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 570, Wed Mar 8 17:30:13 2000 UTC revision 691, Tue Jul 25 07:20:24 2000 UTC
# Line 11  Line 11 
11  Date:  Date:
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    ----------------------------------------------------------------------
15    Name: Matthias Blume
16    Date: 2000/07/25 16:20:00 JST
17    Tag: blume-20000725-makedepend
18    Description:
19    
20    Added a script called ml-makedepend.  This can be used in makefiles
21    for Unix' make in a way very similar to the "makedepend" command for
22    C.
23    
24    The script internally uses function CM.sources.
25    
26    Synopsis:
27    
28        ml-makedepend [-f makefile] cmfile targetname
29    
30    The default for the makefile is "makefile" (or "Makefile" should
31    "makefile" not exist).
32    
33    ml-makedepend adds a cmfile/targetname-specific section to this
34    makefile (after removing the previous version of this section).  The
35    section contains a single dependency specification with targetname on
36    the LHS (targetname is an arbitrary name), and a list of files derived
37    from the cmfile on the RHS.  Some of the files on the RHS are
38    ARCH/OPSYS-specific.  Therefore, ml-makedepend inserts references to
39    "make" variables $(ARCH) and $(OPSYS) in place of the corresponding
40    path names.  The makefile writer is responsible for making sure that
41    these variables have correct at the time "make" is invoked.
42    
43    ----------------------------------------------------------------------
44    Name: Matthias Blume
45    Date: 2000/07/22 23:30:00 JST
46    Tag: blume-20000722-urlupdate
47    Description:
48    
49    Changed BOOT and config/srcarchiveurl to point to BL server:
50    
51        ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.29/
52    
53    ----------------------------------------------------------------------
54    Name: Matthias Blume
55    Date: 2000/07/18 18:00:00 JST
56    Tag: blume-20000718-Version_110_29
57    Description:
58    
59    1. Updated src/compiler/TopLevel/main/version.sml to version 110.29
60    
61    2. Updated config/version to 110.29
62    
63    3. Updated config/srcarchiveurl
64    
65    3. New boot files!
66       ftp://ftp.cs.princeton.edu/pub/people/blume/sml/110.29-autofetch
67    
68    ----------------------------------------------------------------------
69    Name: Matthias Blume
70    Date: 2000/07/11 13:58:00 JST
71    Tag: blume-20000711-doctypo
72    Description:
73    
74    Fixed a few typos in CM manual.
75    
76    ----------------------------------------------------------------------
77    Name: Allen Leung
78    Date: 2000/06/15 00:38:00
79    Tag: leunga-20000704-sparc-x86
80    
81    1. x86 peephole improvement sp += k; sp -= k => nop  [from John]
82    2. fix to x86 RET bug [found by Dan Grossman]
83    3. sparc assembly bug fix for ticc instructions [found by Fermin]
84    
85       Affects c-- and moby only
86    
87    ----------------------------------------------------------------------
88    Name: Matthias Blume
89    Date: 2000/07/04 15:26:00
90    Tag: blume-20000704-trigger
91    Description:
92    
93    1. Improvements to CM manual.
94    2. SMLofNJ.Internals.BTrace.trigger reinstated as an alternative way
95       of getting a back-trace.  The function, when called, raises an
96       internal exception which explicitly carries the full back-trace history,
97       so it is unaffected by any intervening handle-raise pairs ("trivial"
98       or not).  The interactive loop will print that history once it arrives
99       at top level.
100       Short of having all exceptions implicitly carry the full history, the
101       recommended way of using this facility is:
102         - compile your program with instrumentation "on"
103         - run it, when it raises an exception, look at the history
104         - if the history is "cut off" because of some handler, go and modify
105           your program so that it explicitly calls BTrace.trigger
106         - recompile (still instrumented), and rerun; look at the full history
107    
108    ----------------------------------------------------------------------
109    Name: Matthias Blume
110    Date: 2000/07/03 15:36:00 JST
111    Tag: blume-20000702-manual
112    Description:
113    
114    Small corrections and updates to CM manual.
115    
116    ----------------------------------------------------------------------
117    Name: Matthias Blume
118    Date: 2000/06/29 16:04:00 JST
119    Tag: blume-20000629-yacctool
120    Description:
121    
122    Changes:
123    
124    1. Class "mlyacc" now takes separate arguments to pass options to
125       generated .sml- and .sig-files independently.
126    2. Corresponding CM manual updates.
127    3. BTrace module now also reports call sites.  (However, for loop clusters
128       it only shows from where the cluster was entered.)  There are associated
129       modifications to core.sml, internals.{sig,sml}, btrace.sml, and btimp.sml.
130    
131    ----------------------------------------------------------------------
132    Name: Matthias Blume
133    Date: 2000/06/27 16:51:00 JST
134    Tag: blume-20000627-noweb
135    Description:
136    
137    Changes:
138    
139     1. Implemented "subdir" and "witness" options for noweb tool.
140        This caused some slight internal changes in CM's tool implementation.
141     2. Fixed bug in "tool plugin" mechanism.  This is essentially cleaning
142        some remaining issues from earlier path anchor changes.
143     3. Updated CM manual accordingly.
144    
145     4. Changed implementation of back-tracing so that I now consider it
146        ready for prime-time.
147    
148        In particular, you don't have to explicitly trigger the back-trace
149        anymore.  Instead, if you are running BTrace-instrumented code and
150        there is an uncaught exception (regardless of whether or not it was
151        raised in instrumented code), the top-level evalloop will print
152        the back-trace.
153    
154        Features:
155    
156          - Instrumented and uninstrumented code work together seemlessly.
157            (Of course, uninstrumented code is never mentioned in actual
158             back-traces.)
159    
160          - Asymptotic time- and space-complexity of instrumented code is
161            equal to that of uninstrumented code.  (This means that
162            tail-recursion is preserved by the instrumentation phase.)
163    
164          - Modules whose code has been instrumented in different sessions
165            work together without problem.
166    
167          - There is no penalty whatsoever on uninstrumented code.
168    
169          - There is no penalty on "raise" expressions, even in
170            instrumented code.
171    
172        A potential bug (or perhaps it is a feature, too):
173    
174          A back-trace reaches no further than the outermost instrumented
175          non-trivial "raise".  Here, a "trivial" raise is one that is the
176          sole RHS of a "handle" rule.  Thus, back-traces reach trough
177    
178               <exp> handle e => raise e
179    
180          and even
181    
182               <exp> handle Foo => raise Bar
183    
184          and, of course, through
185    
186               <exp> handle Foo => ...
187    
188         if the exception was not Foo.
189    
190         Back-traces always reach right through any un-instrumented code
191         including any of its "handle" expressions, trivial or not.
192    
193       To try this out, do the following:
194    
195         - Erase all existing binfiles for your program.
196           (You may keep binfiles for those modules where you think you
197            definitely don't need back-tracing.)
198         - Turn on back-trace instrumentation:
199              SMLofNJ.Internals.BTrace.mode (SOME true);
200         - Recompile your program.  (I.e., run "CM.make" or "use".)
201         - You may now turn instrumentation off again (if you want):
202              SMLofNJ.Internals.BTrace.mode (SOME false);
203         - Run your program as usual.  If it raises an exception that
204           reaches the interactive toplevel, then a back-trace will
205           automatically be printed.  After that, the toplevel loop
206           will print the exception history as usual.
207    
208    ----------------------------------------------------------------------
209    Name: Matthias Blume
210    Date: 2000/06/26 09:56:46 JST
211    Tag: blume-20000626-setup
212    Description:
213    
214    CM: - setup-parameter to "sml" added; this can be used to run arbitrary
215          ML code before and after compiling a file (e.g., to set compiler
216          flags)
217    
218    Compiler: - improved btrace API (in core.sml, internals.{sig,sml})
219              - associated changes to btrace.sml (BTrace instrumentation pass)
220              - cleaner implementation of btimp.sml (BTrace tracing and report
221                module)
222    
223    CM manual: * new path encoding documented
224               * description of setup-parameter to "sml" added
225    
226    The biggest user-visible change to back-tracing is that it is no
227    longer necessary to compile all traced modules within the same
228    session.  (This was a real limitation.)
229    
230    ----------------------------------------------------------------------
231    Name: Matthias Blume
232    Date: 2000/06/24 12:40:00 JST
233    Tag: blume-20000624-startup
234    Description:
235    
236    Fixes startup slowdown problem.  (I was calling SrcPath.sync a _tad_
237    bit too often -- to put it mildly. :)
238    
239    ----------------------------------------------------------------------
240    Name: Matthias Blume
241    Date: 2000/06/23 18:20:00 JST
242    Tag: blume-20000623-btrace
243    Description:
244    
245    This updates adds a backtrace facility to aid programmers in debugging
246    their programs.  This involves the following changes:
247    
248    1. Module system/smlnj/init/core.sml (structure _Core) now has hooks for
249       keeping track of the current call stack.  When programs are compiled
250       in a special mode, the compiler will insert calls to these hooks
251       into the user program.
252       "Hook" means that it is possible for different implementations of
253       back-tracing to register themselves (at different times).
254    
255    2. compiler/MiscUtil/profile/btrace.sml implements the annotation phase
256       as an Absyn.dec->Absyn.dec rewrite.  Normally this phase is turned off.
