Diff of /sml/trunk/HISTORY

Parent Directory | Revision Log | Patch

	revision 756, Thu Dec 14 16:01:33 2000 UTC	revision 828, Fri May 11 18:32:07 2001 UTC
#	Line 11	Line 11
11	Date:	Date:
12	Tag: <post-commit CVS tag>	Tag: <post-commit CVS tag>
13	Description:	Description:
14
15		----------------------------------------------------------------------
16		Name: Matthias Blume
17		Date: 2001/05/11 14:35:00 EDT
18		Tag: blume-20010511-ml-nlffi
19		Description:
20
21		I finally took the plunge and added my new FFI code to the main
22		repository.  For x86-linux it is now ready for prime-time.
23
24		There are two new subdirectories of "src":
25
26		- ml-nlffi-lib:
27		The utility library for programs using the FFI interface.
28		Here is the implementation of $/c.cm and its associated low-level
29		partners $/c-int.cm and $/memory.cm.
30		- ml-nlffigen:
31		A stand-alone program for generating ML glue code from C source
32		code.
33
34		Building ml-nlffigen requires $/ckit-lib.cm.
35
36		The config/install.sh script has been updates to do the Right Thing
37		(hopefully).
38
39		Notice that the source tree for the C-Kit will not be put under "src"
40		but directly under the installation root directory.  (This is the
41		structure that currently exists on the CVS server when you check out
42		module "sml".)  Fortunately, config/install.sh knows about this oddity.
43
44		Bugs: No documentation yet.
45
46		----------------------------------------------------------------------
47		Name: Matthias Blume
48		Date: 2001/05/09 16:35:00 EDT
49		Tag: blume-20010509-cpscontract
50		Description:
51
52		Fixed a bug in the accounting code in cpsopt/contract.sml.  (The
53		wrapper/unwrapper elimination did not decrement usage counts and some
54		dead variables got overlooked by the dead-up logic.)
55
56		----------------------------------------------------------------------
57		Name: Lal George
58		Date: 2001/05/08  17:26:09 EDT
59		Tag: george-20010508-omit-frameptr
60		Description:
61
62		Changes to implement the omit-frame-pointer optimization to support
63		raw C calls. For now, there is only support on the Intel x86, but
64		other architectures will follow as more experience is gained with this.
65
66
67		----------------------------------------------------------------------
68		Name: Matthias Blume
69		Date: 2001/05/07 14:40:00 EDT
70		Tag: blume-20010507-proxies
71		Description:
72
73		I made into "proxy libraries" all libraries that qualify for such a
74		change.  (A qualifying library is a library that has another library or
75		groups as its sole member and repeats that member's export list
76		verbatim.  A proxy library avoids this repetition by omitting its export
77		list, effectively inheriting the list that its (only) member exports.
78		See the CM manual for more explanation.)
79		The main effect is that explicit export lists for these libraries
80		do not have to be kepts in sync, making maintenance a bit easier.
81
82		I also added copyright notices to many .cm-files.
83
84		Last but not least, I made a new set of bootfiles.
85
86		----------------------------------------------------------------------
87		Name: Matthias Blume
88		Date: 2001/05/04 17:00:00 EDT
89		Tag: blume-20010504-cm-lsplit
90		Description:
91
92		0. John merged pending changes to $/smlnj-lib.cm
93
94		1. Allen's previous change accidentally backed out of one of Lal's
95		earlier changes.  I undid this mistake (re-introducing Lal's change).
96
97		2. I used the new topOrder' function from graph-scc.sml (from $/smlnj-lib.cm)
98		within the compiler where applicable.  There is some code simplification
99		because of that.
100
101		3. The "split" phase (in FLINT) is now part of the default list of phases.
102		Compiler.Control.LambdaSplitting.* can be used to globally control the
103		lambda-splitting (cross-module-inlining) engine.  In addition to that,
104		it can now also be controlled on a per-source basis: CM has been taught
105		a new tool parameter applicable to ML source files.
106
107		- To turn lambda-splitting off completely:
108		local open Compiler.Control.LambdaSplitting in
109		val _ = set Off
110		end
111		- To make "no lambda-splitting" the global default (but allow per-source
112		overriding); this is the initial setting:
113		local open Compiler.Control.LambdaSplitting in
114		val _ = set (Default NONE)
115		end
116		- To make "lambda-splitting with aggressiveness a" the global default
117		(and allow per-source overriding):
118		local open Compiler.Control.LambdaSplitting in
119		val _ = set (Default (SOME a))
120		end
121
122		- To turn lambda-splitting off for a given ML souce file (say: a.sml)
123		write (in the respective .cm-file):
124		a.sml (lambdasplitting:off)
