13 |
Description: |
Description: |
14 |
---------------------------------------------------------------------- |
---------------------------------------------------------------------- |
15 |
Name: Matthias Blume |
Name: Matthias Blume |
16 |
|
Date: 2001/05/07 14:40:00 EDT |
17 |
|
Tag: blume-20010507-proxies |
18 |
|
Description: |
19 |
|
|
20 |
|
I made into "proxy libraries" all libraries that qualify for such a |
21 |
|
change. (A qualifying library is a library that has another library or |
22 |
|
groups as its sole member and repeats that member's export list |
23 |
|
verbatim. A proxy library avoids this repetition by omitting its export |
24 |
|
list, effectively inheriting the list that its (only) member exports. |
25 |
|
See the CM manual for more explanation.) |
26 |
|
The main effect is that explicit export lists for these libraries |
27 |
|
do not have to be kepts in sync, making maintenance a bit easier. |
28 |
|
|
29 |
|
I also added copyright notices to many .cm-files. |
30 |
|
|
31 |
|
Last but not least, I made a new set of bootfiles. |
32 |
|
|
33 |
|
---------------------------------------------------------------------- |
34 |
|
Name: Matthias Blume |
35 |
|
Date: 2001/05/04 17:00:00 EDT |
36 |
|
Tag: blume-20010504-cm-lsplit |
37 |
|
Description: |
38 |
|
|
39 |
|
0. John merged pending changes to $/smlnj-lib.cm |
40 |
|
|
41 |
|
1. Allen's previous change accidentally backed out of one of Lal's |
42 |
|
earlier changes. I undid this mistake (re-introducing Lal's change). |
43 |
|
|
44 |
|
2. I used the new topOrder' function from graph-scc.sml (from $/smlnj-lib.cm) |
45 |
|
within the compiler where applicable. There is some code simplification |
46 |
|
because of that. |
47 |
|
|
48 |
|
3. The "split" phase (in FLINT) is now part of the default list of phases. |
49 |
|
Compiler.Control.LambdaSplitting.* can be used to globally control the |
50 |
|
lambda-splitting (cross-module-inlining) engine. In addition to that, |
51 |
|
it can now also be controlled on a per-source basis: CM has been taught |
52 |
|
a new tool parameter applicable to ML source files. |
53 |
|
|
54 |
|
- To turn lambda-splitting off completely: |
55 |
|
local open Compiler.Control.LambdaSplitting in |
56 |
|
val _ = set Off |
57 |
|
end |
58 |
|
- To make "no lambda-splitting" the global default (but allow per-source |
59 |
|
overriding); this is the initial setting: |
60 |
|
local open Compiler.Control.LambdaSplitting in |
61 |
|
val _ = set (Default NONE) |
62 |
|
end |
63 |
|
- To make "lambda-splitting with aggressiveness a" the global default |
64 |
|
(and allow per-source overriding): |
65 |
|
local open Compiler.Control.LambdaSplitting in |
66 |
|
val _ = set (Default (SOME a)) |
67 |
|
end |
68 |
|
|
69 |
|
- To turn lambda-splitting off for a given ML souce file (say: a.sml) |
70 |
|
write (in the respective .cm-file): |
71 |
|
a.sml (lambdasplitting:off) |
72 |
|
- To turn lambda-splitting for a.sml on with minimal aggressiveness: |
73 |
|
a.sml (lambdasplitting:on) |
74 |
|
- To turn lambda-splitting for a.sml on with aggressiveness <a> (where |
75 |
|
<a> is a decimal non-negative integer): |
76 |
|
a.sml (lambdasplitting:<a>) |
77 |
|
- To turn lambda-splitting for a.sml on with maximal aggressiveness: |
78 |
|
a.sml (lambdasplitting:infinity) |
79 |
|
- To use the global default for a.sml: |
80 |
|
a.sml (lambdasplitting:default) |
81 |
|
or simply |
82 |
|
a.sml |
83 |
|
|
84 |
|
---------------------------------------------------------------------- |
85 |
|
Name: Allen Leung |
86 |
|
Date: 2001/05/04 01:57:00 EDT |
87 |
|
Tag: leunga-20010504-sync |
88 |
|
Description: |
89 |
|
|
90 |
|
MLRISC features. |
91 |
|
|
92 |
|
1. Fix to CMPXCHG instructions. |
93 |
|
2. Changed RA interface to allow annotations in callbacks. |
94 |
|
