ----------------------------------------------------------------------
531    Name: Lal George
532    Date: 2000/12/08 09:54:02 EST 2000
533    Tag: Release_110_31
534    Description:
535    
536    - Version 110.31
537    ----------------------------------------------------------------------
538    Name: Allen Leung
539    Date: Thu Dec  7 22:01:04 EST 2000
540    Tag:  leunga-20001207-cell-monster-hack
541    Description:
542    
543    Major MLRISC internal changes.  Affect all clients.
544    Summary:
545    
546    1.  Type CELLS.cell = int is now replaced by a datatype.
547        As a result, the old regmap is now gone.  Almost all interfaces
548        in MLRISC change as a consequence.
549    
550    2.  A new brand version of machine description tool (v3.0) that generates
551        modules expecting the new interface.  The old version is removed.
552    
553    3.  The RA interface has been further abstracted into two new functors.
554        RISC_RA and X86RA.  These functors have much simpler interfaces.
555        [See also directory MLRISC/demo.]
556    
557    4.  Some other new source->source code generation tools are available:
558    
559        a. MLRISC/Tools/RewriteGen -- generate rewriters from rules.
560        b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules.
561           I use this tool to generate the peephole optimizers---with the new
562           cell type changes, peephole rules are becoming difficult to write
563           without conditional pattern matching.
564    
565    5.  More Intmap -> IntHashTable change.  Previous changes by Matthias didn't
566        cover the entire MLRISC source tree so many things broke.
567    
568    6.  CM files have been moved to the subdirectory MLRISC/cm.
569        They are moved because there are a lot of them and they clutter up the
570        root dir.
571    
572    7.  More detailed documentation to come...
573    
574        NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do
575        a makeml -rebuild first.  This is because of other other
576        changes that Matthias has made (see below).
577    
578    
579    ----------------------------------------------------------------------
580    Name: Matthias Blume
581    Date: 2000/11/30 23:12:00 JST
582    Tag: blume-20001130-filereorg
583    Description:
584    
585    Some manual updates and some file reorganizations in CM.
586    
587    ----------------------------------------------------------------------
588    Name: Matthias Blume
589    Date: 2000/11/24 17:45:00 JST
590    Tag: blume-20001124-link
591    Description:
592    
593    Drastically improved link traversal code for the case that the dynamic
594    value was already loaded at bootstrap time.  As a result, CM and CMB
595    now both load blazingly fast -- even on a very slow machine.  Also,
596    memory consumption has been further reduced by this.
597    
598    Warning: The format of the PIDMAP file has changed.  THerefore, to
599    bootstrap you have to do this:
600    
601    1. Run CMB.make
602    2. Make a symbolic link for the boot directory:
603         ln -s sml.boot.ARCH-OS xxx
604    3. "Rebuild" the boot directory:
605         ./makeml -boot xxx -rebuild sml ; rm xxx
606    4. Boot normally:
607          ./makeml
608    
609    ----------------------------------------------------------------------
610    Name: Matthias Blume
611    Date: 2000/11/21 21:20:00 JST
612    Tag: blume-20001121-tools
613    Description:
614    
615    Continued hacking on autoloading problem -- with success this time.
616    Also changed tool-plugin mechanism.  See new CM manual.
617    
618    ----------------------------------------------------------------------
619    Name: Matthias Blume
620    Date: 2000/11/19 14:30:00 JST
621    Tag:  blume-20001119-autoload
622    Description:
623    
624    Some hacking to make autoloading faster.  Success for CMB, no success
625    so far for CM.  There is a reduced structure CM' that autoloads faster.
626    (This is a temporary, non-documented hack to be eliminated again when
627    the general problem is solved.)
628    
629    ----------------------------------------------------------------------
630    Name: Matthias Blume
631    Date: 2000/11/17 14:10:00 JST
632    Tag: blume-20001117-pickle-lib
633    Description:
634    
635    1. Eliminated comp-lib.cm
636    2. Made pickle-lib.cm
637    3. Eliminated all uses of intset.sml (from comp-lib.cm)
638    4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with
639       equivalent constructs from smlnj-lib.cm (INtHashTable).
640    5. Point 4. also goes for those uses of intmap.* in MLRISC.
641       Duplicated intmap modules thrown out.
642    6. Hunted down all duplicated SCC code and replaced it with
643       equivalent stuff (GraphSCCFn from smlnj-lib.cm).
644    7. Rewrote Feedback module.
645    8. Moved sortedlist.sml into viscomp-lib.cm.  Eventually it
646       should be thrown out and equivalent modules from smlnj-lib.cm
647       should be used (IntRedBlackSet, IntListSet, ...).
648    
649    Confirmed that compiler compiles to fixpoint.
650    
651    ----------------------------------------------------------------------
652    Name: Allen Leung
653    Date: 2000/11/10 18:00:00
654    Tag: leunga-20001110-new-x86-fp
655    
656    A new x86 floating point code generator has been added.
657    By default this is turned off.  To turn this on, do:
658    
659        CM.autoload "$smlnj/compiler.cm";
660        Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true;
661    
662    Changes:
663    
664    1.  Changed FTAN to FPTAN so that the assembly output is correct.
665    2.  Changed the extension callback for FTANGENT to generate:
666    
667              fptan
668              fstp  %st(0)
669        instead of
670              fptan
671              fstpl ftempmem
672    
673    3.  Numerous assembly fixes for x86.
