SCM Repository
Annotation of /sml/trunk/HISTORY
Parent Directory
| 
Revision Log
Revision 641 - (view) (download)
| 1 : | dbm | 570 | This is the HISTORY file for the Yale SML/NJ CVS repository. |
| 2 : | |||
| 3 : | An entry should be made for _every_ commit to the repository. | ||
| 4 : | The entries in this file will be used when creating the README | ||
| 5 : | for new versions, so keep that in mind when writing the | ||
| 6 : | description. | ||
| 7 : | |||
| 8 : | The form of an entry should be: | ||
| 9 : | |||
| 10 : | Name: | ||
| 11 : | Date: | ||
| 12 : | Tag: <post-commit CVS tag> | ||
| 13 : | Description: | ||
| 14 : | leunga | 591 | ---------------------------------------------------------------------- |
| 15 : | leunga | 640 | Name: Allen Leung |
| 16 : | leunga | 641 | Date: 2000/05/10 21:17 EDT |
| 17 : | Tag: leunga-20000510-moby-c--ssa | ||
| 18 : | Description: | ||
| 19 : | |||
| 20 : | Various bug fixes and new features for C--, Moby and MLRISC optimizations. | ||
| 21 : | None of these affect SML/NJ. | ||
| 22 : | |||
| 23 : | 1. Register Allocation | ||
| 24 : | |||
| 25 : | a. A new ra spilling module (ra/ra-spill-with-renaming) is implemented. | ||
| 26 : | This module tries to remove local (i.e. basic block level) redundancies | ||
| 27 : | during spilling. | ||
| 28 : | |||
| 29 : | b. A new framework for performing region based register allocation. | ||
| 30 : | Not yet entirely functional. | ||
| 31 : | |||
| 32 : | 2. X86 | ||
| 33 : | |||
| 34 : | a. DefUse for POP was missing the stack pointer [found by Lal] | ||
| 35 : | b. Reload for CALL was incorrect in X86Spill [found by John] | ||
| 36 : | c. Various fixes in X86Spill so that it can be used correctly for | ||
| 37 : | the new spilling module. | ||
| 38 : | |||
| 39 : | 3. SSA/IR | ||
| 40 : | |||
| 41 : | a. New module ir/dj-dataflow.sml implements elimination based | ||
| 42 : | data flow analysis. | ||
| 43 : | |||
| 44 : | 4. MLRiscGen | ||
| 45 : | |||
| 46 : | a. Fix for gc type annotation | ||
| 47 : | |||
| 48 : | 5. MDGen | ||
| 49 : | |||
| 50 : | Various fixes for machine description -> ml code translation. For ssa | ||
| 51 : | only. | ||
| 52 : | |||
| 53 : | ---------------------------------------------------------------------- | ||
| 54 : | Name: Allen Leung | ||
| 55 : | leunga | 640 | Date: 2000/05/08 22:17 EDT |
| 56 : | Tag: leunga-20000508-labexp | ||
| 57 : | Description: | ||
| 58 : | |||
| 59 : | Fermin has found a few assembly problems with constant expressions | ||
| 60 : | generated in LabelExp. Mostly, the problems involve extra parentheses, | ||
| 61 : | which choke on dumb assemblers. This is his fix. | ||
| 62 : | |||
| 63 : | ---------------------------------------------------------------------- | ||
| 64 : | dbm | 635 | Name: Dave MacQueen |
| 65 : | Date: 2000/04/09 14:00 EDT | ||
| 66 : | Tag: dbm-20000502-Version_110_28 | ||
| 67 : | Description: | ||
| 68 : | |||
| 69 : | 1. Updated src/compiler/TopLevel/main/version.sml to version 110.28 | ||
| 70 : | |||
| 71 : | 2. Updated config/version to 110.28 | ||
| 72 : | |||
| 73 : | 3. Updated config/srcarchiveurl | ||
| 74 : | |||
| 75 : | 3. New boot files! | ||
| 76 : | ftp://ftp.research.bell-labs.com/dist/smlnj/working/110.28/ | ||
| 77 : | |||
| 78 : | ---------------------------------------------------------------------- | ||
| 79 : | blume | 634 | Name: Matthias Blume |
| 80 : | Date: 2000/05/01 19:05:00 JST | ||
| 81 : | Tag: blume-20000501-noweb | ||
| 82 : | Description: | ||
| 83 : | |||
| 84 : | A new noweb tool has been added. The existing system is entirely | ||
| 85 : | unaffected by this, but some CM users have asked for renewed noweb | ||
| 86 : | support. Everything is documented in the CM manual. | ||
| 87 : | |||
| 88 : | New (plugin) libraries: | ||
| 89 : | |||
| 90 : | noweb-tool.cm | ||
| 91 : | nw-ext.cm | ||
| 92 : | |||
| 93 : | ---------------------------------------------------------------------- | ||
| 94 : | dbm | 633 | Name: Dave MacQueen |
| 95 : | Date: 2000/04/30 12:40PM EDT | ||
| 96 : | Tag: dbm-20000430-bug_fixes | ||
| 97 : | Description: | ||
| 98 : | |||
| 99 : | 1. Fix for bug 1498 | ||
| 100 : | smlnj/src/system/Basis/Implementation/Unsafe/object.sig | ||
| 101 : | smlnj/src/system/Basis/Implementation/Unsafe/object.sml | ||
| 102 : | added toRealArray function | ||
| 103 : | smlnj/src/compiler/MiscUtil/print/ppobj.sml | ||
| 104 : | added check for tag Obj.RealArray to array printing case in ppObj | ||
| 105 : | |||
| 106 : | 2. Fix for bug 1510 | ||
| 107 : | smlnj/src/compiler/Semant/types/typesutil.sml | ||
| 108 : | fixed definition of dummyargs (used by equalTycon) so that | ||
| 109 : | dummy args are distinct types | ||
| 110 : | |||
| 111 : | ---------------------------------------------------------------------- | ||
| 112 : | blume | 629 | Name: Matthias Blume |
| 113 : | blume | 632 | Date: 2000/04/30 01:00:00 JST |
| 114 : | Tag: blume-20000430-versions | ||
| 115 : | Description: | ||
| 116 : | |||
| 117 : | 1. CM version numbering added. This is an implementation of Lal's | ||
| 118 : | proposal for adding version numbers and version checking to .cm | ||
| 119 : | files. Lal said that his proposal was just that -- a proposal. | ||
| 120 : | For the time being I went ahead and implemented it so that people | ||
| 121 : | can comment on it. Everything is completely backward-compatible | ||
| 122 : | (except for the stable library format, i.e., new bootfiles!). | ||
| 123 : | |||
| 124 : | As usual, see the CM manual for details. | ||
| 125 : | |||
| 126 : | 2. An alternative syntax for anchored paths has been implemented. | ||
| 127 : | Dave has recently voiced the same concerns that I had when I did | ||