3. Added a new method to the stream interface to allow annotations updates. |
95 |
|
|
96 |
|
---------------------------------------------------------------------- |
97 |
|
Name: Matthias Blume |
98 |
|
Date: 2001/05/01 11:45:00 EDT |
99 |
|
Tag: blume-20010501-pcedittmp |
100 |
|
Description: |
101 |
|
|
102 |
|
Changed install.sh to use the current working directory instead of |
103 |
|
/usr/tmp for a temporary file (pcedittmp). The previous choice |
104 |
|
of /usr/tmp caused trouble with MacOS X because of file premission |
105 |
|
problems. |
106 |
|
|
107 |
|
---------------------------------------------------------------------- |
108 |
|
Name: Matthias Blume |
109 |
|
Date: 2001/04/20 11:10:00 EDT |
110 |
|
Tag: blume-20010420-inMLflag |
111 |
|
Description: |
112 |
|
|
113 |
|
- added vp_limitPtrMask to vproc-state.h |
114 |
|
(for use by the raw-C-calls mechanism to implement proper interrupt |
115 |
|
handling) |
116 |
|
- made the ML compiler aware of various data-structure offsets so it |
117 |
|
can generate code for accessing the vp_inML flag and vp_limitPtrMask |
118 |
|
- tweaked mlriscGen.sml to have it emit interrupt-handling code for |
119 |
|
raw C-calls |
120 |
|
|
121 |
|
---------------------------------------------------------------------- |
122 |
|
Name: Lal George |
123 |
|
Date: 2001/04/20 09:15:28 EDT |
124 |
|
Tag: george-20010420-macosX |
125 |
|
Description: |
126 |
|
|
127 |
|
- Changes to port to Mac OS X; Darwin. |
128 |
|
|
129 |
|
- In the process I found that sqrt was broken on the PPC, because the |
130 |
|
fsqrt instruction is not implemented. |
131 |
|
|
132 |
|
---------------------------------------------------------------------- |
133 |
|
Name: Matthias Blume |
134 |
|
Date: 2001/04/18 12:45:00 EDT |
135 |
|
Tag: blume-20010418-ccalls |
136 |
|
Description: |
137 |
|
|
138 |
|
- fixed two off-by-4 errors in the x86-specific c-calls implementation |
139 |
|
(this bug prevented structure arguments containing pointers from being |
140 |
|
passed correctly) |
141 |
|
- changed the raw-C-call code in mlriscGen.sml in such a way that |
142 |
|
structure arguments are represented as a pointer to the beginning |
143 |
|
of the structure (instead of having a series of synthesized arguments, |
144 |
|
one for each structure member) |
145 |
|
|
146 |
|
- made makeml script's verbosity level configurable via environment |
147 |
|
variable (MAKEML_VERBOSITY) |
148 |
|
|
149 |
|
- eliminated placeholder implementations for f32l, w16s, i16s, and f32s |
150 |
|
in rawmem-x86.sml; we are now using the real thing |
151 |
|
|
152 |
|
---------------------------------------------------------------------- |
153 |
|
Name: Matthias Blume |
154 |
|
Date: 2001/03/22 16:25:00 EST |
155 |
|
Tag: blume-20010322-bootfiles |
156 |
|
Description: |
157 |
|
|
158 |
|
Created a new set of bootfiles (for your automatic installation convenience). |
159 |
|
|
160 |
|
---------------------------------------------------------------------- |
161 |
|
Name: Matthias Blume |
162 |
|
Date: 2001/03/22 15:10:00 EST |
163 |
|
Tag: blume-20010322-rawmem-parcm |
164 |
|
Description: |
165 |
|
|
166 |
|
1. All "raw memory access" primitives for the new FFI are implemented now |
167 |
|
(at least on the x86). |
168 |
|
2. Some further cleanup of CM's parallel make mechanism. |
169 |
|
|
170 |
|
---------------------------------------------------------------------- |
171 |
|
Name: Matthias Blume |
172 |
|
Date: 2001/03/19 17:53:00 EST |
173 |
|
Tag: blume-20010319-parallel |
174 |
|
Description: |
175 |
|
|
176 |
|
Parallel make (using compile servers) now works again. |
177 |
|
|
178 |
|
To this end, CM.stabilize and CMB.make have been modified to work in |
179 |
|
two passes when compile servers are attached: |
180 |
|
1. Compile everything, do not perform stabilization; this pass |