257       It can be turned on using this call:
258         SMLofNJ.Internals.BTrace.mode (SOME true);
259       Turning it off again:
260         SMLofNJ.Internals.BTrace.mode (SOME false);
261       Querying the current status:
262         SMLofNJ.Internals.BTrace.mode NONE;
263       Annotated programs are about twice as big as normal ones, and they
264       run a factor of 2 to 4 slower with a dummy back-trace plugin (one
265       where all hooks do nothing).  The slowdown with a plugin that is
266       actually useful (such as the one supplied by default) is even greater,
267       but in the case of the default plugin it is still only an constant
268       factor (amortized).
269    
270    3. system/Basis/Implementation/NJ/internals.{sig,sml} have been augmented
271       with a sub-structure BTrace for controlling back-tracing.  In particular,
272       the above-mentioned function "mode" controls whether the annotation
273       phase is invoked by the compiler.  Another important function is
274       "trigger": when called it aborts the current execution and causes
275       the top-level loop to print a full back-trace.
276    
277    4. compiler/MiscUtil/profile/btimp.sml is the current default plugin
278       for back-tracing.  It keeps track of the dynamic call stack and in
279       addition to that it keeps a partial history at each "level" of that
280       stack.  For example, if a tail-calls b, b tail-calls c, and c tail-calls
281       d and b (at separate times, dynamically), then the report will show:
282    
283       GOTO   d
284             /c
285       GOTO  \b
286       CALL   a
287    
288       This shows that there was an initial non-tail call of a, then a
289       tail-call to b or c, looping behavior in a cluster of functions that
290       consist of b and c, and then a goto from that cluster (i.e., either from
291       b or from c) to d.
292    
293       Note that (depending on the user program) the amount of information
294       that the back-trace module has to keep track of at each level is bounded
295       by a constant.  Thus, the whole implementation has the same asymptotical
296       complexity as the original program (both in space and in time).
297    
298    5. compiler/TopLevel/interact/evalloop.sml has been modified to
299       handle the special exception SMLofNJ.Internals.BTrace.BTrace
300       which is raised by the "trigger" function mentioned above.
301    
302    Notes on usage:
303    
304    - Annotated code works well together with unannotated code:
305    Unannotated calls simply do not show up at all in the backtrace.
306    
307    - It is not a good idea to let modules that were annotated during
308    different sessions run at the same time.  This is because the compiler
309    chooses small integers to identify individual functions, and there
310    will be clashes if different modules were compiled in separate sessions.
311    (Nothing will crash, and you will even be told about the clashes, but
312    back-trace information will in general not be useful.)
313    
314    - Back-tracing can be confused by callcc and capture.
315    
316    - The only way of getting a back-trace right now is to explicitly
317    invoke the "trigger" function from your user program.  Eventually, we
318    should make every exception carry back-trace information (if
319    available).  But since this creates more overhead at "raise"-time
320    (similar to the current exnHistory overhead), I have not yet
321    implemented this.  (The implementation will be rather easy.)  With
322    exceptions carrying back-trace information, this facility will be even
323    more useful because users don't need to modify their programs...
324    
325    - While it is possible to compile the compiler with back-trace
326    annotations turned on (I did it to get some confidence in
327    correctness), you must make absolutely sure that core.sml and
328    btimp.sml are compiled WITHOUT annotation!  (core.sml cannot actually
329    be compiled with annotation because there is no core access yet, but
330    if you compile btimp.sml with annotation, then the system will go into
331    an infinite recursion and crash.)
332    Since CM currently does not know about BTrace, the only way to turn
333    annotations on and off for different modules of the compiler is to
334    interrupt CMB.make, change the settings, and re-invoke it.  Of course,
335    this is awkward and clumsy.
336    
337    Sample sessions:
338    
339    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
340    - SMLofNJ.Internals.BTrace.mode (SOME true);
341    [autoloading]
342    [autoloading done]
343    val it = false : bool
344    - structure X = struct
345    -     fun main n = let
346    -         fun a (x, 0) = d x
347    -           | a (x, n) = b (x, n - 1)
348    -         and b (x, n) = c (x, n)
349    -         and c (x, n) = a (x, n)
350    -         and d x = e (x, 3)
351    -         and e (x, 0) = f x
352    -           | e (x, n) = e (x, n - 1)
353    -         and f 0 = SMLofNJ.Internals.BTrace.trigger ()
354    -           | f n = n * g (n - 1)
355    -         and g n = a (n, 3)
356    -     in
357    -         f n
358    -     end
359    - end;
360    structure X : sig val main : int -> int end
361    - X.main 3;
362    *** BACK-TRACE ***
363    GOTO   stdIn:4.2-13.20: X.main[2].f
364    GOTO-( stdIn:4.2-13.20: X.main[2].e
365    GOTO   stdIn:4.2-13.20: X.main[2].d
366         / stdIn:4.2-13.20: X.main[2].a
367         | stdIn:4.2-13.20: X.main[2].b
368    GOTO-\ stdIn:4.2-13.20: X.main[2].c
369    CALL   stdIn:4.2-13.20: X.main[2].g
370    GOTO   stdIn:4.2-13.20: X.main[2].f
371    GOTO-( stdIn:4.2-13.20: X.main[2].e
372    GOTO   stdIn:4.2-13.20: X.main[2].d
373         / stdIn:4.2-13.20: X.main[2].a
374         | stdIn:4.2-13.20: X.main[2].b
375    GOTO-\ stdIn:4.2-13.20: X.main[2].c
376    CALL   stdIn:4.2-13.20: X.main[2].g
377    GOTO   stdIn:4.2-13.20: X.main[2].f
378    GOTO-( stdIn:4.2-13.20: X.main[2].e
379    GOTO   stdIn:4.2-13.20: X.main[2].d
380         / stdIn:4.2-13.20: X.main[2].a
381         | stdIn:4.2-13.20: X.main[2].b
382    GOTO-\ stdIn:4.2-13.20: X.main[2].c
383    CALL   stdIn:4.2-13.20: X.main[2].g
384    GOTO   stdIn:4.2-13.20: X.main[2].f
385    CALL   stdIn:2.15-17.4: X.main[2]
386    -
387    
388    (Note that because of a FLINt bug the above code currently does not
389    compile without BTrace turned on.)
390    
391    Here is another example, using my modified Tiger compiler:
392    
393    Standard ML of New Jersey v110.28.1 [FLINT v1.5], June 5, 2000
394    - SMLofNJ.Internals.BTrace.mode (SOME true);
395    [autoloading]
396    [autoloading done]
397    val it = false : bool
398    - CM.make "sources.cm";
399    [autoloading]
400    ...
401    [autoloading done]
402    [scanning sources.cm]
403    [parsing (sources.cm):parse.sml]
404    [creating directory CM/SKEL ...]
405    [parsing (sources.cm):tiger.lex.sml]
406    ...
407    [wrote CM/sparc-unix/semant.sml]
408    [compiling (sources.cm):main.sml]
409    [wrote CM/sparc-unix/main.sml]
410    [New bindings added.]
411    val it = true : bool
412    - Main.compile ("../testcases/merge.tig", "foo.out");
413    *** BACK-TRACE ***
414    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trvar
415    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
416    CALL   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
417    GOTO   lib/semant.sml:289.3-295.22: SemantFun[2].transExp.trexp.check[2]
418    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
419    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
420    CALL   lib/semant.sml:488.3-505.6: SemantFun[2].transDec.trdec[2].transBody[2]
421         / lib/semant.sml:411.65-543.8: SemantFun[2].transDec
422    CALL-\ lib/semant.sml:413.2-540.9: SemantFun[2].transDec.trdec[2]
423    CALL   lib/semant.sml:99.2-396.21: SemantFun[2].transExp.trexp
424    CALL   lib/semant.sml:8.52-558.4: SemantFun[2].transProg[2]
425    CALL   main.sml:1.18-118.4: Main.compile[2]
426    -
427    
428    ----------------------------------------------------------------------
429    Name: Matthias Blumen
430    Date: 2000/06/21 18:00:00 JST
431    Tag: blume-20000621-manual
432    Description:
433    
434    CM manual update: Path environments documented.
435    
436    ----------------------------------------------------------------------
437    Name: Matthias Blume
438    Date: 2000/06/19 13:40:00
439    Tag: blume-20000619-manual
440    Description:
441    
442    CM manual and system/README update.  This only covers the fact that
443    there are no more implicit anchors.  (Path environments and the "bind"
444    option to "cm" have yet to be documented.)
445    
446    ----------------------------------------------------------------------
447    Name: Matthias Blume
448    Date: 2000/06/19 11:05:00 JST
449    Tag: blume-20000619-chdir-bugfix
450    Description:
451    
452    Fixed a bug in new SrcPath module that sometimes led to a bad chDir call.
453    
454    ----------------------------------------------------------------------
455    Name: Matthias Blume
456    Date: 2000/06/18 22:00:10 JST
457    Tag: blume-20000618-implicit-anchors-really-gone
458    Description:
459    
460    I updates the previous HISTORY entry where I forgot to mention that
461    implicit anchors are no longer with us.
462    
463    The current update also gets rid of the (now useless) controller
464    CM.Control.implicit_anchors.
465    
466    ----------------------------------------------------------------------
467    Name: Matthias Blume
468    Date: 2000/06/16 17:30:00 JST
469    Tag: blume-20000616-anchorenv
470    Description:
471    
472    This patch implements the long anticipated (just kidding :) "anchor
473    environment" mechanism.  In the course of doing this, I also
474    re-implemented CM's internal "SrcPath" module from scratch.  The new
475    one should be more robust in certain boundary cases.  In any case, it
476    is a lot cleaner than its predecessor (IMHO).