| 128 : | this, so there should be some support. My take is that eventually | ||
| 129 : | I will let support for the current syntax (where anchors are | ||
| 130 : | "implicit") fade out in favor of the new, explicit syntax. | ||
| 131 : | In order to be backward-compatible, both old and new syntax are | ||
| 132 : | currently supported. | ||
| 133 : | |||
| 134 : | Again, see the CM manual for details. | ||
| 135 : | |||
| 136 : | 3. Parallel make is trying to be slightly smarter: When the master | ||
| 137 : | process finds a "bottleneck", i.e., when there is only one | ||
| 138 : | compilation unit that can be compiled and everybody else is | ||
| 139 : | waiting on it, then it will simply compile it directly instead | ||
| 140 : | of clumsily telling one of the slaves to do it. | ||
| 141 : | |||
| 142 : | 4. Support for "unsharing" added. This is necessary in order to be | ||
| 143 : | able to have two different versions of the same library running | ||
| 144 : | at the same time (e.g., for trying out a new MLRISC while still | ||
| 145 : | having the old MLRISC linked into the current compiler, etc.) | ||
| 146 : | See the CM manual. | ||
| 147 : | |||
| 148 : | 5. Simple "makedepend" functionality added for generating Makefile | ||
| 149 : | dependency information. (This is rather crude at the moment. | ||
| 150 : | Expect some changes here in the future.) | ||
| 151 : | |||
| 152 : | 6. ".fun" added as a recognized suffix for ML files. Also documented | ||
| 153 : | explicitly in the manual that the fallback behavior (unknown suffix | ||
| 154 : | -> ML file) is not an official feature! | ||
| 155 : | |||
| 156 : | 7. Small changes to the pickler for stable libraries. | ||
| 157 : | |||
| 158 : | 8. Several internal changes to CM (for cleanup/improvement). | ||
| 159 : | |||
| 160 : | |||
| 161 : | !!!! NEW BINFILES !!!! | ||
| 162 : | |||
| 163 : | ---------------------------------------------------------------------- | ||
| 164 : | Name: Matthias Blume | ||
| 165 : | blume | 631 | Date: 2000/04/28 17:30:00 JST |
| 166 : | Tag: blume-20000428-pathconfig | ||
| 167 : | Description: | ||
| 168 : | |||
| 169 : | 1. I changed config/install.sh to remove duplicate entries from the | ||
| 170 : | lib/pathconfig file at the end. Moreover, the final version of | ||
| 171 : | lib/pathconfig is sorted alphabetically. The same (sorting) is done | ||
| 172 : | in src/system/installml. | ||
| 173 : | |||
| 174 : | 2. The config/install.sh script now consistently uses relative | ||
| 175 : | pathnames in lib/pathconfig whenever the anchor is in the lib | ||
| 176 : | directory. (So far this was true for the libraries that come | ||
| 177 : | pre-compiled and bundled as part of the bootfiles but not for | ||
| 178 : | libraries that are compiled by the script itself.) | ||
| 179 : | |||
| 180 : | ---------------------------------------------------------------------- | ||
| 181 : | Name: Matthias Blume | ||
| 182 : | blume | 629 | Date: 2000/04/26 13:10:00 JST |
| 183 : | Tag: blume-20000426-fun_suffix | ||
| 184 : | Description: | ||
| 185 : | |||
| 186 : | Added ".fun" as a recognized file name suffix (for ML code). | ||
| 187 : | |||
| 188 : | ---------------------------------------------------------------------- | ||
| 189 : | leunga | 628 | Name: Allen Leung |
| 190 : | Date: 2000/04/25 17:00:00 EST | ||
| 191 : | Tag: leunga-20000425-alpha-ra | ||
| 192 : | Description: | ||
| 193 : | |||
| 194 : | 1. Alpha | ||
| 195 : | |||
| 196 : | PSEUDOARITH was missing in AlphaRewrite. This causes an endless loop | ||
| 197 : | in C--. | ||
| 198 : | |||
| 199 : | 2. RA | ||
| 200 : | |||
| 201 : | Added a flag "ra-dump-size" to print out the size of the flowgraph | ||
| 202 : | and the interference graph. | ||
| 203 : | |||
| 204 : | ---------------------------------------------------------------------- | ||
| 205 : | dbm | 627 | Name: Dave MacQueen |
| 206 : | Date: 2000/04/25/ | ||
| 207 : | Tag: dbm-20000425-mlyacc_doc_examples | ||
| 208 : | Description: | ||
| 209 : | Updated mlyacc.tex sections 5 and 7 for SML '97 and CM. | ||
| 210 : | Updated all three examples in src/ml-yacc/examples to run | ||
| 211 : | under 110.* using CM.make. | ||
| 212 : | |||
| 213 : | ---------------------------------------------------------------------- | ||
| 214 : | leunga | 624 | Name: Allen Leung |
| 215 : | Date: 2000/04/20 23:04:00 EST | ||
| 216 : | Tag: leunga-20000420-ssa-c---stuff | ||
| 217 : | Description: | ||
| 218 : | |||
| 219 : | This update synchronizes my repository with Yale's. Most of these | ||
| 220 : | changes, however, do not affect SML/NJ at all (the RA is an exception). | ||
| 221 : | |||
| 222 : | 1. Register Allocator | ||
| 223 : | |||
| 224 : | a. An improvement in the interference graph construction: | ||
| 225 : | Given a copy | ||
| 226 : | |||
| 227 : | s <- t | ||
| 228 : | |||
| 229 : | leunga | 628 | no interference edge between s and t is added for this definition of s. |
| 230 : | leunga | 624 | |
| 231 : | b. I've added two new spill heuristic modules that Fermin and I developed | ||
| 232 : | (in the new library RA.cm). These are unused in SML/NJ but maybe | ||
| 233 : | useful for others (Moby?) | ||
| 234 : | |||
| 235 : | 2. X86 | ||
| 236 : | |||
| 237 : | a. Various fixes in the backend provided by Fermin [C--] and Lal. | ||
| 238 : | |||
| 239 : | 3. Alpha | ||
| 240 : | |||
| 241 : | a. Added the BSR instruction and code generation that goes with it [C--] | ||
| 242 : | b. Other fixes too numerous to recount provided by Fermin [C--] | ||
| 243 : | |||
| 244 : | 4. Regmaps | ||
| 245 : | |||
| 246 : | a. The regmaps are not initialized with the identity physical bindings | ||