674    
675    5.  Cleaned up the machine code output module x86/x86MC.sml and added
676        support for a whole bunch of instructions and addressing modes:
677    
678          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st, %st(n)
679          faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp  %st, %st(n)
680          fadd/fsub/fsubr/fmul/fdiv/fdivr  %st(n), %st
681          fiadd/fisub/fisubr/fimul/fidiv/fidivr mem
682          fxch %st(n)
683          fld %st(n)
684          fst %st(n)
685          fst mem
686          fstp %st(n)
687          fucom %st(n)
688          fucomp %st(n)
689    
690        All these are now generated when the fast fp mode is turned on.
691    
692    6.  Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs
693    
694    ----------------------------------------------------------------------
695    Name: Matthias Blume
696    Date: 2000/11/09 11:20:00 JST
697    Tag: blume-20001109-scc
698    Description:
699    
700    Eliminated some code duplication:
701    
702    1. Added "where" clause to GraphSCCFn in SML/NJ Library.
703       (Otherwise the functor is useless.)
704    2. Used GraphSCCFn where SCCUtilFun was used previously.
705    3. Got rid of SCCUtilFun (in comp-lib.cm).
706    
707    ----------------------------------------------------------------------
708    Name: Lal George
709    Date: 2000/11/06 09:02:21 EST 2000
710    Tag: Release_110_30
711    Description:
712    
713    - Version 110.30
714    ----------------------------------------------------------------------
715    Name: Matthias Blume
716    Date: 2000/11/04 14:45:00
717    Tag: blume-20001104-mlbuild
718    Description:
719    
720    - Made ml-build faster on startup.
721    - Documentation fixes.
722    
723    ----------------------------------------------------------------------
724    Name: Matthias Blume
725    Date: 2000/11/02 17:00:00 JST
726    Tag: blume-20001102-condcomp
727    Description:
728    
729    - Small tweaks to pickler -- new BOOTFILES!
730    - Version bumped to 110.29.2.
731    - Added conditional compilation facility to init.cmi (see comment there).
732    ----------------------------------------------------------------------
733    Name: Allen Leung
734    Date: 2000/10/23 19:31:00
735    Tag: leunga-20001023-demo-ra
736    
737    1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only)
738    2. Test programs for the graph library updated
739    3. Some new MLRISC demo programs added
740    
741    ----------------------------------------------------------------------
742    Name: Matthias Blume
743    Date: 2000/08/31 22:15:00 JST
744    Tag: blume-20001017-errmsg
745    Description:
746    
747    More error message grief: Where there used to be no messages, there
748    now were some that had bogus error regions.  Fixed.
749    
750    ----------------------------------------------------------------------
751    Name: Matthias Blume
752    Date: 2000/08/31 17:30:00 JST
753    Tag: blume-20001017-v110p29p1
754    Description:
755    
756    I made a version 110.29.1 with new bootfiles.
757    
758    Changes:  Modified pickler/unpickler for faster and leaner unpickling.
759              CM documentation changes and a small bugfix in CM's error reporting.
760    
761    ----------------------------------------------------------------------
762    Name: Lal George
763    Date: 2000/09/27 14:42:35 EDT
764    Tag: george-20000927-nodestatus
765    Description:
766    
767    Changed the type of the nodestatus, so that:
768    
769            SPILLED(~1)             is now SPILLED
770            SPILLED(m) where m>=0   is now MEMREG(m)
771            SPILLED(s) where s<~1   is now SPILL_LOC(~s)
772    
773    ----------------------------------------------------------------------
774    Name: Matthias Blume
775    Date: 2000/09/07 14:45:00 JST
776    Tag: blume-20000907-cmerrmsg
777    Description:
778    
779    Small tweak to CM to avoid getting ML syntax error messages twice.
780    
781    ----------------------------------------------------------------------
782    Name: Matthias Blume
783    Date: 2000/08/31 18:00:00 JST
784    Tag: blume-20000831-cvsbootfiles
785    Description:
786    
787    New URL for boot files (because the 110.29 files on the BL server do
788    now work correctly with my updated install scripts for yacc and lex).
789    
790    ----------------------------------------------------------------------
791    Name: Matthias Blume
792    Date: 2000/08/08 12:33:00 JST
793    Tag: blume-20000808-manual
794    Description:
795    
796    Tiny update to CM manual.
797    
798    ----------------------------------------------------------------------
799    Name: Allen Leung
800    Date: 2000/08/7 19:31:00
801    Tag: leunga-20000807-a-whole-bunch-of-stuff
802    
803      Moby, C--, SSA, x86, machine descriptions etc.  Should only affect C--
804    and Mobdy.
805    
806    1.  x86
807    
808       a.  Fixes to peephole module by John and Dan.
809       b.  Assembly fix to SETcc by Allen.
810       c.  Fix to c-call by John.
811       d.  Fix to spilling by John.  (This one deals with the missing FSTPT case)
812       e.  Instruction selection optimization to SETcc as suggested by John.
813    
814           For example,
815    
816            MV(32, x, COND(32, CMP(32, LT, a, b), LI 1, LI 0))
817    
818           should generate:
819    
820            MOVL a, x
821            SUBL b, x
822            SHRL 31, x
823    
824    2.  IR stuff
825    
826         A bunch of new DJ-graph related algorithms added.  These
827         speed up SSA construction.