477    
478    This time, although there is yet another boot file format change, I
479    kept the unpickler backward-compatible.  As a result, no new bootfiles
480    are necessary and bootstrapping is straightforward.  (You cannot read
481    new bootfiles into an old system, but the other way around is no
482    problem.)
483    
484    Visible changes:
485    
486    ** 0. Implicit path anchors (without the leading $-symbol) are no
487    longer recognized at all. This means that such path names are not
488    illegal either.  For example, the name basis.cm simply refers to a
489    local file called "basis.cm" (i.e, the name is an ordinary path
490    relative to .cm-files directory).  Or, to put it differently, only
491    names that start with $ are anchored paths.
492    
493    ** 1. The $<singlearc> abbreviation for $/<singlearc> has finally
494    vanished.
495    
496    John (Reppy) had critizised this as soon as I originally proposed and
497    implemented it, but at that time I did not really deeply believe
498    him. :) Now I came full-circle because I need the $<singlearc> syntax
499    in another place where it cannot be seen as an abbreviation for
500    $/<singlearc>.  To avoid the confusion, $<singlearc> now means what it
501    seems to mean (i.e., it "expands" into the corresponding anchor
502    value).
503    
504    However, when paths are used as members in CM description files, it
505    continues to be true that there must be at least another arc after the
506    anchor.  This is now enforced separately during semantic analysis
507    (i.e., from a lexical/syntactical point of view, the notation is ok.)
508    
509    ** 2. The "cm" class now accepts an option "bind".  The option's value
510    is a sub-option list of precisely two items -- one labeled "anchor"
511    and the other one labeled "value".  As you might expect, "anchor" is
512    used to specify an anchor name to be bound, and "value" specifies what
513    the anchor is being bound to.
514    
515    The value must be a directory name and can be given in either standard
516    syntax (including the possibility that it is itself an anchored path)
517    or native syntax.
518    
519    Examples:
520    
521       foo.cm (bind:(anchor:bar value:$mystuff/bar))
522       lib.cm (bind:(anchor:a value:"H:\\x\\y\\z"))  (* only works under windows *)
523    
524    and so on.
525    
526    The meaning of this is that the .cm-file will be processed with an
527    augmented anchor environment where the given anchor(s) is/are bound to
528    the given values(s).
529    
530    The rationale for having this feature is this: Suppose you are trying
531    to use two different (already stable) libraries a.cm and b.cm (that
532    you perhaps didn't write yourself).  Further, suppose each of these
533    two libraries internally uses its own auxiliary library $aux/lib.cm.
534    Normally you would now have a problem because the anchor "lib" can not
535    be bound to more than one value globally.  Therefore, the project that
536    uses both a.cm and b.cm must locally redirect the anchor to some other
537    place:
538    
539       a.cm (bind:(anchor:lib value:/usr/lib/smlnj/a-stuff))
540       b.cm (bind:(anchor:lib value:/usr/lib/smlnj/b-stuff))
541    
542    This hard-wires $lib/aux.cm to /usr/lib/smlnj/a-stuff/aux.cm or
543    /usr/lib/smlnj/b-stuff/aux.cm, respectively.
544    
545    Hard-wiring path names is a bit inflexible (and CM will verbosely warn
546    you when you do so at the time of CM.stabilize).  Therefore, you can
547    also use an anchored path as the value:
548    
549      a.cm (bind:(anchor:lib value:$a-lib))
550      b.cm (bind:(anchor:lib value:$b-lib))
551    
552    Now you can globally configure (using the usual CM.Anchor.anchor or
553    pathconfig machinery) bindings for "a-lib" and "b-lib".  Since "lib"
554    itself is always locally bound, setting it globally is no longer
555    meaningful or necessary (but it does not hurt either).  In fact, "lib"
556    can still be used as a global anchor for separate purposes.  As a
557    matter of fact, one can locally define "lib" in terms of a global
558    "lib":
559    
560      a.cm (bind:(anchor:lib value:$lib/a))
561      b.cm (bind:(anchor:lib value:$lib/b))
562    
563    ** 3: The encoding of path names has changed.  This affects the way
564    path names are shown in CM's progress report and also the internal
565    protocol encoding used for parallel make.
566    
567    The encoding now uses one or more ':'-separated segments.  Each
568    segments corresponds to a file that has been specified relative to the
569    file given by its preceding segment.  The first segment is either
570    relative to the CWD, absolute, or anchored.  Each segment itself is
571    basically a Unix pathname; all segments but the first are relative.
572    
573    Example:
574    
575       $foo/bar/baz.cm:a/b/c.sml
576    
577    This path denotes the file bar/a/b/c.sml relative to the directory
578    denoted by anchor "foo".  Notice that the encoding also includes
579    baz.cm which is the .cm-file that listed a/b/c.sml.  As usual, such
580    paths are resolved relative to the .cm-files directory, so baz.cm must
581    be ignored to get the "real" pathname.
582    
583    To make this fact more obvious, CM puts the names of such "virtual
584    arcs" into parentheses when they appear in progress reports. (No
585    parentheses will appear in the internal protocol encoding.)  Thus,
586    what you really see is:
587    
588      $foo/bar/(baz.cm):a/b/c.sml
589    
590    I find this notation to be much more informative than before.
591    
592    Another new feature of the encoding is that special characters
593    including parentheses, colons, (back)slashes, and white space are
594    written as \ddd (where ddd is the decimal encoding of the character).
595    
596    *** The CM manual still needs to be updated.
597    
598    ----------------------------------------------------------------------
599    Name: Allen Leung
600    Date: 2000/06/15 00:38:00
601    Tag: leunga-20000615-x86-peephole
602    
603    x86 Peephole fix by Fermin.  Affects c-- and moby only.
604    
605    ----------------------------------------------------------------------
606    Name: Matthias Blume
607    Date: 2000/06/12 11:40:00
608    Tag: blume-20000612-parmakefix
609    Description:
610    
611    More cleanup after changing the file naming scheme: This time I
612    repaired the parallel make mechanism for CMB.make which I broke earlier.
613    
614    ----------------------------------------------------------------------
615    Name: Allen Leung
616    Date: 2000/06/09 01:25:00
617    Tag: leunga-20000609-various
618    
619    None of these things should affect normal SML/NJ operations
620    
621    1. Peephole improvements provided by Fermin (c--)
622    2. New annotation DEFUSE for adding extra dependence (moby)
623    3. New X86 LOCK instructions (moby)
624    4. New machine description language for reservation tables (scheduling)
625    5. Fixes to various optimization/analysis modules (branch chaining, dominator
626       trees etc.)
627    6. I've changed the CM files so that they can work with versions
628       110.0.6, 110.25 and 110.28
629    
630    ----------------------------------------------------------------------
631    Name: Matthias Blume
632    Date: 2000/06/09 12:40:00
633    Tag: blume-20000609-log
634    Description:
635    
636    - Removed all(?) remaining RCS Log entries from sources.
637    
638    - Fixed bug in ml-yacc and ml-lex sources (use explicit anchors for
639      anchored paths).
640    
641    ----------------------------------------------------------------------
642    Name: Matthias Blume
643    Date: 2000/06/07 17:00:00 JST
644    Tag: blume-20000607-no-implicit-anchors
645    Description:
646    
647    1. This update changes the default setting for
648    CM.Control.implicit_anchors from true to false.  This means that
649    implicit anchors are no longer permitted by default.  I also tried to
650    make sure that nothing else still relies on implicit anchors.
651    (This is the next step on the schedule towards a CM that does not even
652    have the notion of implicit anchors anymore.)
653    
654    2. More CM manual updates.
655    
656    3. I managed to track down and fix the pickling bug I mentioned last
657    time.  Because of the previously existing workaround, this entails no
658    immediate practical changes.
659    
660    ----------------------------------------------------------------------
661    Name: Matthias Blume
662    Date: 2000/06/06 11:15:00 JST
663    Tag: blume-20000606-lazierpickle
664    Description:
665    
666    !!!! NEW BOOT FILES !!!!
667    
668    * The main purpose of this update is to make library pickles lazier in
669    order to reduce the initial space penalty for autoloading a library.
670    As a result, it is now possible to have $smlnj/compiler.cm
671    pre-registered.  This should take care of the many complaints or
672    inquiries about missing structure Compiler.  This required changes to
673    CM's internal data structures and small tweaks to some algorithms.
674    
675    As a neat additional effect, it is no longer necessary (for the sake
676    of lean heap image files) to distinguish between a "minimal" CM and a
677    "full" CM.  Now, there is only one CM (i.e., the "full" version:
678    $smlnj/cm.cm aka $smlnj/cm/full.cm), and it is always available at the
679    interactive top level. ($smlnj/cm/minimal.cm is gone.)
680    
681    To make the life of compiler-hackers easier, "makeml" now also
682    pre-registers $smlnj/cmb.cm (aka $smlnj/cmb/current.cm).  In other
683    words, after you bootstrap a new sml for the first time, you will not
684    have to autoload $smlnj/cmb.cm again afterwards.  (The first time
685    around you will still have to do it, though.)