| 247 : | at creation time. This is unneeded. | ||
| 248 : | |||
| 249 : | 5. MLRISC Optimizations | ||
| 250 : | |||
| 251 : | a. The DJ-Graph module can now compute the iterated dominance frontiers | ||
| 252 : | intersects with liveness incrementally in linear time! Woohoo! | ||
| 253 : | This is now used in my new SSA construction algorithm. | ||
| 254 : | |||
| 255 : | b. THe branch reorganization module is now smarter about linear chains of | ||
| 256 : | basic blocks. | ||
| 257 : | |||
| 258 : | |||
| 259 : | ---------------------------------------------------------------------- | ||
| 260 : | blume | 615 | Name: Matthias Blume |
| 261 : | Date: 2000/04/12 13:52:00 JST | ||
| 262 : | Tag: blume_main_v110p27_1 | ||
| 263 : | Description: | ||
| 264 : | |||
| 265 : | Changed install.sh script to handle archive files without version number | ||
| 266 : | and to use "boot.<arch>-<os>" instead of "sml.boot.<arch>-<os>" for the | ||
| 267 : | name of the boot file archive. | ||
| 268 : | |||
| 269 : | ---------------------------------------------------------------------- | ||
| 270 : | dbm | 607 | Name: Dave MacQueen |
| 271 : | Date: 2000/04/09 14:00 EDT | ||
| 272 : | Tag: dbm-20000410-Version_110_27 | ||
| 273 : | Description: | ||
| 274 : | |||
| 275 : | 1. Updated src/compiler/TopLevel/main/version.sml to version 110.27 | ||
| 276 : | |||
| 277 : | 2. Updated src/config/version to 110.27 | ||
| 278 : | |||
| 279 : | 3. New boot files! | ||
| 280 : | |||
| 281 : | ---------------------------------------------------------------------- | ||
| 282 : | leunga | 606 | Name: Allen Leung |
| 283 : | Date: 2000/04/09 19:09:00 EST | ||
| 284 : | Tag: leunga-20000409-misc | ||
| 285 : | Description: | ||
| 286 : | |||
| 287 : | 1. Yet another fix for x86 assembly for idivl, imull, mull and friends. | ||
| 288 : | |||
| 289 : | 2. Miscellaneous improvements to MLRISC (unused in sml/nj) | ||
| 290 : | |||
| 291 : | ---------------------------------------------------------------------- | ||
| 292 : | monnier | 605 | Name: Stefan |
| 293 : | Date: 2000/04/07 10:00:00 EDT | ||
| 294 : | Tag: monnier-20000406-branch-handling | ||
| 295 : | Description: | ||
| 296 : | |||
| 297 : | Improved handling of branches (mostly those generated from | ||
| 298 : | polymorphic equality), removed switchoff and changed the | ||
| 299 : | default optimization settings (more cpsopt and less flintopt). | ||
| 300 : | |||
| 301 : | ---------------------------------------------------------------------- | ||
| 302 : | leunga | 601 | Name: Allen Leung |
| 303 : | leunga | 602 | Date: 2000/04/06 01:30:00 EST |
| 304 : | Tag: leunga-20000406-peephole-x86-SSA-2 | ||
| 305 : | Description: | ||
| 306 : | |||
| 307 : | Forgot a few files. | ||
| 308 : | |||
| 309 : | ---------------------------------------------------------------------- | ||
| 310 : | Name: Allen Leung | ||
| 311 : | leunga | 601 | Date: 2000/04/06 00:36:00 EST |
| 312 : | Tag: leunga-20000406-peephole-x86-SSA | ||
| 313 : | Description: | ||
| 314 : | |||
| 315 : | 1. New Peephole code | ||
| 316 : | |||
| 317 : | 2. Minor improvement to X86 instruction selection | ||
| 318 : | |||
| 319 : | 3. Various fixes to SSA and machine description -> code translator | ||
| 320 : | |||
| 321 : | ---------------------------------------------------------------------- | ||
| 322 : | blume | 595 | Name: Matthias Blume |
| 323 : | Date: 2000/04/05 12:30:00 JST | ||
| 324 : | Tag: blume_main_v110p26p2_3 | ||
| 325 : | Description: | ||
| 326 : | |||
| 327 : | This update just merges three minor cosmetic updates to CM's sources | ||
| 328 : | to get ready for the 110.27 code freeze on Friday. No functionality | ||
| 329 : | has changed. | ||
| 330 : | |||
| 331 : | ---------------------------------------------------------------------- | ||
| 332 : | leunga | 593 | Name: Allen Leung |
| 333 : | leunga | 594 | Date: 2000/04/04 19:39:00 EST |
| 334 : | Tag: leunga-20000404-x86-asm | ||
| 335 : | Description: | ||
| 336 : | |||
| 337 : | 1. Fixed a problem in X86 assembly. | ||
| 338 : | |||
| 339 : | Things like | ||
| 340 : | |||
| 341 : | jmp %eax | ||
| 342 : | jmp (%eax) | ||
| 343 : | |||
| 344 : | should be output as | ||
| 345 : | |||
| 346 : | jmp *%eax | ||
| 347 : | jmp *(%eax) | ||
| 348 : | |||
| 349 : | 2. Assembly output | ||
| 350 : | |||
| 351 : | Added a new flag | ||
| 352 : | |||
| 353 : | "asm-indent-copies" (default to false) | ||
| 354 : | |||
| 355 : | When this flag is on, parallel copies will be indented an extra level. | ||
| 356 : | |||
| 357 : | ---------------------------------------------------------------------- | ||
| 358 : | Name: Allen Leung | ||
| 359 : | leunga | 593 | Date: 2000/04/04 03:18:00 EST |
| 360 : | Tag: leunga-20000404-C--Moby | ||
| 361 : | Description: | ||
| 362 : | |||
| 363 : | All of these fixes are related to C--, Moby, and my own optimization | ||
| 364 : | stuff; so they shouldn't affect SML/NJ. | ||
| 365 : | |||
| 366 : | 1. X86 | ||
| 367 : | |||
| 368 : | Various fixes related floating point, and extensions. | ||
| 369 : | |||
| 370 : | 2. Alpha | ||
| 371 : | |||
| 372 : | Some extra patterns related to loads with signed/zero extension | ||
| 373 : | provided by Fermin. | ||
| 374 : | |||
| 375 : | 3. Assembly | ||
| 376 : | |||
| 377 : | When generating assemby, resolve the value of client defined constants, | ||
| 378 : | instead of generating symbolic values. This is controlled by the | ||
| 379 : | new flag "asm-resolve-constants", which is default to true. | ||
| 380 : | |||
| 381 : | 4. Machine Descriptions | ||
| 382 : | |||
| 383 : | a. The precedence parser was slightly broken when parsing infixr symbols. | ||