828    
829    3.  SSA + Scheduling
830    
831         Added code for SSA and scheduling to the repository
832    
833    ----------------------------------------------------------------------
834    Name: Lal George
835    Date: 2000/07/27 11:53:14 EDT
836    
837    Tag: lal-20000727-linux-ppc
838    Description:
839    
840     Made changes to support Linux PPC.
841     p.s. I have confirmation that the 110.29 boot files work fine.
842    
843    ----------------------------------------------------------------------
844    Name: Matthias Blume
845    Date: 2000/07/27 17:40:00 JST
846    Tag: blume-20000727-scripts
847    Description:
848    
849    !!!! WARNING !!!!
850    You must recompile the runtime system!
851    !!!! WARNING !!!!
852    
853    This is basically another round of script-enhancements:
854    
855    1. sml, ml-build, and ml-makedepend accept options -D and -U to define
856       and undefine CM preprocessor symbols.
857    
858    2. ml-build avoids generating a new heap image if it finds that the
859       existing one is still ok.  (The condition is that no ML file had to
860       be recompiled and all ML files are found to be older that the heap
861       file.)
862    
863       To make this work smoothly, I also hacked the runtime system as
864       well as SMLofNJ.SysInfo to get access to the heap image suffix
865       (.sparc-solaris, ...) that is currently being used.
866    
867       Moreover, the signature of CM.mk_standalone has changed.  See the
868       CM manual.
869    
870    3. ml-makedepend accepts additional options -n, -a, and -o.  (See the
871       CM manual for details.)
872    
873    4. More CM manual updates:
874        - all of the above has been documented.
875        - there is now a section describing the (CM-related) command line
876          arguments that are accepted by the "sml" command
877    
878    ----------------------------------------------------------------------
879    Name: Matthias Blume
880    Date: 2000/07/25 16:20:00 JST
881    Tag: blume-20000725-makedepend
882    Description:
883    
884    Added a script called ml-makedepend.  This can be used in makefiles
885    for Unix' make in a way very similar to the "makedepend" command for
886    C.
887    
888    The script internally uses function CM.sources.
889    
890    Synopsis:
891    
892        ml-makedepend [-f makefile] cmfile targetname
893    
894    The default for the makefile is "makefile" (or "Makefile" should
895    "makefile" not exist).
896    
897    ml-makedepend adds a cmfile/targetname-specific section to this
898    makefile (after removing the previous version of this section).  The
899    section contains a single dependency specification with targetname on
900    the LHS (targetname is an arbitrary name), and a list of files derived
901    from the cmfile on the RHS.  Some of the files on the RHS are
902    ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
903    "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
904    path names.  The makefile writer is responsible for making sure that
905    these variables have correct at the time "make" is invoked.
906    
907    ----------------------------------------------------------------------
908    Name: Matthias Blume
909    Date: 2000/07/22 23:30:00 JST
910    Tag: blume-20000722-urlupdate
911    Description:
912    
913    Changed BOOT and config/srcarchiveurl to point to BL server:
914    
915        ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
916    
917    ----------------------------------------------------------------------
918    Name: Matthias Blume
919    Date: 2000/07/18 18:00:00 JST
920    Tag: blume-20000718-Version_110_29
921    Description:
922    
923    1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
924    
925    2. Updated config/version to 110.29
926    
927    3. Updated config/srcarchiveurl
928    
929    3. New boot files!
930       ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
931    
932    ----------------------------------------------------------------------
933    Name: Matthias Blume
934    Date: 2000/07/11 13:58:00 JST
935    Tag: blume-20000711-doctypo
936    Description:
937    
938    Fixed a few typos in CM manual.
939    
940    ----------------------------------------------------------------------
941    Name: Allen Leung
942    Date: 2000/06/15 00:38:00
943    Tag: leunga-20000704-sparc-x86
944    
945    1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
946    2. fix to x86 RET bug [found by Dan Grossman]
947    3. sparc assembly bug fix for ticc instructions [found by Fermin]
948    
949       Affects c-- and moby only
950    
951    ----------------------------------------------------------------------
952    Name: Matthias Blume
953    Date: 2000/07/04 15:26:00
954    Tag: blume-20000704-trigger
955    Description:
956    
957    1. Improvements to CM manual.
958    2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
959       of getting a back-trace.  The function, when called, raises an
960       internal exception which explicitly carries the full back-trace history,
961       so it is unaffected by any intervening handle-raise pairs ("trivial"
962       or not).  The interactive loop will print that history once it arrives
963       at top level.
964       Short of having all exceptions implicitly carry the full history, the
965       recommended way of using this facility is:
966         - compile your program with instrumentation "on"
967         - run it, when it raises an exception, look at the history
968         - if the history is "cut off" because of some handler, go and modify
969           your program so that it explicitly calls BTrace.trigger
970         - recompile (still instrumented), and rerun; look at the full history
971    
972    ----------------------------------------------------------------------
973    Name: Matthias Blume
974    Date: 2000/07/03 15:36:00 JST
975    Tag: blume-20000702-manual
976    Description:
977    
978    Small corrections and updates to CM manual.
979    
980    ----------------------------------------------------------------------
981    Name: Matthias Blume
982    Date: 2000/06/29 16:04:00 JST
983    Tag: blume-20000629-yacctool
984    Description:
985    
986    Changes:
987    
988    1. Class "mlyacc" now takes separate arguments to pass options to
989       generated .sml- and .sig-files independently.