686    
687    * A second change consists of major updates to the CM manual.  There
688    are now several appendices with summary information and also a full
689    specification of the CM description file syntax.
690    
691    * In directory src/system I added the script "allcross".  This script
692    invokes sml and cross-compiles the compiler for all supported
693    architectures.  (Useful when providing a new set of boot files.)
694    
695    * There seems to be a latent bug in my "lazy pickles" mechanism.  I
696    added a small tweak to pickle-util.sml to work around this problem,
697    but it is not a proper fix yet.  I will investigate further.  (The
698    effect of the bug was an inflation of library pickle size.)
699    
700    * Version number increased to 110.28.1 (to avoid compatibility problems).
701    
702    ----------------------------------------------------------------------
703    Name: Allen Leung
704    Date: 2000/05/25 17:28 EDT
705    Tag: leunga-20000525-ra
706    Description:
707    
708      Fixed a bug in freezing phase of the register allocator.
709    
710    ----------------------------------------------------------------------
711    Name: Allen Leung
712    Date: 2000/05/15 22:53 EDT
713    Tag: leunga-20000515-alpha-x86-ra
714    Description:
715    
716      1. Alpha
717    
718          Slight cleanup.  Removed the instruction SGNXL
719    
720      2. X86
721    
722          Added the following instructions to the instruction set:
723    
724            ROLx, RORx,
725            BTx, BTSx, BTLx, BTRx,
726            XCHGx, and variants with the LOCK prefix
727    
728      3. Register Allocation
729    
730          The module ra-rewrite-with-renaming has been improved.
731    
732      These have no effect on SML/NJ.
733    
734    ----------------------------------------------------------------------
735    Name: Matthias Blume
736    Date: 2000/05/15 16:20:00 JST
737    Tag: blume-20000515-lightrebuild
738    Description:
739    
740    1. I added an alternative to "-rebuild" to "makeml".  The difference is
741       that prior to calling CMB.make' the CM-variable "LIGHT" will be
742       defined.  In effect, the command will not build any cross-compiler
743       backends and therefore finish more quickly.
744    
745       The "fixpt" script also takes a "-light" switch to be able to use
746       this new facility while compiling for a fixpoint.
747    
748    2. I replaced all mentions of anchored paths in group owner specifications
749       with simple relative paths (usually starting with "..").
750       The rationale is that a library's internal workings should not be
751       compromised by the lack of some anchor.  (An anchor is necessary
752       for someone who wants to refer to the library by an anchored path,
753       but it should not be necessary to build the same library in the first
754       place.)
755    
756    3. I changed the way CM's tool mechanism determines the shell command
757       string used for things like ml-yacc etc. so that it does not break
758       when CM.Control.implicit_anchors is turned off.
759    
760    ----------------------------------------------------------------------
761    Name: Matthias Blume
762    Date: 2000/05/12 18:20:00 JST
763    Tag: blume-20000512-ml-build
764    Description:
765    
766    Fixed a bug in config/_ml-build that prevented ml-yacc and ml-lex from
767    getting installed properly (by config/install.sh).
768    
769    ----------------------------------------------------------------------
770    Name: Matthias Blume
771    Date: 2000/05/12 17:30:00 JST
772    Tag: blume-20000512-anchors
773    Description:
774    
775    !!! NEW BOOT FILES !!!
776    
777    This change is in preparation of fading out support for "implicitly
778    anchored path names".  I went through all sources and used the
779    explicit (and relatively new) $-notation.  See system/README and the
780    CM manual for more info on this.
781    
782    I also modified the anchoring scheme for some things such as "smlnj",
783    "MLRISC", "cm", etc. to take advantage of the fact that explicit
784    anchors are more expressive: anchor name and first arc do not have to
785    coincide.  This entails the following user-visible change:
786    
787    You have to write $smlnj/foo/bar instead of smlnj/foo/bar.  In
788    particular, when you fire up sml with a command-line argument, say,
789    e.g.:
790    
791       sml '$smlnj/cmb.cm'
792    
793    At the ML toplevel prompt:
794    
795       CM.autoload "$smlnj/cmb.cm";
796    
797    There is also a new controller in CM.Control that can be used to turn
798    off all remaining support for implicit anchors by saying:
799    
800        CM.autoload "$smlnj/
801        #set CM.Control.implicit_anchors false;
802    
803    This causes CM to reject implicitly anchored paths.  This is (for the
804    time being) less permissive than the "final" version where there will
805    be no more such implicit anchors and relative paths will be just that:
806    relative.
807    
808    The next step (version after next version?) will be to make the
809    default for CM.Control.implicit_anchors false.  After the dust has
810    settled, I can then produce the "final" version of this...
811    
812    Note: Since bootstrapping is a bit tricky, I provided new boot files.
813    
814    ----------------------------------------------------------------------
815    Name: Matthias Blume
816    Date: 2000/05/11 16:30:00 JST
817    Tag: blume-20000511-sources
818    Description:
819    
820    The main change is that I added function CM.sources as a generalized
821    version of the earlier CM.makedepend.  This entails the following
822    additional changes:
823    
824      - CM.makedepend has been dropped.
825    
826      - CM manual has been updated.
827    
828      - TOOLS signature and API have been changed.
829    
830    ----------------------------------------------------------------------
831    Name: Allen Leung
832    Date: 2000/05/10 21:17 EDT
833    Tag: leunga-20000510-moby-c--ssa
834    Description:
835    
836      Various bug fixes and new features for C--, Moby and MLRISC optimizations.
837    None of these affect SML/NJ.
838    
839    1. Register Allocation
840    
841        a. A new ra spilling module (ra/ra-spill-with-renaming) is implemented.
842           This module tries to remove local (i.e. basic block level) redundancies
843           during spilling.
844    
845        b. A new framework for performing region based register allocation.
846           Not yet entirely functional.
847    
848    2. X86
849    
850       a. DefUse for POP was missing the stack pointer [found by Lal]
851       b. Reload for CALL was incorrect in X86Spill [found by John]
852       c. Various fixes in X86Spill so that it can be used correctly for
853          the new spilling module.
854    
855    3. SSA/IR
856    
857       a. New module ir/dj-dataflow.sml implements elimination based
858          data flow analysis.
859    
860    4. MLRiscGen
861    
862       a. Fix for gc type annotation
863    
864    5. MDGen
865    
866       Various fixes for machine description -> ml code translation.  For ssa
867       only.
868    
869    ----------------------------------------------------------------------
870    Name: Allen Leung
871    Date: 2000/05/08 22:17 EDT
872    Tag: leunga-20000508-labexp
873    Description:
874    
875      Fermin has found a few assembly problems with constant expressions
876      generated in LabelExp.  Mostly, the problems involve extra parentheses,
877      which choke on dumb assemblers.  This is his fix.
878    
879    ----------------------------------------------------------------------
880    Name: Dave MacQueen
881    Date: 2000/04/09 14:00 EDT
882    Tag: dbm-20000502-Version_110_28
883    Description:
884    
885    1. Updated src/compiler/TopLevel/main/version.sml to version 110.28
886    
887    2. Updated config/version to 110.28
888    
889    3. Updated config/srcarchiveurl
890    
891    3. New boot files!
892       ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.28/
893    
894    ----------------------------------------------------------------------
895    Name: Matthias Blume
896    Date: 2000/05/01 19:05:00 JST
897    Tag: blume-20000501-noweb
898    Description:
899    
900    A new noweb tool has been added.  The existing system is entirely
901    unaffected by this, but some CM users have asked for renewed noweb
902    support.  Everything is documented in the CM manual.
903    
904    New (plugin) libraries:
905    
906       noweb-tool.cm
907       nw-ext.cm
908    
909    ----------------------------------------------------------------------
910    Name: Dave MacQueen
911    Date: 2000/04/30 12:40PM EDT
912    Tag: dbm-20000430-bug_fixes
913    Description:
914    
915    1. Fix for bug 1498
916       smlnj/src/system/Basis/Implementation/Unsafe/object.sig
917       smlnj/src/system/Basis/Implementation/Unsafe/object.sml
918         added toRealArray function
919       smlnj/src/compiler/MiscUtil/print/ppobj.sml
920         added check for tag Obj.RealArray to array printing case in ppObj
921    
922    2. Fix for bug 1510
923       smlnj/src/compiler/Semant/types/typesutil.sml
924         fixed definition of dummyargs (used by equalTycon) so that
925         dummy args are distinct types
926    
927    ----------------------------------------------------------------------
928    Name: Matthias Blume
929    Date: 2000/04/30 01:00:00 JST
930    Tag: blume-20000430-versions
931    Description:
932    
933    1. CM version numbering added.  This is an implementation of Lal's
934       proposal for adding version numbers and version checking to .cm
935       files.  Lal said that his proposal was just that -- a proposal.
936       For the time being I went ahead and implemented it so that people
937       can comment on it.  Everything is completely backward-compatible
938       (except for the stable library format, i.e., new bootfiles!).
939    
940       As usual, see the CM manual for details.
941    
942    2. An alternative syntax for anchored paths has been implemented.
943       Dave has recently voiced the same concerns that I had when I did
944       this, so there should be some support.  My take is that eventually
945       I will let support for the current syntax (where anchors are
946       "implicit") fade out in favor of the new, explicit syntax.
947       In order to be backward-compatible, both old and new syntax are
948       currently supported.