| 384 : | b. The type generalizing code had the bound variables reversed, resulting | ||
| 385 : | in a problem during arity raising. | ||
| 386 : | c. Various fixes in machine descriptions. | ||
| 387 : | |||
| 388 : | ---------------------------------------------------------------------- | ||
| 389 : | blume | 592 | Name: Matthias Blume |
| 390 : | Date: 2000/04/03 16:05:00 JST | ||
| 391 : | Tag: blume_main_v110p26p2_2 | ||
| 392 : | Description: | ||
| 393 : | |||
| 394 : | I eliminated coreEnv from compInfo. Access to the "Core" structure is | ||
| 395 : | now done via the ordinary static environment that is context to each | ||
| 396 : | compilation unit. | ||
| 397 : | |||
| 398 : | To this end, I arranged that instead of "structure Core" as "structure | ||
| 399 : | _Core" is bound in the pervasive environment. Core access is done via | ||
| 400 : | _Core (which can never be accidentially rebound because _Core is not a | ||
| 401 : | legal surface-syntax symbol). | ||
| 402 : | |||
| 403 : | The current solution is much cleaner because the core environment is | ||
| 404 : | now simply part of the pervasive environment which is part of every | ||
| 405 : | compilation unit's context anyway. In particular, this eliminates all | ||
| 406 : | special-case handling that was necessary until now in order to deal | ||
| 407 : | with dynamic and symbolic parts of the core environment. | ||
| 408 : | |||
| 409 : | Remaining hackery (to bind the "magic" symbol _Core) is localized in the | ||
| 410 : | compilation mananger's bootstrap compiler (actually: in the "init group" | ||
| 411 : | handling). See the comments in src/system/smlnj/init/init.cmi for | ||
| 412 : | more details. | ||
| 413 : | |||
| 414 : | I also tried to track down all mentions of "Core" (as string argument | ||
| 415 : | to Symbol.strSymbol) in the compiler and replaced them with a | ||
| 416 : | reference to the new CoreSym.coreSym. Seems cleaner since the actual | ||
| 417 : | name appears in one place only. | ||
| 418 : | |||
| 419 : | Binfile and bootfile format have not changed, but the switchover from | ||
| 420 : | the old "init.cmi" to the new one is a bit tricky, so I supplied new | ||
| 421 : | bootfiles anyway. | ||
| 422 : | |||
| 423 : | ---------------------------------------------------------------------- | ||
| 424 : | leunga | 591 | Name: Allen Leung |
| 425 : | Date: 2000/04/02 21:17:00 EST | ||
| 426 : | Tag: leunga-20000402-mltree | ||
| 427 : | Description: | ||
| 428 : | leunga | 585 | |
| 429 : | leunga | 591 | 1. Renamed the constructor CALL in MLTREE by popular demand. |
| 430 : | 2. Added a bunch of files from my repository. These are currently | ||
| 431 : | used by other non-SMLNJ backends. | ||
| 432 : | |||
| 433 : | leunga | 576 | ---------------------------------------------------------------------- |
| 434 : | leunga | 590 | Name: Allen Leung |
| 435 : | Date: 2000/03/31 21:15:00 EST | ||
| 436 : | Tag: leunga-20000331-aliasing | ||
| 437 : | Description: | ||
| 438 : | |||
| 439 : | This update contains a rewritten (and hopefully more correct) module | ||
| 440 : | for extracting aliasing information from CPS. | ||
| 441 : | |||
| 442 : | To turn on this feature: | ||
| 443 : | |||
| 444 : | Compiler.Control.CG.memDisambiguate := true | ||
| 445 : | |||
| 446 : | To pretty print the region information with assembly | ||
| 447 : | |||
| 448 : | Compiler.Control.MLRISC.getFlag "asm-show-region" := true; | ||
| 449 : | |||
| 450 : | To control how many levels of aliasing information are printed, use: | ||
| 451 : | |||
| 452 : | Compiler.Control.MLRISC.getInt "points-to-show-level" := n | ||
| 453 : | |||
| 454 : | The default of n is 3. | ||
| 455 : | |||
| 456 : | ---------------------------------------------------------------------- | ||
| 457 : | dbm | 589 | Name: David MacQueen |
| 458 : | Date: 2000/03/31 11:15:00 EST | ||
| 459 : | Tag: dbm-20000331-runtime_fix | ||
| 460 : | Description: | ||
| 461 : | |||
| 462 : | This update contains: | ||
| 463 : | |||
| 464 : | 1. runtime/c-lib/c-libraries.c | ||
| 465 : | includes added in revision 1.2 caused compilation errors on hppa-hpux | ||
| 466 : | |||
| 467 : | 2. fix for bug 1556 | ||
| 468 : | system/Basis/Implementation/NJ/internal-signals.sml | ||
| 469 : | |||
| 470 : | ---------------------------------------------------------------------- | ||
| 471 : | blume | 587 | Name: Matthias Blume |
| 472 : | blume | 588 | Date: 2000/03/31 18:00:00 JST |
| 473 : | Tag: blume_main_v110p26p2_1 | ||
| 474 : | Description: | ||
| 475 : | |||
| 476 : | This update contains: | ||
| 477 : | |||
| 478 : | 1. A small change to CM's handling of stable libraries: | ||
| 479 : | CM now maintains one "global" modmap that is used for all stable | ||
| 480 : | libraries. The use of such a global modmap maximizes sharing and | ||
| 481 : | minimizes the need for re-traversing parts of environments during | ||
| 482 : | modmap construction. (However, this has minor impact since modmap | ||
| 483 : | construction seems to account for just one percent or less of total | ||
| 484 : | compile time.) | ||
| 485 : | |||
| 486 : | 2. I added a "genmap" phase to the statistics. This is where I got the | ||
| 487 : | "one percent" number (see above). | ||
| 488 : | |||
| 489 : | 3. CM's new tool parameter mechanism just became _even_ better. :) | ||
| 490 : | - The parser understands named parameters and recursive options. | ||
| 491 : | - The "make" and "shell" tools use these new features. | ||
| 492 : | (This makes it a lot easier to cascade these tools.) | ||
| 493 : | - There is a small syntax change: named parameters use a | ||
| 494 : | |||
| 495 : | <name> : ( <option> ... ) or | ||