181 |
|
uses compile servers |
182 |
|
2. Stabilize everything; this pass does not use compile servers |
183 |
|
If there are no compile servers, the two passes are combined into one |
184 |
|
(as before). Splitting the passes increases the inherent parallelism |
185 |
|
in the dependency graph because the entire graph including all |
186 |
|
libraries is available at the same time. This, in turn, improves |
187 |
|
server utilization. The downside is that the master process will |
188 |
|
have to do some extra work after compilation is done (because for |
189 |
|
technical reasons it must re-read all the binfiles during stabilization). |
190 |
|
|
191 |
|
---------------------------------------------------------------------- |
192 |
|
Name: Matthias Blume |
193 |
|
Date: 2001/03/16 12:22:00 EST |
194 |
|
Tag: blume-20010316-bootfiles |
195 |
|
Description: |
196 |
|
|
197 |
|
Created a new set of bootfiles (for your automatic installation convenience). |
198 |
|
|
199 |
|
---------------------------------------------------------------------- |
200 |
|
Name: Matthias Blume |
201 |
|
Date: 2001/03/16 11:00:00 EST |
202 |
|
Tag: blume-20010316-MLTREE-fixup |
203 |
|
Description: |
204 |
|
|
205 |
|
This is a minor fixup for an (untagged) earlier commit by Allen. |
206 |
|
(A file was missing). |
207 |
|
|
208 |
|
---------------------------------------------------------------------- |
209 |
|
Name: Allen Leung |
210 |
|
Date: Mon Mar 5 18:54:57 EST 2001 |
211 |
|
Tag: leunga-20010305-cut-support |
212 |
|
|
213 |
|
1. New support for alternative control-flow in MLTREE. |
214 |
|
Currently we support |
215 |
|
|
216 |
|
FLOW_TO(CALL ...., [k1,...,kn]) |
217 |
|
|
218 |
|
This is needed for 'cuts to' in C-- and try/handle-like constructs |
219 |
|
in Moby |
220 |
|
|
221 |
|
New assembler flag "asm-show-cutsto" to turn on control-flow debugging. |
222 |
|
|
223 |
|
2. Register Allocator |
224 |
|
|
225 |
|
Changes in interface [from Fermin, John] |
226 |
|
|
227 |
|
3. Alpha 8-bit SLL support [Fermin] |
228 |
|
|
229 |
|
4. All architectures |
230 |
|
|
231 |
|
A new module (ClusterExpandCopies) for expanding parallel copies. |
232 |
|
|
233 |
|
---------------------------------------------------------------------- |
234 |
|
Name: Allen Leung |
235 |
|
Date: 2001/02/27 23:07:00 EST |
236 |
|
Tag: leunga-20010227-minor-stuff |
237 |
|
|
238 |
|
1. Alpha bug fix for CMOVNE |
239 |
|
2. Handle mltree COND(..,FCMP ...,...) |
240 |
|
3. Bug fix in simplifier |
241 |
|
|
242 |
|
---------------------------------------------------------------------- |
243 |
|
Name: Matthias Blume |
244 |
|
Date: 2001/01/30 17:50:00 EST |
245 |
|
Tag: blume-20010130-sync |
246 |
|
Description: |
247 |
|
|
248 |
|
This is just a minor update to sync my devel branch with the main brach. |
249 |
|
The only visible change is the addition of some README files. |
250 |
|
|
251 |
|
---------------------------------------------------------------------- |
252 |
|
Name: Matthias Blume |
253 |
|
Date: 2001/01/12 23:30:00 JST |
254 |
|
Tag: blume-20010112-bootfiles |
255 |
|
Description: |
256 |
|
|
257 |
|
Made a new set of bootfiles that goes with the current state of the |
258 |
|
repository. |
259 |
|
|
260 |
|
---------------------------------------------------------------------- |
261 |
|
Name: Matthias Blume |
262 |
|
Date: 2001/01/12 21:20:00 JST |
263 |
|
Tag: blume-20010112-sync |
264 |
|
Description: |
265 |
|
|
266 |
|
I am just flushing out some minor changes that had accumulated in |
267 |
|
my private branch in order to sync with the main tree. (This is |
268 |
|
mainly because I had CVS trouble when trying to merge _into_ my |
269 |
|
private branch.) |
270 |
|
|
271 |
|
Most people should be completely unaffected by this. |
272 |
|
|
273 |
