Diff of /sml/trunk/HISTORY

Parent Directory | Revision Log | Patch

	revision 876, Wed Jul 11 15:42:20 2001 UTC	revision 937, Thu Sep 20 21:20:28 2001 UTC
#	Line 14	Line 14
14
15	----------------------------------------------------------------------	----------------------------------------------------------------------
16	Name: Matthias Blume	Name: Matthias Blume
17		Date: 2001/09/20 17:20:00 EDT
18		Tag: blume-20010920-canonicalpaths
19		Description:
20
21		0. Updated the BOOT file (something that I forgot to do earlier).
22
23		1. Small internal change to CM so that it avoids "/../" in filenames
24		as much as possible (but only where it is safe).
25
26		2. Changed config/_run-sml (resulting in a changed bin/.run-sml) so
27		that arguments that contain delimiters are passed through correctly.
28		This change also means that all "special" arguments of the form
29		@SMLxxx... must come first.
30
31		3. Changed install script to put relative anchor names for tool commands
32		into pathconfig.
33
34		----------------------------------------------------------------------
35		Name: Matthias Blume
36		Date: 2001/09/18 15:35:00 EDT
37		Tag: blume-20010918-readme11036
38		Description:
39
40		Added README files.
41
42		----------------------------------------------------------------------
43		Name: Matthias Blume
44		Date: 2001/09/18 11:45:00 EDT
45		Tag: Release_110_36 (retag)
46		Description:
47
48		Fixed mistake in config/preloads. Retagged as 110.36.
49
50		----------------------------------------------------------------------
51		Name: Matthias Blume
52		Date: 2001/09/18 09:40:00 EDT
53		Tag: Release_110_36_orig (tag changed)
54		Description:
55
56		New version (110.36).  New bootfiles.
57
58		----------------------------------------------------------------------
59		Name: Matthias Blume
60		Date: 2001/09/14 16:15:00 EDT
61		Tag: blume-20010914-x86fastfp
62		Description:
63
64		John committed some changes that Allen made, in particular a (hopefully)
65		correctly working version of the x86-fp module.
66
67		I changed the default setting of the Control.MLRISC.getFlag "x86-fast-fp"
68		flag to "true".  Everything seems to compile to a fixpoint ok, and
69		"mandelbrot" speeds up by about 15%.
70
71		----------------------------------------------------------------------
72		Name: Matthias Blume
73		Date: 2001/09/13 11:20:00 EDT
74		Tag: blume-20010913-minimal
75		Description:
76
77		1. Stefan Monnier's patch to fix a miscompilation problem that
78		was brought to light by John Reppy's work on Moby.
79
80		2. Implemented a minimal "structure Compiler" that contains just
81		"version" and "architecture".  The minimal version will be
82		available when the full version is not.  This is for backward-
83		compatibility with code that wants to test Compiler.version.
84
85		----------------------------------------------------------------------
86		Name: Matthias Blume
87		Date: 2001/08/28 14:03:00 EDT
88		Tag: blume-20010828-ml-lex
89		Description:
90
91		Fix for bug 1581, received from Neophytos Michael.
92
93		----------------------------------------------------------------------
94		Name: Matthias Blume
95		Date: 2001/08/27 11:20:00 EDT
96		Tag: blume-20010827-readme11035
97		Description:
98
99		Fleshed out the README file for 110.35.
100
101		----------------------------------------------------------------------
102		Name: Matthias Blume
103		Date: 2001/08/24 17:10:00 EDT
104		Tag: Release_110_35
105		Description:
106
107		New version number (110.35).  New bootfiles.
108
109		----------------------------------------------------------------------
110		Name: Lal George
111		Date: 2001/08/24 13:47:18 EDT 2001
112		Tag: george-20010824-MLRISC-graphs
113		Description:
114
115		removed clusters from MLRISC completely and replaced with graphs.
116
117		----------------------------------------------------------------------
118		Name: Matthias Blume
119		Date: 2001/08/23 17:50:00 EDT
120		Tag: blume-20010823-toplevel
121		Description:
122
123		- some reorganization of the code that implements various kinds of
124		environments in the compiler (static, dynamic, symbolic, combined)
125		- re-implemented the EnvRef module so that evalStream works properly
126		(if the stream contains references to "use", "CM.make", etc.)
127		- cleaned up evalloop.sml and interact.sml (but they need more cleaning)
128
129		----------------------------------------------------------------------
130		Name: Matthias Blume