| 496 : | <name> : <string> | ||
| 497 : | |||
| 498 : | syntax. Previously, named parameters were implemented in an | ||
| 499 : | ad-hoc fashion by each tool individually (by parsing strings) | ||
| 500 : | and had the form | ||
| 501 : | |||
| 502 : | <name>=<string> | ||
| 503 : | |||
| 504 : | See the CM manual for a full description of these issues. | ||
| 505 : | |||
| 506 : | ---------------------------------------------------------------------- | ||
| 507 : | Name: Matthias Blume | ||
| 508 : | blume | 587 | Date: 2000/03/30 18:00:00 JST |
| 509 : | Tag: blume_main_v110p26p2_0 | ||
| 510 : | Description: | ||
| 511 : | |||
| 512 : | !!!!! WARNING !!!!!! | ||
| 513 : | !! New binfiles !! | ||
| 514 : | !!!!!!!!!!!!!!!!!!!! | ||
| 515 : | |||
| 516 : | This update contains: | ||
| 517 : | |||
| 518 : | 1. Moderate changes to CM: | ||
| 519 : | |||
| 520 : | - Changes to CM's tools mechanism. In particular, it is now possible | ||
| 521 : | to have tools that accept additional "command line" parameters | ||
| 522 : | (specified in the .cm file at each instance where the tool's class is | ||
| 523 : | used). | ||
| 524 : | |||
| 525 : | This was done to accomodate the new "make" and "shell" tools which | ||
| 526 : | facilitate fairly seemless hookup to portions of code managed using | ||
| 527 : | Makefiles or Shell scripts. | ||
| 528 : | |||
| 529 : | There are no classes "shared" or "private" anymore. Instead, the | ||
| 530 : | sharing annotation is now a parameter to the "sml" class. | ||
| 531 : | |||
| 532 : | There is a bit of generic machinery for implementing one's own | ||
| 533 : | tools that accept command-line parameters. However, I am not yet fully | ||
| 534 : | satisfied with that part, so expect changes here in the future. | ||
| 535 : | |||
| 536 : | All existing tools are described in the CM manual. | ||
| 537 : | |||
| 538 : | - Slightly better error handling. (CM now surpresses many followup | ||
| 539 : | error messages that tended to be more annoying than helpful.) | ||
| 540 : | |||
| 541 : | 2. Major changes to the compiler's static environment data structures. | ||
| 542 : | |||
| 543 : | - no CMStaticEnv anymore. | ||
| 544 : | - no CMEnv, no "BareEnvironment" (actually, _only_ BareEnvironment, | ||
| 545 : | but it is called Environment), no conversions between different | ||
| 546 : | kinds of static environments | ||
| 547 : | |||
| 548 : | - There is still a notion of a "modmap", but such modmaps are generated | ||
| 549 : | on demand at the time when they are needed. This sounds slow, but I | ||
| 550 : | sped up the code that generates modmaps enough for this not to lead to | ||
| 551 : | a slowdown of the compiler (at least I didn't detect any). | ||
| 552 : | |||
| 553 : | - To facilitate rapid modmap generation, static environments now | ||
| 554 : | contain an (optional) "modtree" structure. Modtree annotations are | ||
| 555 : | constructed by the unpickler during unpickling. (This means that | ||
| 556 : | the elaborator does not have to worry about modtrees at all.) | ||
| 557 : | Modtrees have the advantage that they are compositional in the same | ||
| 558 : | way as the environment data structure itself is compositional. | ||
| 559 : | As a result, modtrees never hang on to parts of an environment that | ||
| 560 : | has already been rendered "stale" by filtering or rebinding. | ||
| 561 : | |||
| 562 : | - I went through many, many trials and errors before arriving at the | ||
| 563 : | current solution. (The initial idea of "linkpaths" did not work.) | ||
| 564 : | But the result of all this is that I have touched a lot of files that | ||
| 565 : | depend on the "modules" and "types" data structures (most of the | ||
| 566 : | elaborator). There were a lot of changes during my "linkpath" trials | ||
| 567 : | that could have been reverted to their original state but weren't. | ||
| 568 : | Please, don't be too harsh on me for messing with this code a bit more | ||
| 569 : | than what was strictly necessary... (I _did_ resist the tempation | ||
| 570 : | of doing any "global reformatting" to avoid an untimely death at | ||
| 571 : | Dave's hands. :) | ||
| 572 : | |||
| 573 : | - One positive aspect of the previous point: At least I made sure that | ||
| 574 : | all files that I touched now compile without warnings (other than | ||
| 575 : | "polyEqual"). | ||
| 576 : | |||
| 577 : | - compiler now tends to run "leaner" (i.e., ties up less memory in | ||
| 578 : | redundant modmaps) | ||
| 579 : | |||
| 580 : | ---------------------------------------------------------------------- | ||
| 581 : | leunga | 580 | Name: Allen Leung |
| 582 : | leunga | 585 | Date: 2000/03/29 18:00:00 |
| 583 : | Tag: leunga-20000327-mlriscGen_hppa_alpha_x86 | ||
| 584 : | leunga | 586 | Boot files (optional): ftp://react-ilp.cs.nyu.edu/leunga/110.26.1-sml.boot.x86-unix-20000330.tar.gz |
| 585 : | leunga | 585 | Description: |
| 586 : | |||
| 587 : | This update contains *MAJOR* changes to the way code is generated from CPS | ||
| 588 : | in the module mlriscGen, and in various backend modules. | ||
| 589 : | |||
| 590 : | CHANGES | ||
| 591 : | ======= | ||
| 592 : | |||
| 593 : | 1. MLRiscGen: forward propagation fix. | ||
| 594 : | |||
| 595 : | There was a bug in forward propagation introduced at about the same time | ||
| 596 : | as the MLRISC x86 backend, which prohibits coalescing to be | ||
| 597 : | performed effectively in loops. | ||
| 598 : | |||
| 599 : | Effect: speed up of loops in RISC architectures. | ||
| 600 : | By itself, this actually slowed down certain benchmarks on the x86. | ||
| 601 : | |||
| 602 : | 2. MLRiscGen: forward propagating addresses from consing. | ||