125		- To turn lambda-splitting for a.sml on with minimal aggressiveness:
126		a.sml (lambdasplitting:on)
127		- To turn lambda-splitting for a.sml on with aggressiveness <a> (where
128		<a> is a decimal non-negative integer):
129		a.sml (lambdasplitting:<a>)
130		- To turn lambda-splitting for a.sml on with maximal aggressiveness:
131		a.sml (lambdasplitting:infinity)
132		- To use the global default for a.sml:
133		a.sml (lambdasplitting:default)
134		or simply
135		a.sml
136
137		----------------------------------------------------------------------
138		Name: Allen Leung
139		Date: 2001/05/04 01:57:00 EDT
140		Tag: leunga-20010504-sync
141		Description:
142
143		MLRISC features.
144
145		1. Fix to CMPXCHG instructions.
146		2. Changed RA interface to allow annotations in callbacks.
147		3. Added a new method to the stream interface to allow annotations updates.
148
149		----------------------------------------------------------------------
150		Name: Matthias Blume
151		Date: 2001/05/01 11:45:00 EDT
152		Tag: blume-20010501-pcedittmp
153		Description:
154
155		Changed install.sh to use the current working directory instead of
156		/usr/tmp for a temporary file (pcedittmp).  The previous choice
157		of /usr/tmp caused trouble with MacOS X because of file premission
158		problems.
159
160		----------------------------------------------------------------------
161		Name: Matthias Blume
162		Date: 2001/04/20 11:10:00 EDT
163		Tag: blume-20010420-inMLflag
164		Description:
165
166		- added vp_limitPtrMask to vproc-state.h
167		(for use by the raw-C-calls mechanism to implement proper interrupt
168		handling)
169		- made the ML compiler aware of various data-structure offsets so it
170		can generate code for accessing the vp_inML flag and vp_limitPtrMask
171		- tweaked mlriscGen.sml to have it emit interrupt-handling code for
172		raw C-calls
173
174		----------------------------------------------------------------------
175		Name: Lal George
176		Date: 2001/04/20 09:15:28 EDT
177		Tag: george-20010420-macosX
178		Description:
179
180		- Changes to port to Mac OS X; Darwin.
181
182		- In the process I found that sqrt was broken on the PPC, because the
183		fsqrt instruction is not implemented.
184
185		----------------------------------------------------------------------
186		Name: Matthias Blume
187		Date: 2001/04/18 12:45:00 EDT
188		Tag: blume-20010418-ccalls
189		Description:
190
191		- fixed two off-by-4 errors in the x86-specific c-calls implementation
192		(this bug prevented structure arguments containing pointers from being
193		passed correctly)
194		- changed the raw-C-call code in mlriscGen.sml in such a way that
195		structure arguments are represented as a pointer to the beginning
196		of the structure (instead of having a series of synthesized arguments,
197		one for each structure member)
198
199		- made makeml script's verbosity level configurable via environment
200		variable (MAKEML_VERBOSITY)
201
202		- eliminated placeholder implementations for f32l, w16s, i16s, and f32s
203		in rawmem-x86.sml; we are now using the real thing
204
205		----------------------------------------------------------------------
206		Name: Matthias Blume
207		Date: 2001/03/22 16:25:00 EST
208		Tag: blume-20010322-bootfiles
209		Description:
210
211		Created a new set of bootfiles (for your automatic installation convenience).
212
213		----------------------------------------------------------------------
214		Name: Matthias Blume
215		Date: 2001/03/22 15:10:00 EST
216		Tag: blume-20010322-rawmem-parcm
217		Description:
218
219		1. All "raw memory access" primitives for the new FFI are implemented now
220		(at least on the x86).
221		2. Some further cleanup of CM's parallel make mechanism.
222
223		----------------------------------------------------------------------
224		Name: Matthias Blume
225		Date: 2001/03/19 17:53:00 EST
226		Tag: blume-20010319-parallel
227		Description:
228
229		Parallel make (using compile servers) now works again.
230
231		To this end, CM.stabilize and CMB.make have been modified to work in
232		two passes when compile servers are attached:
233		1. Compile everything, do not perform stabilization; this pass
234		uses compile servers
235		2. Stabilize everything; this pass does not use compile servers
236		If there are no compile servers, the two passes are combined into one
237		(as before).  Splitting the passes increases the inherent parallelism
238		in the dependency graph because the entire graph including all
239		libraries is available at the same time.  This, in turn, improves
240		server utilization.  The downside is that the master process will
241		have to do some extra work after compilation is done (because for