131		Date: 2001/08/20 15:50 EDT
132		Tag: blume20010820-slipup
133		Description:
134
135		I forgot to commit a few files.  Here they are...
136
137		----------------------------------------------------------------------
138		Name: Matthias Blume
139		Date: 2001/08/20 15:35:00 EDT
140		Tag: blume-20010820-debugprof
141		Description:
142
143		!!!! NEW BOOTFILES !!!!
144
145		This is another round of reorganizing the compiler sources.  This
146		time the main goal was to factor out all the "instrumentation"
147		passes (for profiling and backtracing) into their own library.
148		The difficulty was to do it in such a way that it does not depend
149		on elaborate.cm but only on elabdata.cm.
150
151		Therefore there have been further changes to both elaborate.cm and
152		elabdata.cm -- more "generic" things have been moved from the former
153		to the latter.  As a result, I was forced to split the assignment
154		of numbers indicating "primtyc"s into two portions: SML-generic and
155		SML/NJ-specific.  Since it would have been awkward to maintain,
156		I bit the bullet and actually _changed_ the mapping between these
157		numbers and primtycs.  The bottom line of this is that you need
158		a new set of bin- and bootfiles.
159
160		I have built new bootfiles for all architectures, so doing a fresh
161		checkout and config/install.sh should be all you need.
162
163		The newly created library's name is
164
165		$smlnj/viscomp/debugprof.cm
166
167		and its sources live under
168
169		src/compiler/DebugProf
170
171		----------------------------------------------------------------------
172		Name: Matthias Blume
173		Date: 2001/08/15 17:15:00 EDT
174		Tag: blume-20010815-compreorg
175		Description:
176
177		This is a first cut at reorganizing the CM libraries that make up the
178		core of the compiler.  The idea is to separate out pieces that could
179		be used independently by tools, e.g., the parser, the typechecker, etc.
180
181		The current status is a step in this direction, but it is not quite
182		satisfactory yet.  Expect more changes in the future.
183
184		Here is the current (new) organization...
185
186		What used to be $smlnj/viscomp/core.cm is now divided into
187		six CM libraries:
188
189		$smlnj/viscomp/basics.cm
190		/parser.cm
191		/elabdata.cm
192		/elaborate.cm
193		/execute.cm
194		/core.cm
195
196		The CM files for these libraries live under src/system/smlnj/viscomp.
197		All these libraries are proxy libraries that contain precisely
198		one CM library component.  Here are the locations of the components
199		(all within the src/compiler tree):
200
201		Basics/basics.cm
202		Parse/parser.cm
203		ElabData/elabdata.cm
204		Elaborator/elaborate.cm
205		Execution/execute.cm
206		core.cm
207
208		[This organization is the same that has been used already
209		for a while for the architecture-specific parts of the visible
210		compiler and for the old version of core.cm.]
211
212		As you will notice, many source files have been moved from their
213		respective original locations to a new home in one of the above
214		subtrees.
215
216		The division of labor between the new libraries is the following:
217
218		basics.cm:
219		- Simple, basic definitions that pertain to many (or all) of
220		the other libraries.
221		parser.cm:
222		- The SML parser, producing output of type Ast.dec.
223		- The type family for Ast is also defined and exported here.
224		elabdata.cm:
225		- The datatypes that describe input and output of the elaborator.
226		This includes types, absyn, and static environments.
227		elaborator.cm:
228		- The SML/NJ type checker and elaborator.
229		This maps an Ast.dec (with a given static environment) to
230		an Absyn.dec (with a new static environment).
231		- This libraries implements certain modules that used to be
232		structures as functors (to remove dependencies on FLINT).
233		execute.cm:
234		- Everything having to do with executing binary code objects.
235		- Dynamic environments.
236		core.cm:
237		- SML/NJ-specific instantiations of the elaborator and MLRISC.
238		- Top-level modules.
239		- FLINT (this should eventually become its own library)
240
241		Notes:
242
243		I am not 100% happy with the way I separated the elaborator (and its
244		data structures) from FLINT.  Two instances of the same problem:
245
246		1. Data structures contain certain fields that carry FLINT-specific
247		information.  I hacked around this using exn and the property list