|
---------------------------------------------------------------------- |
274 |
|
Name: Allen Leung |
275 |
|
Date: Thu Jan 11 21:03:00 EST 2001 |
276 |
|
Tag: leunga-20010111-labexp=mltree |
277 |
|
Description: |
278 |
|
|
279 |
|
1. Removed the type LabelExp and replace it by MLTree. |
280 |
|
2. Rewritten mltree-simplify with the pattern matcher tool. |
281 |
|
3. There were some bugs in alpha code generator which would break |
282 |
|
64-bit code generation. |
283 |
|
4. Redo the tools to generate code with the |
284 |
|
5. The CM files in MLRISC (and in src/system/smlnj/MLRISC) |
285 |
|
are now generated by perl scripts. |
286 |
|
|
287 |
|
---------------------------------------------------------------------- |
288 |
|
Name: Matthias Blume |
289 |
|
Date: 2001/01/10 21:55:00 JST |
290 |
|
Tag: blume-20010110-rcc |
291 |
|
Description: |
292 |
|
|
293 |
|
The RCC stuff now seems to work (but only on the x86). |
294 |
|
This required hacking of the c-calls interface (and -implementation) in |
295 |
|
MLRISC. |
296 |
|
|
297 |
|
Normal compiler users should be unaffected. |
298 |
|
|
299 |
|
---------------------------------------------------------------------- |
300 |
|
Name: Matthias Blume |
301 |
|
Date: 2001/01/09 01:20:00 JST |
302 |
|
Tag: blume-20010109-rcc |
303 |
|
Description: |
304 |
|
|
305 |
|
This is a fairly big patch, flushing out a large number of pending |
306 |
|
changes that I made to my development copy over the last couple of days. |
307 |
|
|
308 |
|
Of practical relevance at this moment is a workaround for a pickling |
309 |
|
bug that Allen ran into the other day. The cause of the bug itself is |
310 |
|
still unknown and it might be hard to fix it properly, but the |
311 |
|
workaround has some merits of its own (namely somewhat reducing pickling |
312 |
|
overhead for certain libraries). Therefore, I think this solution should |
313 |
|
be satisfactory at this time. |
314 |
|
|
315 |
|
The rest of the changes (i.e., the vast majority) has to do with my |
316 |
|
ongoing efforts of providing direct support for C function calls from |
317 |
|
ML. At the moment there is a new primop "RAW_CCALL", typing magic |
318 |
|
in types/cproto.sml (invoked from FLINT/trans/translate.sml), a new |
319 |
|
case in the FLINT CPS datatype (RCC), changes to cps/convert.sml to |
320 |
|
translate uses of RAW_CCALL into RCC, and changes to mlriscGen.sml to |
321 |
|
handle RCC. |
322 |
|
|
323 |
|
The last part (the changes to mlriscGen.sml) are still known to be |
324 |
|
wrong on the x86 and not implemented on all other architectures. But |
325 |
|
the infrastructure is in place. I had to change a few functor |
326 |
|
signatures in the backend to be able to route the CCalls interface |
327 |
|
from MLRISC there, and I had to specialize the mltree type (on the |
328 |
|
x86) to include the necessary extensions. (The extensions themselves |
329 |
|
were already there and redy to go in MLRISC/x86). |
330 |
|
|
331 |
|
Everything should be very happy as soon as someone helps me with |
332 |
|
mlriscGen.sml... |
333 |
|
|
334 |
|
In any case, nothing of this should matter to anyone as long as the |
335 |
|
new primop is not being used (which is going to be the case unless you |
336 |
|
find it where I hid it :). The rest of the compiler is completely |
337 |
|
unaffected. |
338 |
|
|
339 |
|
---------------------------------------------------------------------- |
340 |
|
Name: Matthias Blume |
341 |
|
Date: 2001/01/05 00:30:00 JST |
342 |
|
Tag: blume-20010105-primops |
343 |
|
Description: |
344 |
|
|
345 |
|
Added some experimental support for work that I am doing right now. |
346 |
|
These changes mostly concern added primops, but there is also a new |
347 |
|
experimental C library in the runtime system (but currently not enabled |