242		technical reasons it must re-read all the binfiles during stabilization).
243
244		----------------------------------------------------------------------
245		Name: Matthias Blume
246		Date: 2001/03/16 12:22:00 EST
247		Tag: blume-20010316-bootfiles
248		Description:
249
250		Created a new set of bootfiles (for your automatic installation convenience).
251
252		----------------------------------------------------------------------
253		Name: Matthias Blume
254		Date: 2001/03/16 11:00:00 EST
255		Tag: blume-20010316-MLTREE-fixup
256		Description:
257
258		This is a minor fixup for an (untagged) earlier commit by Allen.
259		(A file was missing).
260
261		----------------------------------------------------------------------
262		Name: Allen Leung
263		Date: Mon Mar  5 18:54:57 EST 2001
264		Tag: leunga-20010305-cut-support
265
266		1. New support for alternative control-flow in MLTREE.
267		Currently we support
268
269		FLOW_TO(CALL ...., [k1,...,kn])
270
271		This is needed for 'cuts to' in C-- and try/handle-like constructs
272		in Moby
273
274		New assembler flag "asm-show-cutsto" to turn on control-flow debugging.
275
276		2. Register Allocator
277
278		Changes in interface [from Fermin, John]
279
280		3. Alpha 8-bit SLL support [Fermin]
281
282		4. All architectures
283
284		A new module (ClusterExpandCopies) for expanding parallel copies.
285
286		----------------------------------------------------------------------
287		Name: Allen Leung
288		Date: 2001/02/27 23:07:00 EST
289		Tag: leunga-20010227-minor-stuff
290
291		1. Alpha bug fix for CMOVNE
292		2. Handle mltree COND(..,FCMP ...,...)
293		3. Bug fix in simplifier
294
295		----------------------------------------------------------------------
296		Name: Matthias Blume
297		Date: 2001/01/30 17:50:00 EST
298		Tag: blume-20010130-sync
299		Description:
300
301		This is just a minor update to sync my devel branch with the main brach.
302		The only visible change is the addition of some README files.
303
304		----------------------------------------------------------------------
305		Name: Matthias Blume
306		Date: 2001/01/12 23:30:00 JST
307		Tag: blume-20010112-bootfiles
308		Description:
309
310		Made a new set of bootfiles that goes with the current state of the
311		repository.
312
313		----------------------------------------------------------------------
314		Name: Matthias Blume
315		Date: 2001/01/12 21:20:00 JST
316		Tag: blume-20010112-sync
317		Description:
318
319		I am just flushing out some minor changes that had accumulated in
320		my private branch in order to sync with the main tree.  (This is
321		mainly because I had CVS trouble when trying to merge _into_ my
322		private branch.)
323
324		Most people should be completely unaffected by this.
325
326		----------------------------------------------------------------------
327		Name: Allen Leung
328		Date: Thu Jan 11 21:03:00 EST 2001
329		Tag: leunga-20010111-labexp=mltree
330		Description:
331
332		1.  Removed the type LabelExp and replace it by MLTree.
333		2.  Rewritten mltree-simplify with the pattern matcher tool.
334		3.  There were some bugs in alpha code generator which would break
335		64-bit code generation.
336		4.  Redo the tools to generate code with the
337		5.  The CM files in MLRISC (and in src/system/smlnj/MLRISC)
338		are now generated by perl scripts.
339
340		----------------------------------------------------------------------
341		Name: Matthias Blume
342		Date: 2001/01/10 21:55:00 JST
343		Tag: blume-20010110-rcc
344		Description:
345
346		The RCC stuff now seems to work (but only on the x86).
347		This required hacking of the c-calls interface (and -implementation) in
348		MLRISC.
349
350		Normal compiler users should be unaffected.
351
352		----------------------------------------------------------------------
353		Name: Matthias Blume
354		Date: 2001/01/09 01:20:00 JST
355		Tag: blume-20010109-rcc
356		Description:
357
358		This is a fairly big patch, flushing out a large number of pending
359		changes that I made to my development copy over the last couple of days.
360
361		Of practical relevance at this moment is a workaround for a pickling
362		bug that Allen ran into the other day.  The cause of the bug itself is
363		still unknown and it might be hard to fix it properly, but the
364		workaround has some merits of its own (namely somewhat reducing pickling
365		overhead for certain libraries).  Therefore, I think this solution should
366		be satisfactory at this time.
367
368		The rest of the changes (i.e., the vast majority) has to do with my
369		ongoing efforts of providing direct support for C function calls from
370		ML.  At the moment there is a new primop "RAW_CCALL", typing magic