| 603 : | |||
| 604 : | I've changed the way consing code is generated. Basically I separated | ||
| 605 : | out the initialization part: | ||
| 606 : | |||
| 607 : | store tag, offset(allocptr) | ||
| 608 : | store elem1, offset+4(allocptr) | ||
| 609 : | store elem2, offset+8(allocptr) | ||
| 610 : | ... | ||
| 611 : | store elemn, offset+4n(allocptr) | ||
| 612 : | |||
| 613 : | and the address computation part: | ||
| 614 : | |||
| 615 : | celladdr <- offset+4+alloctpr | ||
| 616 : | |||
| 617 : | and move the address computation part | ||
| 618 : | |||
| 619 : | Effect: register pressure is generally lower as a result. This | ||
| 620 : | makes compilation of certain expressions much faster, such as | ||
| 621 : | long lists with non-trivial elements. | ||
| 622 : | |||
| 623 : | [(0,0), (0,0), .... (0,0)] | ||
| 624 : | |||
| 625 : | 3. MLRiscGen: base pointer elimination. | ||
| 626 : | |||
| 627 : | As part of the linkage mechanism, we generate the sequence: | ||
| 628 : | |||
| 629 : | L: ... <- start of the code fragment | ||
| 630 : | |||
| 631 : | L1: | ||
| 632 : | base pointer <- linkreg - L1 + L | ||
| 633 : | |||
| 634 : | The base pointer was then used for computing relocatable addresses | ||
| 635 : | in the code fragment. Frequently (such as in lots of continuations) | ||
| 636 : | this is not needed. We now eliminate this sequence whenever possible. | ||
| 637 : | |||
| 638 : | For compile time efficiency, I'm using a very stupid local heuristic. | ||
| 639 : | But in general, this should be done as a control flow analysis. | ||
| 640 : | |||
| 641 : | Effect: Smaller code size. Speed up of most programs. | ||
| 642 : | |||
| 643 : | 4. Hppa back end | ||
| 644 : | |||
| 645 : | Long jumps in span dependence resolution used to depend on the existence | ||
| 646 : | of the base pointer. | ||
| 647 : | |||
| 648 : | A jump to a long label L was expanded into the following sequence: | ||
| 649 : | |||
| 650 : | LDIL %hi(L-8192), %r29 | ||
| 651 : | LDO %lo(L-8192)(%r29), %r29 | ||
| 652 : | ADD %r29, baseptr, %r29 | ||
| 653 : | BV,n %r0(%r29) | ||
| 654 : | |||
| 655 : | In the presence of change (3) above, this will not work. I've changed | ||
| 656 : | it so that the following sequence of instructions are generated, which | ||
| 657 : | doesn't mention the base pointer at all: | ||
| 658 : | |||
| 659 : | BL,n L', %r29 /* branch and link, L' + 4 -> %r29 */ | ||
| 660 : | L': ADDIL L-(L'+4), %r29 /* Compute address of L */ | ||
| 661 : | BV,n %r0(%r29) /* Jump */ | ||
| 662 : | |||
| 663 : | 5. Alpha back end | ||
| 664 : | |||
| 665 : | New alpha instructions LDB/LDW have been added, as per Fermin's | ||
| 666 : | suggestions. This is unrelated to all other changes. | ||
| 667 : | |||
| 668 : | 6. X86 back end | ||
| 669 : | |||
| 670 : | I've changed andl to testl in the floating point test sequence | ||
| 671 : | whenever appropriate. The Intel optimization guide states that | ||
| 672 : | testl is perferable to andl. | ||
| 673 : | |||
| 674 : | 7. RA (x86 only) | ||
| 675 : | |||
| 676 : | I've improved the spill propagation algorithm, using an approximation | ||
| 677 : | of maximal weighted independent sets. This seems to be necessary to | ||
| 678 : | alleviate the negative effect in light of the slow down in (1). | ||
| 679 : | |||
| 680 : | I'll write down the algorithm one of these days. | ||
| 681 : | |||
| 682 : | 8. MLRiscGen: frequencies | ||
| 683 : | |||
| 684 : | I've added an annotation that states that all call gc blocks have zero | ||
| 685 : | execution frequencies. This improves register allocation on the x86. | ||
| 686 : | |||
| 687 : | BENCHMARKS | ||
| 688 : | ========== | ||
| 689 : | |||
| 690 : | I've only perform the comparison on 110.25. | ||
| 691 : | |||
| 692 : | The platforms are: | ||
| 693 : | |||
| 694 : | HPPA A four processor HP machine (E9000) with 5G of memory. | ||
| 695 : | X86 A 300Hhz Pentium II with 128M of memory, and | ||
| 696 : | SPARC An Ultra sparc 2 with 512M of memory. | ||
| 697 : | |||
| 698 : | I used the following parameters for the SML benchmarks: | ||
| 699 : | |||
| 700 : | @SMLalloc | ||
| 701 : | HPPA 256k | ||
| 702 : | SPARC 512k | ||
| 703 : | X86 256k | ||
| 704 : | |||
| 705 : | COMPILATION TIME | ||
| 706 : | ---------------- | ||
| 707 : | Here are the numbers comparing the compilation times of the compilers. | ||
| 708 : | I've only compared 110.25 compiling the new sources versus | ||
| 709 : | a fixpoint version of the new compiler compiling the same. | ||
| 710 : | |||
| 711 : | 110.25 New | ||
| 712 : | Total Time in RA Spill+Reload Total Time In RA Spill+Reload | ||
| 713 : | HPPA 627s 116s 2684+3584 599s 95s 1003+1879 | ||
| 714 : | SPARC 892s 173s 2891+3870 708s 116s 1004+1880 | ||
| 715 : | X86 999s 315s 94006+130691 987s 296s 108877+141957 | ||
| 716 : | |||
| 717 : | 110.25 New | ||
| 718 : | Code Size Code Size | ||
| 719 : | HPPA 8596736 8561421 | ||
| 720 : | SPARC 8974299 8785143 | ||
| 721 : | X86 9029180 8716783 | ||
| 722 : | |||
| 723 : | So in summary, things are at least as good as before. Dramatic | ||
| 724 : | reduction in compilation is obtained on the Sparc; I can't explain it, | ||
| 725 : | but it is reproducible. Perhaps someone should try to reproduce this | ||
| 726 : | on their own machines. | ||
| 727 : | |||
| 728 : | SML BENCHMARKS | ||
| 729 : | -------------- | ||
| 730 : | |||
| 731 : | On the average, all benchmarks perform at least as well as before. | ||
| 732 : | |||
| 733 : | HPPA Compilation Time Spill+Reload Run Time | ||
| 734 : | 110.25 New 110.25 New 110.25 New | ||
| 735 : | |||
| 736 : | barnesHut 3.158 3.015 4.75% 1+1 0+0 2.980 2.922 2.00% | ||