348 |
|
anywhere except on Linux/X86). |
349 |
|
|
350 |
|
In the course of adding primops (and playing with them), I discovered that |
351 |
|
Zhong's INL_PRIM hack (no type info for certain primops) was, in fact, badly |
352 |
|
broken. (Zhong was very right he labeled this stuff as "major gross hack".) |
353 |
|
To recover, I made type information in INL_PRIM mandatory and changed |
354 |
|
prim.sml as well as built-in.sml accordingly. The InLine structure now |
355 |
|
has complete, correct type information (i.e., no bottom types). |
356 |
|
|
357 |
|
Since all these changes mean that we need new binfiles, I also bumped the |
358 |
|
version number to 110.32.1. |
359 |
|
|
360 |
|
---------------------------------------------------------------------- |
361 |
|
Name: Matthias Blume |
362 |
|
Date: 2000/12/30 22:10:00 JST |
363 |
|
Tag: blume-20001230-various |
364 |
|
Description: |
365 |
|
|
366 |
|
Added proxy libraries for MLRISC and let MLRISC libraries refer |
367 |
|
to each other using path anchors. (See CM manual for explanation.) |
368 |
|
|
369 |
|
Updated CM documentation. |
370 |
|
|
371 |
|
Fixed some bugs in CM. |
372 |
|
|
373 |
|
Implemented "proxy" libraries (= syntactic sugar for CM). |
374 |
|
|
375 |
|
Added "-quiet" option to makeml and changed runtime system accordingly. |
376 |
|
|
377 |
|
Added cleanup handler for exportML to reset timers and compiler stats. |
378 |
|
|
379 |
|
---------------------------------------------------------------------- |
380 |
|
Name: Lal George |
381 |
|
Date: 2000/12/22 22:22:58 EST 2000 |
382 |
|
Tag: Release_110_32 |
383 |
|
Description: |
384 |
|
|
385 |
|
Infinite precision used throughout MLRISC. |
386 |
|
see MLRISC/mltree/machine-int.sig |
387 |
|
|
388 |
|
---------------------------------------------------------------------- |
389 |
|
Name: Matthias Blume |
390 |
|
Date: 2000/12/22 23:16:00 JST |
391 |
|
Tag: blume-20001222-warn |
392 |
|
Description: |
393 |
|
|
394 |
|
Corrected wording and formatting of some CM warning message which I |
395 |
|
broke in my previous patch. |
396 |
|
|
397 |
|
---------------------------------------------------------------------- |
398 |
|
Name: Matthias Blume |
399 |
|
Date: 2000/12/22 21:20:00 JST |
400 |
|
Tag: blume-20001222-anchorenv |
401 |
|
Description: |
402 |
|
|
403 |
|
Fixed CM's handling of anchor environments in connection with CMB.make. |
404 |
|
|
405 |
|
---------------------------------------------------------------------- |
406 |
|
Name: Matthias Blume |
407 |
|
Date: 2000/12/22 13:15:00 JST |
408 |
|
Tag: blume-20001222-cleanup |
409 |
|
Description: |
410 |
|
|
411 |
|
Removed src/cm/ffi which does not (and did not) belong here. |
412 |
|
|
413 |
|
---------------------------------------------------------------------- |
414 |
|
Name: Matthias Blume |
415 |
|
Date: 2000/12/21 23:55:00 JST |
416 |
|
Tag: blume-20001221-exn |
417 |
|
Description: |
418 |
|
|
419 |
|
Probably most important: CM no longer silently swallows all exceptions |
420 |
|
in the compiler. |
421 |
|
Plus: some other minor CM changes. For example, CM now reports some |
422 |
|
sizes for generated binfiles (code, data, envpickle, lambdapickle). |
423 |
|
|
424 |
|
---------------------------------------------------------------------- |
425 |
|
Name: Matthias Blume |
426 |
|
Date: 2000/12/15 00:01:05 JST |
427 |
|
Tag: blume-20001215-dirtool |
428 |
|
Description: |
429 |
|
|
430 |
|
- "dir" tool added. |
431 |
|
- improvements and cleanup to Tools structure |
432 |
|
- documentation updates |
433 |
|
|
434 |
|
---------------------------------------------------------------------- |
435 |
|
Name: Allen Leung |
436 |
|
Date: Thu Dec 14 03:45:24 EST 2000 |
437 |
|
Description: |
438 |
|
Tag: leunga-20001214-int-inf |
439 |
|
Description: |
440 |
|
|