949    
950       Again, see the CM manual for details.
951    
952    3. Parallel make is trying to be slightly smarter:  When the master
953       process finds a "bottleneck", i.e., when there is only one
954       compilation unit that can be compiled and everybody else is
955       waiting on it, then it will simply compile it directly instead
956       of clumsily telling one of the slaves to do it.
957    
958    4. Support for "unsharing" added.  This is necessary in order to be
959       able to have two different versions of the same library running
960       at the same time (e.g., for trying out a new MLRISC while still
961       having the old MLRISC linked into the current compiler, etc.)
962       See the CM manual.
963    
964    5. Simple "makedepend" functionality added for generating Makefile
965       dependency information.  (This is rather crude at the moment.
966       Expect some changes here in the future.)
967    
968    6. ".fun" added as a recognized suffix for ML files. Also documented
969       explicitly in the manual that the fallback behavior (unknown suffix
970       -> ML file) is not an official feature!
971    
972    7. Small changes to the pickler for stable libraries.
973    
974    8. Several internal changes to CM (for cleanup/improvement).
975    
976    
977    !!!! NEW BINFILES !!!!
978    
979  ----------------------------------------------------------------------  ----------------------------------------------------------------------
980  Name:  Name: Matthias Blume
981  Date:  Date: 2000/04/28 17:30:00 JST
982  Tag:  Tag: blume-20000428-pathconfig
983    Description:
984    
985    1. I changed config/install.sh to remove duplicate entries from the
986       lib/pathconfig file at the end.  Moreover, the final version of
987       lib/pathconfig is sorted alphabetically.  The same (sorting) is done
988       in src/system/installml.
989    
990    2. The config/install.sh script now consistently uses relative
991       pathnames in lib/pathconfig whenever the anchor is in the lib
992       directory.  (So far this was true for the libraries that come
993       pre-compiled and bundled as part of the bootfiles but not for
994       libraries that are compiled by the script itself.)
995    
996    ----------------------------------------------------------------------
997    Name: Matthias Blume
998    Date: 2000/04/26 13:10:00 JST
999    Tag: blume-20000426-fun_suffix
1000    Description:
1001    
1002    Added ".fun" as a recognized file name suffix (for ML code).
1003    
1004    ----------------------------------------------------------------------
1005    Name: Allen Leung
1006    Date: 2000/04/25 17:00:00 EST
1007    Tag: leunga-20000425-alpha-ra
1008    Description:
1009    
1010    1. Alpha
1011    
1012        PSEUDOARITH was missing in AlphaRewrite.  This causes an endless loop
1013    in C--.
1014    
1015    2. RA
1016    
1017       Added a flag "ra-dump-size" to print out the size of the flowgraph
1018       and the interference graph.
1019    
1020    ----------------------------------------------------------------------
1021    Name: Dave MacQueen
1022    Date: 2000/04/25/
1023    Tag: dbm-20000425-mlyacc_doc_examples
1024    Description:
1025      Updated mlyacc.tex sections 5 and 7 for SML '97 and CM.
1026      Updated all three examples in src/ml-yacc/examples to run
1027      under 110.* using CM.make.
1028    
1029    ----------------------------------------------------------------------
1030    Name: Allen Leung
1031    Date: 2000/04/20 23:04:00 EST
1032    Tag: leunga-20000420-ssa-c---stuff
1033    Description:
1034    
1035      This update synchronizes my repository with Yale's.  Most of these
1036    changes, however, do not affect SML/NJ at all (the RA is an exception).
1037    
1038    1. Register Allocator
1039    
1040       a. An improvement in the interference graph construction:
1041          Given a copy
1042    
1043                s <- t
1044    
1045          no interference edge between s and t is added for this definition of s.
1046    
1047       b. I've added two new spill heuristic modules that Fermin and I developed
1048          (in the new library RA.cm). These are unused in SML/NJ but maybe
1049          useful for others (Moby?)
1050    
1051    2. X86
1052    
1053       a. Various fixes in the backend provided by Fermin [C--] and Lal.
1054    
1055    3. Alpha
1056    
1057       a. Added the BSR instruction and code generation that goes with it [C--]
1058       b. Other fixes too numerous to recount provided by Fermin [C--]
1059    
1060    4. Regmaps
1061    
1062       a. The regmaps are not initialized with the identity physical bindings
1063          at creation time.  This is unneeded.
1064    
1065    5. MLRISC Optimizations
1066    
1067       a. The DJ-Graph module can now compute the iterated dominance frontiers
1068          intersects with liveness incrementally in linear time! Woohoo!
1069          This is now used in my new SSA construction algorithm.
1070    
1071       b. THe branch reorganization module is now smarter about linear chains of
1072          basic blocks.
1073    
1074    
1075    ----------------------------------------------------------------------
1076    Name: Matthias Blume
1077    Date: 2000/04/12 13:52:00 JST
1078    Tag: blume_main_v110p27_1
1079    Description:
1080    
1081    Changed install.sh script to handle archive files without version number
1082    and to use "boot.<arch>-<os>" instead of "sml.boot.<arch>-<os>" for the
1083    name of the boot file archive.
1084    
1085    ----------------------------------------------------------------------
1086    Name: Dave MacQueen
1087    Date: 2000/04/09 14:00 EDT
1088    Tag: dbm-20000410-Version_110_27
1089    Description:
1090    
1091    1. Updated src/compiler/TopLevel/main/version.sml to version 110.27
1092    
1093    2. Updated src/config/version to 110.27
1094    
1095    3. New boot files!
1096    
1097    ----------------------------------------------------------------------
1098    Name: Allen Leung
1099    Date: 2000/04/09 19:09:00 EST
1100    Tag: leunga-20000409-misc
1101    Description:
1102    
1103    1.  Yet another fix for x86 assembly for idivl, imull, mull and friends.
1104    
1105    2.  Miscellaneous improvements to MLRISC (unused in sml/nj)
1106    
1107    ----------------------------------------------------------------------
1108    Name: Stefan
1109    Date: 2000/04/07 10:00:00 EDT
1110    Tag: monnier-20000406-branch-handling
1111    Description:
1112    
1113    Improved handling of branches (mostly those generated from
1114    polymorphic equality), removed switchoff and changed the
1115    default optimization settings (more cpsopt and less flintopt).
1116    
1117    ----------------------------------------------------------------------
1118    Name: Allen Leung
1119    Date: 2000/04/06 01:30:00 EST
1120    Tag: leunga-20000406-peephole-x86-SSA-2
1121    Description:
1122    
1123       Forgot a few files.
1124    
1125    ----------------------------------------------------------------------
1126    Name: Allen Leung
1127    Date: 2000/04/06 00:36:00 EST
1128    Tag: leunga-20000406-peephole-x86-SSA
1129    Description:
1130    
1131    1.  New Peephole code
1132    
1133    2.  Minor improvement to X86 instruction selection
1134    
1135    3.  Various fixes to SSA and machine description -> code translator
1136    
1137    ----------------------------------------------------------------------
1138    Name: Matthias Blume
1139    Date: 2000/04/05 12:30:00 JST
1140    Tag: blume_main_v110p26p2_3
1141    Description:
1142    
1143    This update just merges three minor cosmetic updates to CM's sources
1144    to get ready for the 110.27 code freeze on Friday.  No functionality
1145    has changed.
1146    
1147    ----------------------------------------------------------------------
1148    Name: Allen Leung
1149    Date: 2000/04/04 19:39:00 EST
1150    Tag: leunga-20000404-x86-asm
1151    Description:
1152    
1153    1.  Fixed a problem in X86 assembly.
1154    
1155        Things like
1156    
1157           jmp %eax
1158           jmp (%eax)
1159    
1160        should be output as
1161    
1162           jmp *%eax
1163           jmp *(%eax)
1164    
1165    2.  Assembly output
1166    
1167          Added a new flag
1168    
1169              "asm-indent-copies" (default to false)
1170    
1171          When this flag is on, parallel copies will be indented an extra level.
1172    
1173    ----------------------------------------------------------------------
1174    Name: Allen Leung
1175    Date: 2000/04/04 03:18:00 EST
1176    Tag: leunga-20000404-C--Moby
1177    Description:
1178    
1179        All of these fixes are related to C--, Moby, and my own optimization
1180        stuff; so they shouldn't affect SML/NJ.
1181    
1182    1.  X86
1183    
1184        Various fixes related floating point, and extensions.
1185    
1186    2.  Alpha
1187    
1188        Some extra patterns related to loads with signed/zero extension
1189        provided by Fermin.
1190    
1191    3.  Assembly
1192    
1193        When generating assembly, resolve the value of client defined constants,
1194        instead of generating symbolic values.  This is controlled by the
1195        new flag "asm-resolve-constants", which is default to true.
1196    
1197    4.  Machine Descriptions
1198    
1199        a. The precedence parser was slightly broken when parsing infixr symbols.
1200        b. The type generalizing code had the bound variables reversed, resulting
1201           in a problem during arity raising.
1202        c. Various fixes in machine descriptions.
1203    
1204    ----------------------------------------------------------------------
1205    Name: Matthias Blume
1206    Date: 2000/04/03 16:05:00 JST
1207    Tag: blume_main_v110p26p2_2
1208    Description:
1209    
1210    I eliminated coreEnv from compInfo.  Access to the "Core" structure is
1211    now done via the ordinary static environment that is context to each
1212    compilation unit.