248		module from smlnj-lib.  But the fact that there are middle-end
249		specific fields around at all is a bit annoying.
250
251		2. The elaborator calculates certain FLINT-related information.  I tried
252		to make this as abstract as I could using functorization, but, again,
253		the fact that the elaborator has to perform calculations on behalf
254		of the middle-end at all is not nice.
255
256		3. Having to used exn and property lists is unfortunate because it
257		weakens type checking.  The other alternative (parameterizing
258		nearly *everything*) is not appealing, though.
259
260		I removed the "rebinding =" warning hack because due to the new organization
261		it was awkward to maintain it.  As a result, the compiler now issues some of
262		these warnings when compiling init.cmi during bootstrap compilation. On
263		the plus side, you also get a warning when you do, for example:
264		val op = = Int32.+
265		which was not the case up to now.
266
267		I placed "assign" and "deref" into the _Core structure so that the
268		code that deals with the "lazy" keyword can find them there.  This
269		removes the need for having access to the primitive environment
270		during elaboration.
271
272		----------------------------------------------------------------------
273		Name: Matthias Blume
274		Date: 2001/08/13
275		Tag: blume-20010813-closures
276		Description:
277
278		This fix was sent to us by Zhong Shao.  It is supposed to improve the
279		performance of certain loops by avoiding needless closure allocation.
280
281		----------------------------------------------------------------------
282		Name: Lal George
283		Date: 2001/07/31 10:03:23 EDT 2001
284		Tag: george-20010731-x86-fmalloc
285		Description: Fixed bug in x86 calls
286
287		There was a bug where call instructions would mysteriously
288		vanish. The call instruction had to be one that returned
289		a floating point value.
290
291		----------------------------------------------------------------------
292		Name: Lal George
293		Date: 2001/07/19 16:36:29 EDT 2001
294		Tag: george-20010719-simple-cells
295		Description:
296
297		I have dramatically simplified the interface for CELLS in MLRISC.
298
299		In summary, the cells interface is broken up into three parts:
300
301		1. CellsBasis : CELLS_BASIS
302
303		CellsBasis is a top level structure and common for all
304		architectures.  it contains the definitions of basic datatypes
305		and utility  functions over these types.
306
307		2. functor Cells() : CELLS
308
309		Cells generates an interface for CELLS that incorporates the
310		specific resources on the target architecture, such as the
311		presence of special register classes, their number and size,
312		and various useful substructures.
313
314		3. <ARCH>CELLS
315
316		e.g. SparcCells: SPARCCELLS
317
318		<ARCH>CELLS usually contains additional bindings for special
319		registers  on the architecture, such as:
320
321		val r0 : cell           (* register zero *)
322		val y : cell            (* Y register *)
323		val psr : cell          (* processor status register *)
324		...
325
326		The structure returned by applying the Cells functor is opened
327		in this interface.
328
329		The main implication of all this is that the datatypes for cells is
330		split between CellsBasis and CELLS -- a fairly simple change for user
331		code.
332
333		In the old scheme the CELLS interface had a definitional binding of
334		the form:
335
336		signature CELLS = sig
337
338		structure CellsBasis = CellsBasis
339
340		...
341
342		end
343
344		With all the sharing constraints that goes on in MLRISC, this old
345		design  quickly leads to errors such as:
346
347		"structure definition spec inside of sharing ... "
348
349
350		and appears to require an unacceptable amount of sharing and where
351		constraint hackery.
352
353		I think this error message (the interaction of definitional specs and
354		sharing) requires more explanation on our web page.
355
356		----------------------------------------------------------------------
357		Name: Matthias Blume
358		Date: 2001/07/19 15:00:00 EDT
359		Tag: blume-20010719-libreorg
360		Description:
361
362		This update puts together a fairly extensive but straightforward change
363		to the way the libraries that implement the interactive system are
364		organized:
365
366		The biggest change is the elimination of structure Compiler.  As a
367		replacement for this structure, there is now a CM library