990    2. Corresponding CM manual updates.
991    3. BTrace module now also reports call sites.  (However, for loop clusters
992       it only shows from where the cluster was entered.)  There are associated
993       modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
994    
995    ----------------------------------------------------------------------
996    Name: Matthias Blume
997    Date: 2000/06/27 16:51:00 JST
998    Tag: blume-20000627-noweb
999    Description:
1000    
1001    Changes:
1002    
1003     1. Implemented "subdir" and "witness" options for noweb tool.
1004        This caused some slight internal changes in CM's tool implementation.
1005     2. Fixed bug in "tool plugin" mechanism.  This is essentially cleaning
1006        some remaining issues from earlier path anchor changes.
1007     3. Updated CM manual accordingly.
1008    
1009     4. Changed implementation of back-tracing so that I now consider it
1010        ready for prime-time.
1011    
1012        In particular, you don't have to explicitly trigger the back-trace
1013        anymore.  Instead, if you are running BTrace-instrumented code and
1014        there is an uncaught exception (regardless of whether or not it was
1015        raised in instrumented code), the top-level evalloop will print
1016        the back-trace.
1017    
1018        Features:
1019    
1020          - Instrumented and uninstrumented code work together seemlessly.
1021            (Of course, uninstrumented code is never mentioned in actual
1022             back-traces.)
1023    
1024          - Asymptotic time- and space-complexity of instrumented code is
1025            equal to that of uninstrumented code.  (This means that
1026            tail-recursion is preserved by the instrumentation phase.)
1027    
1028          - Modules whose code has been instrumented in different sessions
1029            work together without problem.
1030    
1031          - There is no penalty whatsoever on uninstrumented code.
1032    
1033          - There is no penalty on "raise" expressions, even in
1034            instrumented code.
1035    
1036        A potential bug (or perhaps it is a feature, too):
1037    
1038          A back-trace reaches no further than the outermost instrumented
1039          non-trivial "raise".  Here, a "trivial" raise is one that is the
1040          sole RHS of a "handle" rule.  Thus, back-traces reach trough
1041    
1042               <exp> handle e => raise e
1043    
1044          and even
1045    
1046               <exp> handle Foo => raise Bar
1047    
1048          and, of course, through
1049    
1050               <exp> handle Foo => ...
1051    
1052         if the exception was not Foo.
1053    
1054         Back-traces always reach right through any un-instrumented code
1055         including any of its "handle" expressions, trivial or not.
1056    
1057       To try this out, do the following:
1058    
1059         - Erase all existing binfiles for your program.
1060           (You may keep binfiles for those modules where you think you
1061            definitely don't need back-tracing.)
1062         - Turn on back-trace instrumentation:
1063              SMLofNJ.Internals.BTrace.mode (SOME true);
1064         - Recompile your program.  (I.e., run "CM.make" or "use".)
1065         - You may now turn instrumentation off again (if you want):
1066              SMLofNJ.Internals.BTrace.mode (SOME false);
1067         - Run your program as usual.  If it raises an exception that
1068           reaches the interactive toplevel, then a back-trace will
1069           automatically be printed.  After that, the toplevel loop
1070           will print the exception history as usual.
1071    
1072    ----------------------------------------------------------------------
1073    Name: Matthias Blume
1074    Date: 2000/06/26 09:56:46 JST
1075    Tag: blume-20000626-setup
1076    Description:
1077    
1078    CM: - setup-parameter to "sml" added; this can be used to run arbitrary
1079          ML code before and after compiling a file (e.g., to set compiler
1080          flags)
1081    
1082    Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
1083              - associated changes to btrace.sml (BTrace instrumentation pass)
1084              - cleaner implementation of btimp.sml (BTrace tracing and report
1085                module)
1086    
1087    CM manual: * new path encoding documented
1088               * description of setup-parameter to "sml" added
1089    
1090    The biggest user-visible change to back-tracing is that it is no
1091    longer necessary to compile all traced modules within the same
1092    session.  (This was a real limitation.)
1093    
1094    ----------------------------------------------------------------------
1095    Name: Matthias Blume
1096    Date: 2000/06/24 12:40:00 JST
1097    Tag: blume-20000624-startup
1098    Description:
1099    
1100    Fixes startup slowdown problem.  (I was calling SrcPath.sync a _tad_
1101    bit too often -- to put it mildly. :)
1102    
1103    ----------------------------------------------------------------------
1104    Name: Matthias Blume
1105    Date: 2000/06/23 18:20:00 JST
1106    Tag: blume-20000623-btrace
1107    Description:
1108    
1109    This updates adds a backtrace facility to aid programmers in debugging
1110    their programs.  This involves the following changes:
1111    
1112    1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
1113       keeping track of the current call stack.  When programs are compiled
1114       in a special mode, the compiler will insert calls to these hooks
1115       into the user program.
1116       "Hook" means that it is possible for different implementations of
1117       back-tracing to register themselves (at different times).
1118    
1119    2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
1120       as an Absyn.dec->Absyn.dec rewrite.  Normally this phase is turned off.