| 737 : | boyer 6.152 5.708 7.77% 0+0 0+0 0.218 0.213 2.34% | ||
| 738 : | count-graphs 1.168 1.120 4.32% 0+0 0+0 22.705 23.073 -1.60% | ||
| 739 : | fft 0.877 0.792 10.74% 1+3 1+3 0.602 0.587 2.56% | ||
| 740 : | knuthBendix 3.180 2.857 11.32% 0+0 0+0 0.675 0.662 2.02% | ||
| 741 : | lexgen 6.190 5.290 17.01% 0+0 0+0 0.913 0.788 15.86% | ||
| 742 : | life 0.803 0.703 14.22% 25+25 0+0 0.153 0.140 9.52% | ||
| 743 : | logic 2.048 2.007 2.08% 6+6 1+1 4.133 4.008 3.12% | ||
| 744 : | mandelbrot 0.077 0.080 -4.17% 0+0 0+0 0.765 0.712 7.49% | ||
| 745 : | mlyacc 22.932 20.937 9.53% 154+181 32+57 0.468 0.430 8.91% | ||
| 746 : | nucleic 5.183 5.060 2.44% 2+2 0+0 0.125 0.120 4.17% | ||
| 747 : | ratio-regions 3.357 3.142 6.84% 0+0 0+0 116.225 113.173 2.70% | ||
| 748 : | ray 1.283 1.290 -0.52% 0+0 0+0 2.887 2.855 1.11% | ||
| 749 : | simple 6.307 6.032 4.56% 28+30 5+7 3.705 3.658 1.28% | ||
| 750 : | tsp 0.888 0.862 3.09% 0+0 0+0 7.040 6.893 2.13% | ||
| 751 : | vliw 24.378 23.455 3.94% 106+127 25+45 2.758 2.707 1.91% | ||
| 752 : | -------------------------------------------------------------------------- | ||
| 753 : | Average 6.12% 4.09% | ||
| 754 : | |||
| 755 : | SPARC Compilation Time Spill+Reload Run Time | ||
| 756 : | 110.25 New 110.25 New 110.25 New | ||
| 757 : | |||
| 758 : | barnesHut 3.778 3.592 5.20% 2+2 0+0 3.648 3.453 5.65% | ||
| 759 : | boyer 6.632 6.110 8.54% 0+0 0+0 0.258 0.242 6.90% | ||
| 760 : | count-graphs 1.435 1.325 8.30% 0+0 0+0 33.672 34.737 -3.07% | ||
| 761 : | fft 0.980 0.940 4.26% 3+9 2+6 0.838 0.827 1.41% | ||
| 762 : | knuthBendix 3.590 3.138 14.39% 0+0 0+0 0.962 0.967 -0.52% | ||
| 763 : | lexgen 6.593 6.072 8.59% 1+1 0+0 1.077 1.078 -0.15% | ||
| 764 : | life 0.972 0.868 11.90% 26+26 0+0 0.143 0.140 2.38% | ||
| 765 : | logic 2.525 2.387 5.80% 7+7 1+1 5.625 5.158 9.05% | ||
| 766 : | mandelbrot 0.090 0.093 -3.57% 0+0 0+0 0.855 0.728 17.39% | ||
| 767 : | mlyacc 26.732 23.827 12.19% 162+189 32+57 0.550 0.560 -1.79% | ||
| 768 : | nucleic 6.233 6.197 0.59% 3+3 0+0 0.163 0.173 -5.77% | ||
| 769 : | ratio-regions 3.780 3.507 7.79% 0+0 0+0 133.993 131.035 2.26% | ||
| 770 : | ray 1.595 1.550 2.90% 1+1 0+0 3.440 3.418 0.63% | ||
| 771 : | simple 6.972 6.487 7.48% 29+32 5+7 3.523 3.525 -0.05% | ||
| 772 : | tsp 1.115 1.063 4.86% 0+0 0+0 7.393 7.265 1.77% | ||
| 773 : | vliw 27.765 24.818 11.87% 110+135 25+45 2.265 2.135 6.09% | ||
| 774 : | ---------------------------------------------------------------------------- | ||
| 775 : | Average 6.94% 2.64% | ||
| 776 : | |||
| 777 : | X86 Compilation Time Spill+Reload Run Time | ||
| 778 : | 110.25 New 110.25 New 110.25 New | ||
| 779 : | |||
| 780 : | barnesHut 5.530 5.420 2.03% 593+893 597+915 3.532 3.440 2.66% | ||
| 781 : | boyer 8.768 7.747 13.19% 493+199 301+289 0.327 0.297 10.11% | ||
| 782 : | count-graphs 2.040 2.010 1.49% 298+394 315+457 26.578 28.660 -7.26% | ||
| 783 : | fft 1.327 1.302 1.92% 112+209 115+210 1.055 0.962 9.71% | ||
| 784 : | knuthBendix 5.218 5.475 -4.69% 451+598 510+650 0.928 0.932 -0.36% | ||
| 785 : | lexgen 9.970 9.623 3.60% 1014+841 1157+885 0.947 0.928 1.97% | ||
| 786 : | life 1.183 1.183 0.00% 162+182 145+148 0.127 0.103 22.58% | ||
| 787 : | logic 3.285 3.512 -6.45% 514+684 591+836 5.682 5.577 1.88% | ||
| 788 : | mandelbrot 0.147 0.143 2.33% 38+41 33+54 0.703 0.690 1.93% | ||
| 789 : | mlyacc 35.457 32.763 8.22% 3496+4564 3611+4860 0.552 0.550 0.30% | ||
| 790 : | nucleic 7.100 6.888 3.07% 239+168 201+158 0.175 0.173 0.96% | ||
| 791 : | ratio-regions 6.388 6.843 -6.65% 1182+257 981+300 120.142 120.345 -0.17% | ||
| 792 : | ray 2.332 2.338 -0.29% 346+398 402+494 3.593 3.540 1.51% | ||
| 793 : | simple 9.912 9.903 0.08% 1475+941 1579+1168 3.057 3.178 -3.83% | ||
| 794 : | tsp 1.623 1.532 5.98% 266+200 250+211 8.045 7.878 2.12% | ||
| 795 : | vliw 33.947 35.470 -4.29% 2629+2774 2877+3171 2.072 1.890 9.61% | ||
| 796 : | ---------------------------------------------------------------------------- | ||
| 797 : | Average 1.22% 3.36% | ||
| 798 : | |||
| 799 : | ---------------------------------------------------------------------- | ||
| 800 : | Name: Allen Leung | ||
| 801 : | leunga | 583 | Date: 2000/03/23 16:25:00 |
| 802 : | Tag: leunga-20000323-fix_x86_alpha | ||
| 803 : | Description: | ||
| 804 : | |||
| 805 : | 1. X86 fixes/changes | ||
| 806 : | |||
| 807 : | a. The old code generated for SETcc was completely wrong. | ||
| 808 : | The Intel optimization guide is VERY misleading. | ||
| 809 : | |||
| 810 : | 2. ALPHA fixes/changes | ||
| 811 : | |||
| 812 : | a. Added the instructions LDBU, LDWU, STB, STW as per Fermin's suggestion. | ||
| 813 : | b. Added a new mode byteWordLoadStores to the functor parameter to Alpha() | ||
| 814 : | c. Added reassociation code for address computation. | ||
| 815 : | |||
| 816 : | ---------------------------------------------------------------------- | ||
| 817 : | Name: Allen Leung | ||
| 818 : | leunga | 580 | Date: 2000/03/22 01:23:00 |
| 819 : | Tag: leunga-20000322-fix_x86_hppa_ra | ||
| 820 : | Description: | ||
| 821 : | |||
| 822 : | 1. X86 fixes/changes | ||
| 823 : | |||
| 824 : | a. x86Rewrite bug with MUL3 (found by Lal) | ||
| 825 : | b. Added the instructions FSTS, FSTL | ||
| 826 : | |||
| 827 : | 2. PA-RISC fixes/changes | ||
| 828 : | |||
| 829 : | a. B label should not be a delay slot candidate! Why did this work? | ||
| 830 : | b. ADDT(32, REG(32, r), LI n) now generates one instruction instead of two, | ||
| 831 : | as it should be. | ||
| 832 : | c. The assembly syntax for fstds and fstdd was wrong. | ||
| 833 : | d. Added the composite instruction COMICLR/LDO, which is the immediate | ||
| 834 : | operand variant of COMCLR/LDO. | ||