368		(known as $smlnj/compiler.cm or $smlnj/compiler/current.cm)
369		that exports all the substructures of the original structure Compiler
370		directly.  So instead of saying Compiler.Foo.bar one now simply
371		says Foo.bar.  (The CM libraries actually export a collection of
372		structures that is richer than the collection of substructures of
373		structure Compiler.)
374
375		To make the transition smooth, there is a separate library called
376		$smlnj/compiler/compiler.cm which puts together and exports the
377		original structure Compiler (or at least something very close to it).
378
379		There are five members of the original structure Compiler
380		that are not exported directly but which instead became members
381		of a new structure Backend (described by signature BACKEND).  These are:
382		structure Profile (: PROFILE), structure Compile (: COMPILE), structure
383		Interact (: INTERACT), structure Machine (: MACHINE), and val
384		architecture (: string).
385
386		Structure Compiler.Version has become structure CompilerVersion.
387
388		Cross-compilers for alpha32, hppa, ppc, sparc, and x86 are provided
389		by $smlnj/compiler/<arch>.cm where <arch> is alpha32, hppa, ppc, sparc,
390		or x86, respectively.
391		Each of these exports the same frontend structures that
392		$smlnj/compiler.cm exports.  But they do not have a structure Backend
393		and instead export some structure <Arch>Backend where <Arch> is Alpha32,
394		Hppa, PPC, Sparc, or X86, respectively.
395
396		Library $smlnj/compiler/all.cm exports the union of the exports of
397		$smlnj/compiler/<arch>.cm
398
399		There are no structures <Arch>Compiler anymore, use
400		$smlnj/compiler/<arch>.cm instead.
401
402		Library host-compiler-0.cm is gone.  Instead, the internal library
403		that instantiates CM is now called cm0.cm.  Selection of the host
404		compiler (backend) is no longer done here but. (Responsibility for it
405		now lies with $smlnj/compiler/current.cm.  This seems to be more
406		logical.)
407
408		Many individual files have been moved or renamed.  Some files have
409		been split into multiple files, and some "dead" files have been deleted.
410
411		Aside from these changes to library organization, there are also changes
412		to the way the code itself is organized:
413
414		Structure Binfile has been re-implemented in such a way that it no
415		longer needs any knowledge of the compiler.  It exclusively deals
416		with the details of binfile layout.  It no longer invokes the
417		compiler (for the purpose of creating new prospective binfile
418		content), and it no longer has any knowledge of how to interpret
419		pickles.
420
421		Structure Compile (: COMPILE) has been stripped down to the bare
422		essentials of compilation.  It no longer deals with linking/execution.
423		The interface has been cleaned up considerably.
424
425		Utility routines for dealing with linking and execution have been
426		moved into their own substructures.
427
428		(The ultimate goal of these changes is to provide a light-weight
429		binfile loader/linker (at least for, e.g., stable libraries) that
430		does not require CM or the compiler to be present.)
431
432		CM documentation has been updated to reflect the changes to library
433		organization.
434
435		----------------------------------------------------------------------
436		Name: Matthias Blume
437	Date: 2001/07/10 17:30:00 EDT	Date: 2001/07/10 17:30:00 EDT
438	Tag: Release_110_34	Tag: Release_110_34
439	Description:	Description:
#	Line 2589	Line 3009
3009	elaborator). There were a lot of changes during my "linkpath" trials	elaborator). There were a lot of changes during my "linkpath" trials
3010	that could have been reverted to their original state but weren't.	that could have been reverted to their original state but weren't.
3011	Please, don't be too harsh on me for messing with this code a bit more	Please, don't be too harsh on me for messing with this code a bit more
3012	than what was strictly necessary...  (I _did_ resist the tempation	than what was strictly necessary...  (I _did_ resist the temptation
3013	of doing any "global reformatting" to avoid an untimely death at	of doing any "global reformatting" to avoid an untimely death at
3014	Dave's hands. :)	Dave's hands. :)
3015

---

Colored Diff Long Colored Diff Unidiff Context Diff Side by Side

Legend: Removed from v.876   changed lines   Added in v.937

---

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