1213    
1214    To this end, I arranged that instead of "structure Core" as "structure
1215    _Core" is bound in the pervasive environment.  Core access is done via
1216    _Core (which can never be accidentally rebound because _Core is not a
1217    legal surface-syntax symbol).
1218    
1219    The current solution is much cleaner because the core environment is
1220    now simply part of the pervasive environment which is part of every
1221    compilation unit's context anyway.  In particular, this eliminates all
1222    special-case handling that was necessary until now in order to deal
1223    with dynamic and symbolic parts of the core environment.
1224    
1225    Remaining hackery (to bind the "magic" symbol _Core) is localized in the
1226    compilation manager's bootstrap compiler (actually: in the "init group"
1227    handling).  See the comments in src/system/smlnj/init/init.cmi for
1228    more details.
1229    
1230    I also tried to track down all mentions of "Core" (as string argument
1231    to Symbol.strSymbol) in the compiler and replaced them with a
1232    reference to the new CoreSym.coreSym.  Seems cleaner since the actual
1233    name appears in one place only.
1234    
1235    Binfile and bootfile format have not changed, but the switchover from
1236    the old "init.cmi" to the new one is a bit tricky, so I supplied new
1237    bootfiles anyway.
1238    
1239    ----------------------------------------------------------------------
1240    Name: Allen Leung
1241    Date: 2000/04/02 21:17:00 EST
1242    Tag: leunga-20000402-mltree
1243    Description:
1244    
1245       1. Renamed the constructor CALL in MLTREE by popular demand.
1246       2. Added a bunch of files from my repository.  These are currently
1247          used by other non-SMLNJ backends.
1248    
1249    ----------------------------------------------------------------------
1250    Name: Allen Leung
1251    Date: 2000/03/31 21:15:00 EST
1252    Tag: leunga-20000331-aliasing
1253    Description:
1254    
1255    This update contains a rewritten (and hopefully more correct) module
1256    for extracting aliasing information from CPS.
1257    
1258       To turn on this feature:
1259    
1260            Compiler.Control.CG.memDisambiguate := true
1261    
1262       To pretty print the region information with assembly
1263    
1264           Compiler.Control.MLRISC.getFlag "asm-show-region" := true;
1265    
1266       To control how many levels of aliasing information are printed, use:
1267    
1268           Compiler.Control.MLRISC.getInt "points-to-show-level" := n
1269    
1270       The default of n is 3.
1271    
1272    ----------------------------------------------------------------------
1273    Name: David MacQueen
1274    Date: 2000/03/31 11:15:00 EST
1275    Tag: dbm-20000331-runtime_fix
1276    Description:
1277    
1278    This update contains:
1279    
1280    1. runtime/c-lib/c-libraries.c
1281       includes added in revision 1.2 caused compilation errors on hppa-hpux
1282    
1283    2. fix for bug 1556
1284       system/Basis/Implementation/NJ/internal-signals.sml
1285    
1286    ----------------------------------------------------------------------
1287    Name: Matthias Blume
1288    Date: 2000/03/31 18:00:00 JST
1289    Tag: blume_main_v110p26p2_1
1290    Description:
1291    
1292    This update contains:
1293    
1294    1. A small change to CM's handling of stable libraries:
1295       CM now maintains one "global" modmap that is used for all stable
1296       libraries.  The use of such a global modmap maximizes sharing and
1297       minimizes the need for re-traversing parts of environments during
1298       modmap construction.  (However, this has minor impact since modmap
1299       construction seems to account for just one percent or less of total
1300       compile time.)
1301    
1302    2. I added a "genmap" phase to the statistics.  This is where I got the
1303       "one percent" number (see above).
1304    
1305    3. CM's new tool parameter mechanism just became _even_ better. :)
1306       - The parser understands named parameters and recursive options.
1307       - The "make" and "shell" tools use these new features.
1308         (This makes it a lot easier to cascade these tools.)
1309       - There is a small syntax change: named parameters use a
1310    
1311           <name> : ( <option> ... )            or
1312           <name> : <string>
1313    
1314         syntax.  Previously, named parameters were implemented in an
1315         ad-hoc fashion by each tool individually (by parsing strings)
1316         and had the form
1317    
1318           <name>=<string>
1319    
1320       See the CM manual for a full description of these issues.
1321    
1322    ----------------------------------------------------------------------
1323    Name: Matthias Blume
1324    Date: 2000/03/30 18:00:00 JST
1325    Tag: blume_main_v110p26p2_0
1326    Description:
1327    
1328    !!!!! WARNING !!!!!!
1329    !!  New binfiles  !!
1330    !!!!!!!!!!!!!!!!!!!!
1331    
1332    This update contains:
1333    
1334    1. Moderate changes to CM:
1335    
1336       - Changes to CM's tools mechanism.  In particular, it is now possible
1337       to have tools that accept additional "command line" parameters
1338       (specified in the .cm file at each instance where the tool's class is
1339       used).
1340    
1341       This was done to accommodate the new "make" and "shell" tools which
1342       facilitate fairly seamless hookup to portions of code managed using
1343       Makefiles or Shell scripts.
1344    
1345       There are no classes "shared" or "private" anymore.  Instead, the
1346       sharing annotation is now a parameter to the "sml" class.
1347    
1348       There is a bit of generic machinery for implementing one's own
1349       tools that accept command-line parameters.  However, I am not yet fully
1350       satisfied with that part, so expect changes here in the future.
1351    
1352       All existing tools are described in the CM manual.
1353    
1354       - Slightly better error handling.  (CM now suppresses many followup
1355       error messages that tended to be more annoying than helpful.)
1356    
1357    2. Major changes to the compiler's static environment data structures.
1358    
1359       - no CMStaticEnv anymore.
1360            - no CMEnv, no "BareEnvironment" (actually, _only_ BareEnvironment,
1361              but it is called Environment), no conversions between different
1362              kinds of static environments
1363    
1364       - There is still a notion of a "modmap", but such modmaps are generated
1365         on demand at the time when they are needed.  This sounds slow, but I
1366         sped up the code that generates modmaps enough for this not to lead to
1367         a slowdown of the compiler (at least I didn't detect any).
1368    
1369       - To facilitate rapid modmap generation, static environments now
1370         contain an (optional) "modtree" structure.  Modtree annotations are
1371         constructed by the unpickler during unpickling.  (This means that
1372         the elaborator does not have to worry about modtrees at all.)
1373         Modtrees have the advantage that they are compositional in the same
1374         way as the environment data structure itself is compositional.
1375         As a result, modtrees never hang on to parts of an environment that
1376         has already been rendered "stale" by filtering or rebinding.
1377    
1378       - I went through many, many trials and errors before arriving at the
1379         current solution.  (The initial idea of "linkpaths" did not work.)
1380         But the result of all this is that I have touched a lot of files that
1381         depend on the "modules" and "types" data structures (most of the
1382         elaborator). There were a lot of changes during my "linkpath" trials
1383         that could have been reverted to their original state but weren't.
1384         Please, don't be too harsh on me for messing with this code a bit more
1385         than what was strictly necessary...  (I _did_ resist the tempation
1386         of doing any "global reformatting" to avoid an untimely death at
1387         Dave's hands. :)
1388    
1389       - One positive aspect of the previous point:  At least I made sure that
1390         all files that I touched now compile without warnings (other than
1391         "polyEqual").
1392    
1393       - compiler now tends to run "leaner" (i.e., ties up less memory in
1394         redundant modmaps)
1395    
1396    ----------------------------------------------------------------------
1397    Name: Allen Leung
1398    Date: 2000/03/29 18:00:00
1399    Tag: leunga-20000327-mlriscGen_hppa_alpha_x86
1400    Boot files (optional): ftp://react-ilp.cs.nyu.edu/leunga/110.26.1-sml.boot.x86-unix-20000330.tar.gz
1401    Description:
1402    
1403       This update contains *MAJOR* changes to the way code is generated from CPS
1404    in the module mlriscGen, and in various backend modules.
1405    
1406    CHANGES
1407    =======
1408    
1409    1. MLRiscGen: forward propagation fix.
1410    
1411       There was a bug in forward propagation introduced at about the same time
1412       as the MLRISC x86 backend, which prohibits coalescing to be
1413       performed effectively in loops.
1414    
1415       Effect: speed up of loops in RISC architectures.
1416               By itself, this actually slowed down certain benchmarks on the x86.
1417    
1418    2. MLRiscGen:  forward propagating addresses from consing.
1419    
1420       I've changed the way consing code is generated.  Basically I separated
1421       out the initialization part:
1422    
1423            store tag,   offset(allocptr)
1424            store elem1, offset+4(allocptr)
1425            store elem2, offset+8(allocptr)
1426            ...
1427            store elemn, offset+4n(allocptr)
1428    
1429       and the address computation part:
1430    
1431            celladdr <- offset+4+alloctpr
1432    
1433       and move the address computation part
1434    
1435       Effect:  register pressure is generally lower as a result.  This
1436                makes compilation of certain expressions much faster, such as
1437                long lists with non-trivial elements.
1438    
1439                 [(0,0), (0,0), .... (0,0)]
1440    
1441    3. MLRiscGen: base pointer elimination.