1121       It can be turned on using this call:
1122         SMLofNJ.Internals.BTrace.mode (SOME true);
1123       Turning it off again:
1124         SMLofNJ.Internals.BTrace.mode (SOME false);
1125       Querying the current status:
1126         SMLofNJ.Internals.BTrace.mode NONE;
1127       Annotated programs are about twice as big as normal ones, and they
1128       run a factor of 2 to 4 slower with a dummy back-trace plugin (one
1129       where all hooks do nothing).  The slowdown with a plugin that is
1130       actually useful (such as the one supplied by default) is even greater,
1131       but in the case of the default plugin it is still only an constant
1132       factor (amortized).
1133    
1134    3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
1135       with a sub-structure BTrace for controlling back-tracing.  In particular,
1136       the above-mentioned function "mode" controls whether the annotation
1137       phase is invoked by the compiler.  Another important function is
1138       "trigger": when called it aborts the current execution and causes
1139       the top-level loop to print a full back-trace.
1140    
1141    4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
1142       for back-tracing.  It keeps track of the dynamic call stack and in
1143       addition to that it keeps a partial history at each "level" of that
1144       stack.  For example, if a tail-calls b, b tail-calls c, and c tail-calls
1145       d and b (at separate times, dynamically), then the report will show:
1146    
1147       GOTO   d
1148             /c
1149       GOTO  \b
1150       CALL   a
1151    
1152       This shows that there was an initial non-tail call of a, then a
1153       tail-call to b or c, looping behavior in a cluster of functions that
1154       consist of b and c, and then a goto from that cluster (i.e., either from
1155       b or from c) to d.
1156    
1157       Note that (depending on the user program) the amount of information
1158       that the back-trace module has to keep track of at each level is bounded
1159       by a constant.  Thus, the whole implementation has the same asymptotical
1160       complexity as the original program (both in space and in time).
1161    
1162    5. compiler/TopLevel/interact/evalloop.sml has been modified to
1163       handle the special exception SMLofNJ.Internals.BTrace.BTrace
1164       which is raised by the "trigger" function mentioned above.
1165    
1166    Notes on usage:
1167    
1168    - Annotated code works well together with unannotated code:
1169    Unannotated calls simply do not show up at all in the backtrace.
1170    
1171    - It is not a good idea to let modules that were annotated during
1172    different sessions run at the same time.  This is because the compiler
1173    chooses small integers to identify individual functions, and there
1174    will be clashes if different modules were compiled in separate sessions.
1175    (Nothing will crash, and you will even be told about the clashes, but
1176    back-trace information will in general not be useful.)
1177    
1178    - Back-tracing can be confused by callcc and capture.
1179    
1180    - The only way of getting a back-trace right now is to explicitly
1181    invoke the "trigger" function from your user program.  Eventually, we
1182    should make every exception carry back-trace information (if
1183    available).  But since this creates more overhead at "raise"-time
1184    (similar to the current exnHistory overhead), I have not yet
1185    implemented this.  (The implementation will be rather easy.)  With
1186    exceptions carrying back-trace information, this facility will be even
1187    more useful because users don't need to modify their programs...
1188    
1189    - While it is possible to compile the compiler with back-trace
1190    annotations turned on (I did it to get some confidence in
1191    correctness), you must make absolutely sure that core.sml and
1192    btimp.sml are compiled WITHOUT annotation!  (core.sml cannot actually
1193    be compiled with annotation because there is no core access yet, but
1194    if you compile btimp.sml with annotation, then the system will go into
1195    an infinite recursion and crash.)
1196    Since CM currently does not know about BTrace, the only way to turn
1197    annotations on and off for different modules of the compiler is to
1198    interrupt CMB.make, change the settings, and re-invoke it.  Of course,
1199    this is awkward and clumsy.
1200    
1201    Sample sessions:
1202    
1203    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
1204    - SMLofNJ.Internals.BTrace.mode (SOME true);
1205    [autoloading]
1206    [autoloading done]
1207    val it = false : bool
1208    - structure X = struct
1209    -     fun main n = let
1210    -         fun a (x, 0) = d x
1211    -           | a (x, n) = b (x, n - 1)
1212    -         and b (x, n) = c (x, n)
1213    -         and c (x, n) = a (x, n)
1214    -         and d x = e (x, 3)
1215    -         and e (x, 0) = f x
1216    -           | e (x, n) = e (x, n - 1)
1217    -         and f 0 = SMLofNJ.Internals.BTrace.trigger ()
1218    -           | f n = n * g (n - 1)
1219    -         and g n = a (n, 3)
1220    -     in
1221    -         f n
1222    -     end
1223    - end;
1224    structure X : sig val main : int -> int end
1225    - X.main 3;
1226    *** BACK-TRACE ***
1227    GOTO   stdIn:4.2-13.20: X.main[2].f
1228    GOTO-( stdIn:4.2-13.20: X.main[2].e
1229    GOTO   stdIn:4.2-13.20: X.main[2].d
1230         / stdIn:4.2-13.20: X.main[2].a
1231         | stdIn:4.2-13.20: X.main[2].b
1232    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1233    CALL   stdIn:4.2-13.20: X.main[2].g
1234    GOTO   stdIn:4.2-13.20: X.main[2].f
1235    GOTO-( stdIn:4.2-13.20: X.main[2].e
1236    GOTO   stdIn:4.2-13.20: X.main[2].d
1237         / stdIn:4.2-13.20: X.main[2].a
1238         | stdIn:4.2-13.20: X.main[2].b
1239    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1240    CALL   stdIn:4.2-13.20: X.main[2].g
1241    GOTO   stdIn:4.2-13.20: X.main[2].f
1242    GOTO-( stdIn:4.2-13.20: X.main[2].e
1243    GOTO   stdIn:4.2-13.20: X.main[2].d
1244         / stdIn:4.2-13.20: X.main[2].a
1245         | stdIn:4.2-13.20: X.main[2].b
1246    GOTO-\ stdIn:4.2-13.20: X.main[2].c
1247    CALL   stdIn:4.2-13.20: X.main[2].g
1248    GOTO   stdIn:4.2-13.20: X.main[2].f
1249    CALL   stdIn:2.15-17.4: X.main[2]
1250    -
1251    
1252    (Note that because of a FLINt bug the above code currently does not
1253    compile without BTrace turned on.)