441 |
|
In IntInf, added these standard functions, which are missing from our |
442 |
|
implementation: |
443 |
|
|
444 |
|
andb : int * int -> int |
445 |
|
xorb : int * int -> int |
446 |
|
orb : int * int -> int |
447 |
|
notb : int -> int |
448 |
|
<< : int * word -> int |
449 |
|
~>> : int * word -> int |
450 |
|
|
451 |
|
Not tested, I hope they are correct. |
452 |
|
|
453 |
|
---------------------------------------------------------------------- |
454 |
|
Name: Allen Leung |
455 |
|
Date: Fri Dec 8 19:23:26 EST 2000 |
456 |
|
Description: |
457 |
|
Tag: leunga-20001208-nowhere |
458 |
|
Description: |
459 |
|
|
460 |
|
Slight improvements to the 'nowhere' tool to handle OR-patterns, |
461 |
|
to generate better error messages etc. Plus a brief manual. |
462 |
|
|
463 |
|
---------------------------------------------------------------------- |
464 |
|
Name: Lal George |
465 |
|
Date: 2000/12/08 09:54:02 EST 2000 |
466 |
|
Tag: Release_110_31 |
467 |
|
Description: |
468 |
|
|
469 |
|
- Version 110.31 |
470 |
|
---------------------------------------------------------------------- |
471 |
|
Name: Allen Leung |
472 |
|
Date: Thu Dec 7 22:01:04 EST 2000 |
473 |
|
Tag: leunga-20001207-cell-monster-hack |
474 |
|
Description: |
475 |
|
|
476 |
|
Major MLRISC internal changes. Affect all clients. |
477 |
|
Summary: |
478 |
|
|
479 |
|
1. Type CELLS.cell = int is now replaced by a datatype. |
480 |
|
As a result, the old regmap is now gone. Almost all interfaces |
481 |
|
in MLRISC change as a consequence. |
482 |
|
|
483 |
|
2. A new brand version of machine description tool (v3.0) that generates |
484 |
|
modules expecting the new interface. The old version is removed. |
485 |
|
|
486 |
|
3. The RA interface has been further abstracted into two new functors. |
487 |
|
RISC_RA and X86RA. These functors have much simpler interfaces. |
488 |
|
[See also directory MLRISC/demo.] |
489 |
|
|
490 |
|
4. Some other new source->source code generation tools are available: |
491 |
|
|
492 |
|
a. MLRISC/Tools/RewriteGen -- generate rewriters from rules. |
493 |
|
b. MLRISC/Tools/WhereGen -- expands conditional pattern matching rules. |
494 |
|
I use this tool to generate the peephole optimizers---with the new |
495 |
|
cell type changes, peephole rules are becoming difficult to write |
496 |
|
without conditional pattern matching. |
497 |
|
|
498 |
|
5. More Intmap -> IntHashTable change. Previous changes by Matthias didn't |
499 |
|
cover the entire MLRISC source tree so many things broke. |
500 |
|
|
501 |
|
6. CM files have been moved to the subdirectory MLRISC/cm. |
502 |
|
They are moved because there are a lot of them and they clutter up the |
503 |
|
root dir. |
504 |
|
|
505 |
|
7. More detailed documentation to come... |
506 |
|
|
507 |
|
NOTE: To rebuild from 110.30 (ftp distribution), you'll have to do |
508 |
|
a makeml -rebuild first. This is because of other other |
509 |
|
changes that Matthias has made (see below). |
510 |
|
|
511 |
|
|
512 |
|
---------------------------------------------------------------------- |
513 |
|
Name: Matthias Blume |
514 |
|
Date: 2000/11/30 23:12:00 JST |
515 |
|
Tag: blume-20001130-filereorg |
516 |
|
Description: |
517 |
|
|
518 |
|
Some manual updates and some file reorganizations in CM. |
519 |
|
|
520 |
|
---------------------------------------------------------------------- |
521 |
|
Name: Matthias Blume |
522 |
|
Date: 2000/11/24 17:45:00 JST |
523 |
|
Tag: blume-20001124-link |
524 |
|
Description: |
525 |
|
|
526 |
|
Drastically improved link traversal code for the case that the dynamic |
527 |
|
value was already loaded at bootstrap time. As a result, CM and CMB |
528 |
|
now both load blazingly fast -- even on a very slow machine. Also, |
529 |
|
memory consumption has been further reduced by this. |