371		in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new
372		case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to
373		translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to
374		handle RCC.
375
376		The last part (the changes to mlriscGen.sml) are still known to be
377		wrong on the x86 and not implemented on all other architectures.  But
378		the infrastructure is in place. I had to change a few functor
379		signatures in the backend to be able to route the CCalls interface
380		from MLRISC there, and I had to specialize the mltree type (on the
381		x86) to include the necessary extensions. (The extensions themselves
382		were already there and redy to go in MLRISC/x86).
383
384		Everything should be very happy as soon as someone helps me with
385		mlriscGen.sml...
386
387		In any case, nothing of this should matter to anyone as long as the
388		new primop is not being used (which is going to be the case unless you
389		find it where I hid it :). The rest of the compiler is completely
390		unaffected.
391
392		----------------------------------------------------------------------
393		Name: Matthias Blume
394		Date: 2001/01/05 00:30:00 JST
395		Tag: blume-20010105-primops
396		Description:
397
398		Added some experimental support for work that I am doing right now.
399		These changes mostly concern added primops, but there is also a new
400		experimental C library in the runtime system (but currently not enabled
401		anywhere except on Linux/X86).
402
403		In the course of adding primops (and playing with them), I discovered that
404		Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly
405		broken.  (Zhong was very right he labeled this stuff as "major gross hack".)
406		To recover, I made type information in INL_PRIM mandatory and changed
407		prim.sml as well as built-in.sml accordingly.  The InLine structure now
408		has complete, correct type information (i.e., no bottom types).
409
410		Since all these changes mean that we need new binfiles, I also bumped the
411		version number to 110.32.1.
412
413		----------------------------------------------------------------------
414		Name: Matthias Blume
415		Date: 2000/12/30 22:10:00 JST
416		Tag: blume-20001230-various
417		Description:
418
419		Added proxy libraries for MLRISC and let MLRISC libraries refer
420		to each other using path anchors.  (See CM manual for explanation.)
421
422		Updated CM documentation.
423
424		Fixed some bugs in CM.
425
426		Implemented "proxy" libraries (= syntactic sugar for CM).
427
428		Added "-quiet" option to makeml and changed runtime system accordingly.
429
430		Added cleanup handler for exportML to reset timers and compiler stats.
431
432		----------------------------------------------------------------------
433		Name: Lal George
434		Date: 2000/12/22 22:22:58 EST 2000
435		Tag: Release_110_32
436		Description:
437
438		Infinite precision used throughout MLRISC.
439		see MLRISC/mltree/machine-int.sig
440
441		----------------------------------------------------------------------
442		Name: Matthias Blume
443		Date: 2000/12/22 23:16:00 JST
444		Tag: blume-20001222-warn
445		Description:
446
447		Corrected wording and formatting of some CM warning message which I
448		broke in my previous patch.
449
450		----------------------------------------------------------------------
451		Name: Matthias Blume
452		Date: 2000/12/22 21:20:00 JST
453		Tag: blume-20001222-anchorenv
454		Description:
455
456		Fixed CM's handling of anchor environments in connection with CMB.make.
457
458		----------------------------------------------------------------------
459		Name: Matthias Blume
460		Date: 2000/12/22 13:15:00 JST
461		Tag: blume-20001222-cleanup
462		Description:
463
464		Removed src/cm/ffi which does not (and did not) belong here.
465
466		----------------------------------------------------------------------
467		Name: Matthias Blume
468		Date: 2000/12/21 23:55:00 JST
469		Tag: blume-20001221-exn
470		Description:
471
472		Probably most important: CM no longer silently swallows all exceptions
473		in the compiler.
474		Plus: some other minor CM changes.  For example, CM now reports some
475		sizes for generated binfiles (code, data, envpickle, lambdapickle).
476
477	----------------------------------------------------------------------	----------------------------------------------------------------------
478	Name: Matthias Blume	Name: Matthias Blume
479	Date: 2000/12/15 00:01:05 JST	Date: 2000/12/15 00:01:05 JST

---

Colored Diff Long Colored Diff Unidiff Context Diff Side by Side

Legend: Removed from v.756   changed lines   Added in v.828

---

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