1254    
1255    Here is another example, using my modified Tiger compiler:
1256    
1257    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
1258    - SMLofNJ.Internals.BTrace.mode (SOME true);
1259    [autoloading]
1260    [autoloading done]
1261    val it = false : bool
1262    - CM.make "sources.cm";
1263    [autoloading]
1264    ...
1265    [autoloading done]
1266    [scanning sources.cm]
1267    [parsing (sources.cm):parse.sml]
1268    [creating directory CM/SKEL ...]
1269    [parsing (sources.cm):tiger.lex.sml]
1270    ...
1271    [wrote CM/sparc-unix/semant.sml]
1272    [compiling (sources.cm):main.sml]
1273    [wrote CM/sparc-unix/main.sml]
1274    [New bindings added.]
1275    val it = true : bool
1276    - Main.compile ("../testcases/merge.tig", "foo.out");
1277    *** BACK-TRACE ***
1278    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
1279    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1280    CALL   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1281    GOTO   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
1282    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1283    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1284    CALL   lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
1285         / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
1286    CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
1287    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
1288    CALL   lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
1289    CALL   main.sml:1.18-118.4: Main.compile[2]
1290    -
1291    
1292    ----------------------------------------------------------------------
1293    Name: Matthias Blumen
1294    Date: 2000/06/21 18:00:00 JST
1295    Tag: blume-20000621-manual
1296    Description:
1297    
1298    CM manual update: Path environments documented.
1299    
1300    ----------------------------------------------------------------------
1301    Name: Matthias Blume
1302    Date: 2000/06/19 13:40:00
1303    Tag: blume-20000619-manual
1304    Description:
1305    
1306    CM manual and system/README update.  This only covers the fact that
1307    there are no more implicit anchors.  (Path environments and the "bind"
1308    option to "cm" have yet to be documented.)
1309    
1310    ----------------------------------------------------------------------
1311    Name: Matthias Blume
1312    Date: 2000/06/19 11:05:00 JST
1313    Tag: blume-20000619-chdir-bugfix
1314    Description:
1315    
1316    Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
1317    
1318    ----------------------------------------------------------------------
1319    Name: Matthias Blume
1320    Date: 2000/06/18 22:00:10 JST
1321    Tag: blume-20000618-implicit-anchors-really-gone
1322    Description:
1323    
1324    I updates the previous HISTORY entry where I forgot to mention that
1325    implicit anchors are no longer with us.
1326    
1327    The current update also gets rid of the (now useless) controller
1328    CM.Control.implicit_anchors.
1329    
1330    ----------------------------------------------------------------------
1331    Name: Matthias Blume
1332    Date: 2000/06/16 17:30:00 JST
1333    Tag: blume-20000616-anchorenv
1334    Description:
1335    
1336    This patch implements the long anticipated (just kidding :) "anchor
1337    environment" mechanism.  In the course of doing this, I also
1338    re-implemented CM's internal "SrcPath" module from scratch.  The new
1339    one should be more robust in certain boundary cases.  In any case, it
1340    is a lot cleaner than its predecessor (IMHO).
1341    
1342    This time, although there is yet another boot file format change, I
1343    kept the unpickler backward-compatible.  As a result, no new bootfiles
1344    are necessary and bootstrapping is straightforward.  (You cannot read
1345    new bootfiles into an old system, but the other way around is no
1346    problem.)
1347    
1348    Visible changes:
1349    
1350    ** 0. Implicit path anchors (without the leading $-symbol) are no
1351    longer recognized at all. This means that such path names are not
1352    illegal either.  For example, the name basis.cm simply refers to a
1353    local file called "basis.cm" (i.e, the name is an ordinary path
1354    relative to .cm-files directory).  Or, to put it differently, only
1355    names that start with $ are anchored paths.
1356    
1357    ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
1358    vanished.
1359    
1360    John (Reppy) had critizised this as soon as I originally proposed and
1361    implemented it, but at that time I did not really deeply believe
1362    him. :) Now I came full-circle because I need the $<singlearc> syntax
1363    in another place where it cannot be seen as an abbreviation for
1364    $/<singlearc>.  To avoid the confusion, $<singlearc> now means what it
1365    seems to mean (i.e., it "expands" into the corresponding anchor
1366    value).
1367    
1368    However, when paths are used as members in CM description files, it
1369    continues to be true that there must be at least another arc after the
1370    anchor.  This is now enforced separately during semantic analysis
1371    (i.e., from a lexical/syntactical point of view, the notation is ok.)