530 |
|
|
531 |
|
Warning: The format of the PIDMAP file has changed. THerefore, to |
532 |
|
bootstrap you have to do this: |
533 |
|
|
534 |
|
1. Run CMB.make |
535 |
|
2. Make a symbolic link for the boot directory: |
536 |
|
ln -s sml.boot.ARCH-OS xxx |
537 |
|
3. "Rebuild" the boot directory: |
538 |
|
./makeml -boot xxx -rebuild sml ; rm xxx |
539 |
|
4. Boot normally: |
540 |
|
./makeml |
541 |
|
|
542 |
|
---------------------------------------------------------------------- |
543 |
|
Name: Matthias Blume |
544 |
|
Date: 2000/11/21 21:20:00 JST |
545 |
|
Tag: blume-20001121-tools |
546 |
|
Description: |
547 |
|
|
548 |
|
Continued hacking on autoloading problem -- with success this time. |
549 |
|
Also changed tool-plugin mechanism. See new CM manual. |
550 |
|
|
551 |
|
---------------------------------------------------------------------- |
552 |
|
Name: Matthias Blume |
553 |
|
Date: 2000/11/19 14:30:00 JST |
554 |
|
Tag: blume-20001119-autoload |
555 |
|
Description: |
556 |
|
|
557 |
|
Some hacking to make autoloading faster. Success for CMB, no success |
558 |
|
so far for CM. There is a reduced structure CM' that autoloads faster. |
559 |
|
(This is a temporary, non-documented hack to be eliminated again when |
560 |
|
the general problem is solved.) |
561 |
|
|
562 |
|
---------------------------------------------------------------------- |
563 |
|
Name: Matthias Blume |
564 |
|
Date: 2000/11/17 14:10:00 JST |
565 |
|
Tag: blume-20001117-pickle-lib |
566 |
|
Description: |
567 |
|
|
568 |
|
1. Eliminated comp-lib.cm |
569 |
|
2. Made pickle-lib.cm |
570 |
|
3. Eliminated all uses of intset.sml (from comp-lib.cm) |
571 |
|
4. Replaced all uses of intmap.{sig,sml} (from comp-lib.cm) with |
572 |
|
equivalent constructs from smlnj-lib.cm (INtHashTable). |
573 |
|
5. Point 4. also goes for those uses of intmap.* in MLRISC. |
574 |
|
Duplicated intmap modules thrown out. |
575 |
|
6. Hunted down all duplicated SCC code and replaced it with |
576 |
|
equivalent stuff (GraphSCCFn from smlnj-lib.cm). |
577 |
|
7. Rewrote Feedback module. |
578 |
|
8. Moved sortedlist.sml into viscomp-lib.cm. Eventually it |
579 |
|
should be thrown out and equivalent modules from smlnj-lib.cm |
580 |
|
should be used (IntRedBlackSet, IntListSet, ...). |
581 |
|
|
582 |
|
Confirmed that compiler compiles to fixpoint. |
583 |
|
|
584 |
|
---------------------------------------------------------------------- |
585 |
|
Name: Allen Leung |
586 |
|
Date: 2000/11/10 18:00:00 |
587 |
|
Tag: leunga-20001110-new-x86-fp |
588 |
|
|
589 |
|
A new x86 floating point code generator has been added. |
590 |
|
By default this is turned off. To turn this on, do: |
591 |
|
|
592 |
|
CM.autoload "$smlnj/compiler.cm"; |
593 |
|
Compiler.Control.MLRISC.getFlag "x86-fast-fp" := true; |
594 |
|
|
595 |
|
Changes: |
596 |
|
|
597 |
|
1. Changed FTAN to FPTAN so that the assembly output is correct. |
598 |
|
2. Changed the extension callback for FTANGENT to generate: |
599 |
|
|
600 |
|
fptan |
601 |
|
fstp %st(0) |
602 |
|
instead of |
603 |
|
fptan |
604 |
|
fstpl ftempmem |
605 |
|
|
606 |
|
3. Numerous assembly fixes for x86. |
607 |
|
|
608 |
|
5. Cleaned up the machine code output module x86/x86MC.sml and added |
609 |
|
support for a whole bunch of instructions and addressing modes: |
610 |
|
|
611 |
|
fadd/fsub/fsubr/fmul/fdiv/fdivr %st, %st(n) |
612 |
|
faddp/fsubp/fsubrp/fmulp/fdivp/fdivrp %st, %st(n) |
613 |
|
fadd/fsub/fsubr/fmul/fdiv/fdivr %st(n), %st |
614 |
|
fiadd/fisub/fisubr/fimul/fidiv/fidivr mem |
615 |
|
fxch %st(n) |
616 |
|
fld %st(n) |
617 |
|
fst %st(n) |
618 |
|
fst mem |
619 |
|
fstp %st(n) |
620 |
|
fucom %st(n) |
621 |
|
fucomp %st(n) |
622 |
|
|
623 |
|
All these are now generated when the fast fp mode is turned on. |
624 |
|
|
625 |
|