1442    
1443        As part of the linkage mechanism, we generate the sequence:
1444    
1445         L:  ...  <- start of the code fragment
1446    
1447         L1:
1448             base pointer <- linkreg - L1 + L
1449    
1450         The base pointer was then used for computing relocatable addresses
1451       in the code fragment.  Frequently (such as in lots of continuations)
1452       this is not needed.  We now eliminate this sequence whenever possible.
1453    
1454         For compile time efficiency, I'm using a very stupid local heuristic.
1455       But in general, this should be done as a control flow analysis.
1456    
1457       Effect:  Smaller code size.  Speed up of most programs.
1458    
1459    4. Hppa back end
1460    
1461         Long jumps in span dependence resolution used to depend on the existence
1462      of the base pointer.
1463    
1464         A jump to a long label L was expanded into the following sequence:
1465    
1466          LDIL %hi(L-8192), %r29
1467          LDO  %lo(L-8192)(%r29), %r29
1468          ADD  %r29, baseptr, %r29
1469          BV,n %r0(%r29)
1470    
1471         In the presence of change (3) above, this will not work.  I've changed
1472       it so that the following sequence of instructions are generated, which
1473       doesn't mention the base pointer at all:
1474    
1475             BL,n  L', %r29           /* branch and link, L' + 4 -> %r29 */
1476        L':  ADDIL L-(L'+4), %r29     /* Compute address of L */
1477             BV,n  %r0(%r29)          /* Jump */
1478    
1479    5. Alpha back end
1480    
1481          New alpha instructions LDB/LDW have been added, as per Fermin's
1482       suggestions.   This is unrelated to all other changes.
1483    
1484    6. X86 back end
1485    
1486         I've changed andl to testl in the floating point test sequence
1487         whenever appropriate.  The Intel optimization guide states that
1488         testl is preferable to andl.
1489    
1490    7. RA (x86 only)
1491    
1492         I've improved the spill propagation algorithm, using an approximation
1493       of maximal weighted independent sets.   This seems to be necessary to
1494       alleviate the negative effect in light of the slow down in (1).
1495    
1496         I'll write down the algorithm one of these days.
1497    
1498    8. MLRiscGen: frequencies
1499    
1500         I've added an annotation that states that all call gc blocks have zero
1501       execution frequencies.  This improves register allocation on the x86.
1502    
1503    BENCHMARKS
1504    ==========
1505    
1506       I've only perform the comparison on 110.25.
1507    
1508       The platforms are:
1509    
1510        HPPA  A four processor HP machine (E9000) with 5G of memory.
1511        X86   A 300Hhz Pentium II with 128M of memory, and
1512        SPARC An Ultra sparc 2 with 512M of memory.
1513    
1514       I used the following parameters for the SML benchmarks:
1515    
1516                 @SMLalloc
1517         HPPA    256k
1518         SPARC   512k
1519         X86     256k
1520    
1521    COMPILATION TIME
1522    ----------------
1523       Here are the numbers comparing the compilation times of the compilers.
1524       I've only compared 110.25 compiling the new sources versus
1525       a fixpoint version of the new compiler compiling the same.
1526    
1527                     110.25                                  New
1528               Total  Time in RA  Spill+Reload   Total  Time In RA Spill+Reload
1529         HPPA   627s    116s        2684+3584     599s    95s       1003+1879
1530         SPARC  892s    173s        2891+3870     708s    116s      1004+1880
1531         X86    999s    315s       94006+130691   987s    296s    108877+141957
1532    
1533                   110.25         New
1534                Code Size      Code Size
1535         HPPA   8596736         8561421
1536         SPARC  8974299         8785143
1537         X86    9029180         8716783
1538    
1539       So in summary, things are at least as good as before.   Dramatic
1540       reduction in compilation is obtained on the Sparc; I can't explain it,
1541       but it is reproducible.  Perhaps someone should try to reproduce this
1542       on their own machines.
1543    
1544    SML BENCHMARKS
1545    --------------
1546    
1547        On the average, all benchmarks perform at least as well as before.
1548    
1549          HPPA         Compilation Time     Spill+Reload      Run Time
1550                     110.25  New            110.25    New   110.25  New
1551    
1552          barnesHut  3.158  3.015  4.75%    1+1       0+0   2.980  2.922   2.00%
1553              boyer  6.152  5.708  7.77%    0+0       0+0   0.218  0.213   2.34%
1554       count-graphs  1.168  1.120  4.32%    0+0       0+0  22.705 23.073  -1.60%
1555                fft  0.877  0.792 10.74%    1+3       1+3   0.602  0.587   2.56%
1556        knuthBendix  3.180  2.857 11.32%    0+0       0+0   0.675  0.662   2.02%
1557             lexgen  6.190  5.290 17.01%    0+0       0+0   0.913  0.788  15.86%
1558               life  0.803  0.703 14.22%   25+25      0+0   0.153  0.140   9.52%
1559              logic  2.048  2.007  2.08%    6+6       1+1   4.133  4.008   3.12%
1560         mandelbrot  0.077  0.080 -4.17%    0+0       0+0   0.765  0.712   7.49%
1561             mlyacc 22.932 20.937  9.53%  154+181    32+57  0.468  0.430   8.91%
1562            nucleic  5.183  5.060  2.44%    2+2       0+0   0.125  0.120   4.17%
1563      ratio-regions  3.357  3.142  6.84%    0+0       0+0  116.225 113.173 2.70%
1564                ray  1.283  1.290 -0.52%    0+0       0+0   2.887  2.855   1.11%
1565             simple  6.307  6.032  4.56%   28+30      5+7   3.705  3.658   1.28%
1566                tsp  0.888  0.862  3.09%    0+0       0+0   7.040  6.893   2.13%
1567               vliw 24.378 23.455  3.94%  106+127    25+45  2.758  2.707   1.91%
1568      --------------------------------------------------------------------------
1569       Average                     6.12%                                   4.09%
1570    
1571          SPARC        Compilation Time     Spill+Reload      Run Time
1572                     110.25  New            110.25    New   110.25  New
1573    
1574          barnesHut  3.778  3.592  5.20%    2+2       0+0   3.648  3.453    5.65%
1575              boyer  6.632  6.110  8.54%    0+0       0+0   0.258  0.242    6.90%
1576       count-graphs  1.435  1.325  8.30%    0+0       0+0  33.672 34.737   -3.07%
1577                fft  0.980  0.940  4.26%    3+9       2+6   0.838  0.827    1.41%
1578        knuthBendix  3.590  3.138 14.39%    0+0       0+0   0.962  0.967   -0.52%
1579             lexgen  6.593  6.072  8.59%    1+1       0+0   1.077  1.078   -0.15%
1580               life  0.972  0.868 11.90%   26+26      0+0   0.143  0.140    2.38%
1581              logic  2.525  2.387  5.80%    7+7       1+1   5.625  5.158    9.05%
1582         mandelbrot  0.090  0.093 -3.57%    0+0       0+0   0.855  0.728   17.39%
1583             mlyacc 26.732 23.827 12.19%  162+189    32+57  0.550  0.560   -1.79%
1584            nucleic  6.233  6.197  0.59%    3+3       0+0   0.163  0.173   -5.77%
1585      ratio-regions  3.780  3.507  7.79%    0+0       0+0 133.993 131.035   2.26%
1586                ray  1.595  1.550  2.90%    1+1       0+0   3.440  3.418    0.63%
1587             simple  6.972  6.487  7.48%   29+32      5+7   3.523  3.525   -0.05%
1588                tsp  1.115  1.063  4.86%    0+0       0+0   7.393  7.265    1.77%
1589               vliw 27.765 24.818 11.87%  110+135    25+45  2.265  2.135    6.09%
1590      ----------------------------------------------------------------------------
1591       Average                     6.94%                                    2.64%
1592    
1593          X86          Compilation Time     Spill+Reload      Run Time
1594                     110.25  New            110.25    New   110.25  New
1595    
1596          barnesHut  5.530  5.420  2.03%  593+893   597+915   3.532  3.440   2.66%
1597              boyer  8.768  7.747 13.19%  493+199   301+289   0.327  0.297  10.11%
1598       count-graphs  2.040  2.010  1.49%  298+394   315+457  26.578 28.660  -7.26%
1599                fft  1.327  1.302  1.92%  112+209   115+210   1.055  0.962   9.71%
1600        knuthBendix  5.218  5.475 -4.69%  451+598   510+650   0.928  0.932  -0.36%
1601             lexgen  9.970  9.623  3.60% 1014+841  1157+885   0.947  0.928   1.97%
1602               life  1.183  1.183  0.00%  162+182   145+148   0.127  0.103  22.58%
1603              logic  3.285  3.512 -6.45%  514+684   591+836   5.682  5.577   1.88%
1604         mandelbrot  0.147  0.143  2.33%   38+41     33+54    0.703  0.690   1.93%
1605             mlyacc 35.457 32.763  8.22% 3496+4564 3611+4860  0.552  0.550   0.30%
1606            nucleic  7.100  6.888  3.07%  239+168   201+158   0.175  0.173   0.96%
1607      ratio-regions  6.388  6.843 -6.65% 1182+257   981+300  120.142 120.345 -0.17%
1608                ray  2.332  2.338 -0.29%  346+398   402+494   3.593  3.540   1.51%
1609             simple  9.912  9.903  0.08% 1475+941  1579+1168  3.057  3.178  -3.83%
1610                tsp  1.623  1.532  5.98%  266+200   250+211   8.045  7.878   2.12%
1611               vliw 33.947 35.470 -4.29% 2629+2774 2877+3171  2.072  1.890   9.61%
1612      ----------------------------------------------------------------------------
1613       Average                     1.22%                                     3.36%
1614    
1615    ----------------------------------------------------------------------
1616    Name: Allen Leung
1617    Date: 2000/03/23 16:25:00
1618    Tag: leunga-20000323-fix_x86_alpha
1619    Description:
1620    
1621    1. X86 fixes/changes
1622    
1623       a.  The old code generated for SETcc was completely wrong.