1372    
1373    ** 2. The "cm" class now accepts an option "bind".  The option's value
1374    is a sub-option list of precisely two items -- one labeled "anchor"
1375    and the other one labeled "value".  As you might expect, "anchor" is
1376    used to specify an anchor name to be bound, and "value" specifies what
1377    the anchor is being bound to.
1378    
1379    The value must be a directory name and can be given in either standard
1380    syntax (including the possibility that it is itself an anchored path)
1381    or native syntax.
1382    
1383    Examples:
1384    
1385       foo.cm (bind:(anchor:bar value:$mystuff/bar))
1386       lib.cm (bind:(anchor:a value:"H:\\x\\y\\z"))  (* only works under windows *)
1387    
1388    and so on.
1389    
1390    The meaning of this is that the .cm-file will be processed with an
1391    augmented anchor environment where the given anchor(s) is/are bound to
1392    the given values(s).
1393    
1394    The rationale for having this feature is this: Suppose you are trying
1395    to use two different (already stable) libraries a.cm and b.cm (that
1396    you perhaps didn't write yourself).  Further, suppose each of these
1397    two libraries internally uses its own auxiliary library $aux/lib.cm.
1398    Normally you would now have a problem because the anchor "lib" can not
1399    be bound to more than one value globally.  Therefore, the project that
1400    uses both a.cm and b.cm must locally redirect the anchor to some other
1401    place:
1402    
1403       a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
1404       b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
1405    
1406    This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
1407    /usr/lib/smlnj/b-stuff/aux.cm, respectively.
1408    
1409    Hard-wiring path names is a bit inflexible (and CM will verbosely warn
1410    you when you do so at the time of CM.stabilize).  Therefore, you can
1411    also use an anchored path as the value:
1412    
1413      a.cm (bind:(anchor:lib value:$a-lib))
1414      b.cm (bind:(anchor:lib value:$b-lib))
1415    
1416    Now you can globally configure (using the usual CM.Anchor.anchor or
1417    pathconfig machinery) bindings for "a-lib" and "b-lib".  Since "lib"
1418    itself is always locally bound, setting it globally is no longer
1419    meaningful or necessary (but it does not hurt either).  In fact, "lib"
1420    can still be used as a global anchor for separate purposes.  As a
1421    matter of fact, one can locally define "lib" in terms of a global
1422    "lib":
1423    
1424      a.cm (bind:(anchor:lib value:$lib/a))
1425      b.cm (bind:(anchor:lib value:$lib/b))
1426    
1427    ** 3: The encoding of path names has changed.  This affects the way
1428    path names are shown in CM's progress report and also the internal
1429    protocol encoding used for parallel make.
1430    
1431    The encoding now uses one or more ':'-separated segments.  Each
1432    segments corresponds to a file that has been specified relative to the
1433    file given by its preceding segment.  The first segment is either
1434    relative to the CWD, absolute, or anchored.  Each segment itself is
1435    basically a Unix pathname; all segments but the first are relative.
1436    
1437    Example:
1438    
1439       $foo/bar/baz.cm:a/b/c.sml
1440    
1441    This path denotes the file bar/a/b/c.sml relative to the directory
1442    denoted by anchor "foo".  Notice that the encoding also includes
1443    baz.cm which is the .cm-file that listed a/b/c.sml.  As usual, such
1444    paths are resolved relative to the .cm-files directory, so baz.cm must
1445    be ignored to get the "real" pathname.
1446    
1447    To make this fact more obvious, CM puts the names of such "virtual
1448    arcs" into parentheses when they appear in progress reports. (No
1449    parentheses will appear in the internal protocol encoding.)  Thus,
1450    what you really see is:
1451    
1452      $foo/bar/(baz.cm):a/b/c.sml
1453    
1454    I find this notation to be much more informative than before.
1455    
1456    Another new feature of the encoding is that special characters
1457    including parentheses, colons, (back)slashes, and white space are
1458    written as \ddd (where ddd is the decimal encoding of the character).
1459    
1460    *** The CM manual still needs to be updated.
1461    
1462    ----------------------------------------------------------------------
1463    Name: Allen Leung
1464    Date: 2000/06/15 00:38:00
1465    Tag: leunga-20000615-x86-peephole
1466    
1467    x86 Peephole fix by Fermin.  Affects c-- and moby only.
1468    
1469    ----------------------------------------------------------------------
1470    Name: Matthias Blume
1471    Date: 2000/06/12 11:40:00
1472    Tag: blume-20000612-parmakefix
1473    Description:
1474    
1475    More cleanup after changing the file naming scheme: This time I
1476    repaired the parallel make mechanism for CMB.make which I broke earlier.
1477    
1478    ----------------------------------------------------------------------
1479    Name: Allen Leung
1480    Date: 2000/06/09 01:25:00
1481    Tag: leunga-20000609-various
1482    
1483    None of these things should affect normal SML/NJ operations
1484    
1485    1. Peephole improvements provided by Fermin (c--)
1486    2. New annotation DEFUSE for adding extra dependence (moby)
1487    3. New X86 LOCK instructions (moby)
1488    4. New machine description language for reservation tables (scheduling)
1489    5. Fixes to various optimization/analysis modules (branch chaining, dominator
1490       trees etc.)