6. Removed the dedicated registers %st(0), ..., %st(7) from X86CpsRegs |
626 |
|
|
627 |
|
---------------------------------------------------------------------- |
628 |
|
Name: Matthias Blume |
629 |
|
Date: 2000/11/09 11:20:00 JST |
630 |
|
Tag: blume-20001109-scc |
631 |
|
Description: |
632 |
|
|
633 |
|
Eliminated some code duplication: |
634 |
|
|
635 |
|
1. Added "where" clause to GraphSCCFn in SML/NJ Library. |
636 |
|
(Otherwise the functor is useless.) |
637 |
|
2. Used GraphSCCFn where SCCUtilFun was used previously. |
638 |
|
3. Got rid of SCCUtilFun (in comp-lib.cm). |
639 |
|
|
640 |
|
---------------------------------------------------------------------- |
641 |
|
Name: Lal George |
642 |
|
Date: 2000/11/06 09:02:21 EST 2000 |
643 |
|
Tag: Release_110_30 |
644 |
|
Description: |
645 |
|
|
646 |
|
- Version 110.30 |
647 |
|
---------------------------------------------------------------------- |
648 |
|
Name: Matthias Blume |
649 |
|
Date: 2000/11/04 14:45:00 |
650 |
|
Tag: blume-20001104-mlbuild |
651 |
|
Description: |
652 |
|
|
653 |
|
- Made ml-build faster on startup. |
654 |
|
- Documentation fixes. |
655 |
|
|
656 |
|
---------------------------------------------------------------------- |
657 |
|
Name: Matthias Blume |
658 |
|
Date: 2000/11/02 17:00:00 JST |
659 |
|
Tag: blume-20001102-condcomp |
660 |
|
Description: |
661 |
|
|
662 |
|
- Small tweaks to pickler -- new BOOTFILES! |
663 |
|
- Version bumped to 110.29.2. |
664 |
|
- Added conditional compilation facility to init.cmi (see comment there). |
665 |
|
---------------------------------------------------------------------- |
666 |
|
Name: Allen Leung |
667 |
|
Date: 2000/10/23 19:31:00 |
668 |
|
Tag: leunga-20001023-demo-ra |
669 |
|
|
670 |
|
1. Minor RA changes that improves spilling on x86 (affects Moby and C-- only) |
671 |
|
2. Test programs for the graph library updated |
672 |
|
3. Some new MLRISC demo programs added |
673 |
|
|
674 |
|
---------------------------------------------------------------------- |
675 |
|
Name: Matthias Blume |
676 |
|
Date: 2000/08/31 22:15:00 JST |
677 |
|
Tag: blume-20001017-errmsg |
678 |
|
Description: |
679 |
|
|
680 |
|
More error message grief: Where there used to be no messages, there |
681 |
|
now were some that had bogus error regions. Fixed. |
682 |
|
|
683 |
|
---------------------------------------------------------------------- |
684 |
|
Name: Matthias Blume |
685 |
|
Date: 2000/08/31 17:30:00 JST |
686 |
|
Tag: blume-20001017-v110p29p1 |
687 |
|
Description: |
688 |
|
|
689 |
|
I made a version 110.29.1 with new bootfiles. |
690 |
|
|
691 |
|
Changes: Modified pickler/unpickler for faster and leaner unpickling. |
692 |
|
CM documentation changes and a small bugfix in CM's error reporting. |
693 |
|
|
694 |
|
---------------------------------------------------------------------- |
695 |
|
Name: Lal George |
696 |
|
Date: 2000/09/27 14:42:35 EDT |
697 |
|
Tag: george-20000927-nodestatus |
698 |
|
Description: |
699 |
|
|
700 |
|
Changed the type of the nodestatus, so that: |
701 |
|
|
702 |
|
SPILLED(~1) is now SPILLED |
703 |
|
SPILLED(m) where m>=0 is now MEMREG(m) |
704 |
|
SPILLED(s) where s<~1 is now SPILL_LOC(~s) |
705 |
|
|
706 |
|
---------------------------------------------------------------------- |
707 |
|
Name: Matthias Blume |
708 |
|
Date: 2000/09/07 14:45:00 JST |
709 |
|
Tag: blume-20000907-cmerrmsg |
710 |
|
Description: |
711 |
|
|
712 |
|
Small tweak to CM to avoid getting ML syntax error messages twice. |
713 |
|
|
714 |
|
---------------------------------------------------------------------- |
715 |
|
Name: Matthias Blume |
716 |
Date: 2000/08/31 18:00:00 JST |
Date: 2000/08/31 18:00:00 JST |
717 |
Tag: blume-20000831-cvsbootfiles |
Tag: blume-20000831-cvsbootfiles |
718 |
Description: |
Description: |