1624           The Intel optimization guide is VERY misleading.
1625    
1626    2. ALPHA fixes/changes
1627    
1628       a.  Added the instructions LDBU, LDWU, STB, STW as per Fermin's suggestion.
1629       b.  Added a new mode byteWordLoadStores to the functor parameter to Alpha()
1630       c.  Added reassociation code for address computation.
1631    
1632    ----------------------------------------------------------------------
1633    Name: Allen Leung
1634    Date: 2000/03/22 01:23:00
1635    Tag: leunga-20000322-fix_x86_hppa_ra
1636    Description:
1637    
1638    1. X86 fixes/changes
1639    
1640       a.  x86Rewrite bug with MUL3 (found by Lal)
1641       b.  Added the instructions FSTS, FSTL
1642    
1643    2. PA-RISC fixes/changes
1644    
1645       a.  B label should not be a delay slot candidate!  Why did this work?
1646       b.  ADDT(32, REG(32, r), LI n) now generates one instruction instead of two,
1647           as it should be.
1648       c.  The assembly syntax for fstds and fstdd was wrong.
1649       d.  Added the composite instruction COMICLR/LDO, which is the immediate
1650           operand variant of COMCLR/LDO.
1651    
1652    3. Generic MLRISC
1653    
1654       a.  shuffle.sml rewritten to be slightly more efficient
1655       b.  DIV bug in mltree-simplify fixed (found by Fermin)
1656    
1657    4. Register Allocator
1658    
1659       a.  I now release the interference graph earlier during spilling.
1660           May improve memory usage.
1661    
1662    ----------------------------------------------------------------------
1663    Name: Matthias Blume
1664    Date: 2000/03/14 14:15:32
1665    Tag: blume_main_v110p26p1_2
1666    Description:
1667    
1668    1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an
1669    additional argument called "template" which is an optional string that
1670    specifies the layout of the tool command line.  See the CM manual for
1671    explanation.
1672    
1673    2. A special-purpose tool can be "registered" by simply dropping the
1674    corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same
1675    directory where the .cm file lives that uses this tool.  (The
1676    behavior/misfeature until now was to look for the tool description
1677    files in the current working directory.)  As before, tool description
1678    files could also be anchored -- in which case they can live anywhere
1679    they like.  Following the recent e-mail discussion, this change should
1680    make it easier to have special-purpose tools that are shipped together
1681    with the sources of the program that uses them.
1682    
1683    ----------------------------------------------------------------------
1684    Name: Matthias Blume
1685    Date: 2000/03/10 07:48:34
1686    Tag: blume_main_v110p26p1_1
1687    Description:
1688    
1689    I added a re-written version of Dave's fixpt script to src/system.
1690    Changes relative to the original version:
1691      - sh-ified (not everybody has ksh)
1692      - automatically figures out which architecture it runs on
1693      - uses ./makeml a bit more cleverly
1694      - never invokes ./installml (and, thus, does not clobber your
1695        good and working installation of sml in case something goes wrong)
1696      - accepts max iteration count using option "-iter <n>"
1697      - accepts a "base" name using option "-base <base>"
1698    
1699    It does not build any extraneous heap images but directly rebuilds
1700    bin- and boot-hierarchies using makeml's "-rebuild" switch. Finally,
1701    it can incorporate existing bin- and boot- hierarchies.  For example,
1702    suppose the base is set to "sml" (which is the default).  Then it
1703    successively builds
1704    
1705            sml.bin.<arch>-unix and sml.boot.<arch>-unix
1706    then    sml1.bin.<arch>-unix and sml1.boot.<arch>-unix
1707    then    sml2.bin.<arch>-unix and sml2.boot.<arch>-unix
1708    ...
1709    then    sml<n>.bin.<arch>-unix and sml<n>.boot.<arch>-unix
1710    
1711    and so on.  If any of these already exist, it will just use what's
1712    there.  In particular, many people will have the initial set of bin
1713    and boot files around, so this saves time for at least one full
1714    rebuild.  Having sets of the form <base><k>.{bin,boot}.<arch>-unix for
1715    <k>=1,2,... is normally not a good idea when invoking fixpt.  However,
1716    they might be the result of an earlier partial run of fixpt (which
1717    perhaps got accidentally killed).  In this case, fixpt will quickly
1718    move through what exists before continuing where it left off earlier,
1719    and, thus, saves a lot of time.
1720    
1721    ----------------------------------------------------------------------
1722    Name: Allen Leung
1723    Date: 00/03/10 02:20:00
1724    Tag: leunga-20000310-fix_x86_asm_ra
1725    Description:
1726    
1727    More assembly output problems involving the indexed addressing mode
1728    on the x86 have been found and corrected. Thanks to Fermin Reig for the
1729    fix.
1730    
1731    The interface and implementation of the register allocator have been changed
1732    slightly to accommodate the possibility to skip the register allocation
1733    phases completely and go directly to memory allocation.  This is needed
1734    for C-- use.
1735    
1736    ----------------------------------------------------------------------
1737    Name: Matthias Blume
1738    Date: 00/03/09 10:23:53
1739    Tag: blume_main_v110p26p1_0
1740    Description:
1741    
1742    * Complete re-organization of library names.  Many libraries have been
1743    consolidated so that they share the same path anchor.  For example,
1744    all MLRISC-related libraries are anchored at MLRISC, most libraries that
1745    are SML/NJ-specific are under "smlnj".  Notice that names like
1746    host-cmb.cm or host-compiler.cm no longer exist.  See system/README
1747    for a complete description of the new naming scheme.  Quick reference:
1748    
1749       host-cmb.cm        -> smlnj/cmb.cm
1750       host-compiler.cm   -> smlnj/compiler.cm
1751       full-cm.cm         -> smlnj/cm.cm
1752       <arch>-<os>.cm     -> smlnj/cmb/<arch>-<os>.cm
1753       <arch>-compiler.cm -> smlnj/compiler/<arch>.cm
1754    
1755    * Bug fixes in CM.
1756        - exceptions in user code are being passed through (i.e., reach top level)
1757        - more bugs in paranoia mode fixed
1758        - bug related to checking group owners fixed
1759    
1760    * New install.sh script that automagically fetches archive files:
1761      The new file config/srcarchiveurl must contain the URL of the
1762      (remote) directory that contains bin files (or other source archives).
1763      If install.sh does not find the archive locally, it tries to get
1764      it from that remote directory.
1765      This should simplify installation further:  For machines that have
1766      access to the internet, just fetch <version>-config.tgz, unpack it,
1767      edit config/targets, and go (run config/install.sh).  The script will
1768      fetch everything else that it might need all by itself.
1769    
1770      For CVS users, this mechanism is not relevant for source archives, but
1771      it is convenient for getting new sets of binfiles.
1772    
1773      Archives should be tar files compressed with either gzip, compress, or
1774      bzip2.  The script recognizes .tgz, .tar, tar.gz, tz, .tar.Z, and .tar.bz2.
1775    
1776    ----------------------------------------------------------------------
1777    Name: Matthias Blume
1778    Date: 2000/03/07 04:01:04
1779    Tag: blume_main_v110_26_2
1780  Description:  Description:
1781    - size info in BOOTLIST
1782         * no fixed upper limits for number of bootfiles or length of
1783           bootfile names in runtime
1784         * falling back to old behavior if no BOOTLIST size info found
1785    - allocation size heuristics in .run-sml
1786         * tries to read cache size from /proc/cpuinfo (this is important for
1787            small-cache Celeron systems!)
1788    - install.sh robustified
1789    - CM manual updates
1790    - paranoid mode
1791         * no more CMB.deliver() (i.e., all done by CMB.make())
1792         * can re-use existing sml.boot.* files
1793         * init.cmi now treated as library
1794         * library stamps for consistency checks
1795    - sml.boot.<arch>-<os>/PIDMAP file
1796         * This file is read by the CM startup code.  This is used to minimize
1797           the amount of dynamic state that needs to be stowed away for the
1798           purpose of sharing between interactive system and user code.
1799    - CM.Anchor.anchor instead of CM.Anchor.{set,cancel}
1800         * Upon request by Elsa.  Anchors now controlled by get-set-pair
1801           like most other CM state variables.
1802    - Compiler.CMSA eliminated
1803         * No longer supported by CM anyway.
1804    - fixed bugs in pickler that kept biting Stefan
1805         * past refs to past refs (was caused by the possibility that
1806           ad-hoc sharing is more discriminating than hash-cons sharing)
1807         * integer overflow on LargeInt.minInt
1808    - ml-{lex,yacc} build scripts now use new mechanism
1809      for building standalone programs
1810    - fixed several gcc -Wall warnings that were caused by missing header
1811      files, missing initializations, etc., in runtime (not all warnings
1812      eliminated, though)

Legend:
Removed from v.570  
changed lines
  Added in v.691

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