1491    6. I've changed the CM files so that they can work with versions
1492       110.0.6, 110.25 and 110.28
1493    
1494    ----------------------------------------------------------------------
1495    Name: Matthias Blume
1496    Date: 2000/06/09 12:40:00
1497    Tag: blume-20000609-log
1498    Description:
1499    
1500    - Removed all(?) remaining RCS Log entries from sources.
1501    
1502    - Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
1503      anchored paths).
1504    
1505  ----------------------------------------------------------------------  ----------------------------------------------------------------------
1506  Name: Matthias Blume  Name: Matthias Blume
1507  Date: 2000/06/07 17:00:00 JST  Date: 2000/06/07 17:00:00 JST
# Line 563  Line 2054 
2054    
2055  3.  Assembly  3.  Assembly
2056    
2057      When generating assemby, resolve the value of client defined constants,      When generating assembly, resolve the value of client defined constants,
2058      instead of generating symbolic values.  This is controlled by the      instead of generating symbolic values.  This is controlled by the
2059      new flag "asm-resolve-constants", which is default to true.      new flag "asm-resolve-constants", which is default to true.
2060    
# Line 586  Line 2077 
2077    
2078  To this end, I arranged that instead of "structure Core" as "structure  To this end, I arranged that instead of "structure Core" as "structure
2079  _Core" is bound in the pervasive environment.  Core access is done via  _Core" is bound in the pervasive environment.  Core access is done via
2080  _Core (which can never be accidentially rebound because _Core is not a  _Core (which can never be accidentally rebound because _Core is not a
2081  legal surface-syntax symbol).  legal surface-syntax symbol).
2082    
2083  The current solution is much cleaner because the core environment is  The current solution is much cleaner because the core environment is
# Line 596  Line 2087 
2087  with dynamic and symbolic parts of the core environment.  with dynamic and symbolic parts of the core environment.
2088    
2089  Remaining hackery (to bind the "magic" symbol _Core) is localized in the  Remaining hackery (to bind the "magic" symbol _Core) is localized in the
2090  compilation mananger's bootstrap compiler (actually: in the "init group"  compilation manager's bootstrap compiler (actually: in the "init group"
2091  handling).  See the comments in src/system/smlnj/init/init.cmi for  handling).  See the comments in src/system/smlnj/init/init.cmi for
2092  more details.  more details.
2093    
# Line 711  Line 2202 
2202     (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
2203     used).     used).
2204    
2205     This was done to accomodate the new "make" and "shell" tools which     This was done to accommodate the new "make" and "shell" tools which
2206     facilitate fairly seemless hookup to portions of code managed using     facilitate fairly seamless hookup to portions of code managed using
2207     Makefiles or Shell scripts.     Makefiles or Shell scripts.
2208    
2209     There are no classes "shared" or "private" anymore.  Instead, the     There are no classes "shared" or "private" anymore.  Instead, the
# Line 724  Line 2215 
2215    
2216     All existing tools are described in the CM manual.     All existing tools are described in the CM manual.
2217    
2218     - Slightly better error handling.  (CM now surpresses many followup     - Slightly better error handling.  (CM now suppresses many followup
2219     error messages that tended to be more annoying than helpful.)     error messages that tended to be more annoying than helpful.)
2220    
2221  2. Major changes to the compiler's static environment data structures.  2. Major changes to the compiler's static environment data structures.
# Line 858  Line 2349 
2349    
2350       I've changed andl to testl in the floating point test sequence       I've changed andl to testl in the floating point test sequence
2351       whenever appropriate.  The Intel optimization guide states that       whenever appropriate.  The Intel optimization guide states that
2352       testl is perferable to andl.       testl is preferable to andl.
2353    
2354  7. RA (x86 only)  7. RA (x86 only)
2355    
# Line 1040  Line 2531 
2531    
2532  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an  1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
2533  additional argument called "template" which is an optional string that  additional argument called "template" which is an optional string that
2534  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
2535  explanation.  explanation.
2536    
2537  2. A special-purpose tool can be "regisitered" by simply dropping the  2. A special-purpose tool can be "registered" by simply dropping the
2538  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same  corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
2539  directory where the .cm file lives that uses this tool.  (The  directory where the .cm file lives that uses this tool.  (The
2540  behavior/misfeature until now was to look for the tool description  behavior/misfeature until now was to look for the tool description
# Line 1087  Line 2578 
2578  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
2579  <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,
2580  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
2581  perhaps got accidentially killed).  In this case, fixpt will quickly  perhaps got accidentally killed).  In this case, fixpt will quickly
2582  move through what exists before continuing where it left off earlier,  move through what exists before continuing where it left off earlier,
2583  and, thus, saves a lot of time.  and, thus, saves a lot of time.
2584    
# Line 1137  Line 2628 
2628    it from that remote directory.    it from that remote directory.
2629    This should simplify installation further:  For machines that have    This should simplify installation further:  For machines that have
2630    access to the internet, just fetch <version>-config.tgz, unpack it,    access to the internet, just fetch <version>-config.tgz, unpack it,
2631    edit config/targets, and go (run config/install.sh).  The scipt will    edit config/targets, and go (run config/install.sh).  The script will
2632    fetch everything else that it might need all by itself.    fetch everything else that it might need all by itself.
2633    
2634    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.653  
changed lines
  Added in v.829

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