| 835 : | |||
| 836 : | 3. Generic MLRISC | ||
| 837 : | |||
| 838 : | a. shuffle.sml rewritten to be slightly more efficient | ||
| 839 : | b. DIV bug in mltree-simplify fixed (found by Fermin) | ||
| 840 : | |||
| 841 : | 4. Register Allocator | ||
| 842 : | |||
| 843 : | a. I now release the interference graph earlier during spilling. | ||
| 844 : | May improve memory usage. | ||
| 845 : | |||
| 846 : | ---------------------------------------------------------------------- | ||
| 847 : | blume | 577 | Name: Matthias Blume |
| 848 : | blume | 578 | Date: 2000/03/14 14:15:32 |
| 849 : | Tag: blume_main_v110p26p1_2 | ||
| 850 : | Description: | ||
| 851 : | |||
| 852 : | 1. Tools.registerStdShellCmdTool (from smlnj/cm/tool.cm) takes an | ||
| 853 : | additional argument called "template" which is an optional string that | ||
| 854 : | specifiel the layout of the tool command line. See the CM manual for | ||
| 855 : | explanation. | ||
| 856 : | |||
| 857 : | 2. A special-purpose tool can be "regisitered" by simply dropping the | ||
| 858 : | corresponding <...>-tool.cm (and/or <...>-ext.cm) into the same | ||
| 859 : | directory where the .cm file lives that uses this tool. (The | ||
| 860 : | behavior/misfeature until now was to look for the tool description | ||
| 861 : | files in the current working directory.) As before, tool description | ||
| 862 : | files could also be anchored -- in which case they can live anywhere | ||
| 863 : | they like. Following the recent e-mail discussion, this change should | ||
| 864 : | make it easier to have special-purpose tools that are shipped together | ||
| 865 : | with the sources of the program that uses them. | ||
| 866 : | |||
| 867 : | ---------------------------------------------------------------------- | ||
| 868 : | Name: Matthias Blume | ||
| 869 : | blume | 577 | Date: 2000/03/10 07:48:34 |
| 870 : | Tag: blume_main_v110p26p1_1 | ||
| 871 : | Description: | ||
| 872 : | |||
| 873 : | I added a re-written version of Dave's fixpt script to src/system. | ||
| 874 : | Changes relative to the original version: | ||
| 875 : | - sh-ified (not everybody has ksh) | ||
| 876 : | - automatically figures out which architecture it runs on | ||
| 877 : | - uses ./makeml a bit more cleverly | ||
| 878 : | - never invokes ./installml (and, thus, does not clobber your | ||
| 879 : | good and working installation of sml in case something goes wrong) | ||
| 880 : | - accepts max iteration count using option "-iter <n>" | ||
| 881 : | - accepts a "base" name using option "-base <base>" | ||
| 882 : | |||
| 883 : | It does not build any extraneous heap images but directly rebuilds | ||
| 884 : | bin- and boot-hierarchies using makeml's "-rebuild" switch. Finally, | ||
| 885 : | it can incorporate existing bin- and boot- hierarchies. For example, | ||
| 886 : | suppose the base is set to "sml" (which is the default). Then it | ||
| 887 : | successively builds | ||
| 888 : | |||
| 889 : | sml.bin.<arch>-unix and sml.boot.<arch>-unix | ||
| 890 : | then sml1.bin.<arch>-unix and sml1.boot.<arch>-unix | ||
| 891 : | then sml2.bin.<arch>-unix and sml2.boot.<arch>-unix | ||
| 892 : | ... | ||
| 893 : | then sml<n>.bin.<arch>-unix and sml<n>.boot.<arch>-unix | ||
| 894 : | |||
| 895 : | and so on. If any of these already exist, it will just use what's | ||
| 896 : | there. In particular, many people will have the initial set of bin | ||
| 897 : | and boot files around, so this saves time for at least one full | ||
| 898 : | rebuild. Having sets of the form <base><k>.{bin,boot}.<arch>-unix for | ||
| 899 : | <k>=1,2,... is normally not a good idea when invoking fixpt. However, | ||
| 900 : | they might be the result of an earlier partial run of fixpt (which | ||
| 901 : | perhaps got accidentially killed). In this case, fixpt will quickly | ||
| 902 : | move through what exists before continuing where it left off earlier, | ||
| 903 : | and, thus, saves a lot of time. | ||
| 904 : | |||
| 905 : | ---------------------------------------------------------------------- | ||
| 906 : | leunga | 576 | Name: Allen Leung |
| 907 : | Date: 00/03/10 02:20:00 | ||
| 908 : | Tag: leunga-20000310-fix_x86_asm_ra | ||
| 909 : | Description: | ||
| 910 : | dbm | 570 | |
| 911 : | leunga | 576 | More assembly output problems involving the indexed addressing mode |
| 912 : | on the x86 have been found and corrected. Thanks to Fermin Reig for the | ||
| 913 : | fix. | ||
| 914 : | |||
| 915 : | The interface and implementation of the register allocator have been changed | ||
| 916 : | slightly to accommodate the possibility to skip the register allocation | ||
| 917 : | phases completely and go directly to memory allocation. This is needed | ||
| 918 : | for C-- use. | ||
| 919 : | |||
| 920 : | dbm | 570 | ---------------------------------------------------------------------- |
| 921 : | blume | 572 | Name: Matthias Blume |
| 922 : | blume | 575 | Date: 00/03/09 10:23:53 |
| 923 : | Tag: blume_main_v110p26p1_0 | ||
| 924 : | Description: | ||
| 925 : | |||
| 926 : | * Complete re-organization of library names. Many libraries have been | ||
| 927 : | consolidated so that they share the same path anchor. For example, | ||
| 928 : | all MLRISC-related libraries are anchored at MLRISC, most libraries that | ||
| 929 : | are SML/NJ-specific are under "smlnj". Notice that names like | ||
| 930 : | host-cmb.cm or host-compiler.cm no longer exist. See system/README | ||
| 931 : | for a complete description of the new naming scheme. Quick reference: | ||
| 932 : | |||
| 933 : | host-cmb.cm -> smlnj/cmb.cm | ||
| 934 : | host-compiler.cm -> smlnj/compiler.cm | ||
| 935 : | full-cm.cm -> smlnj/cm.cm | ||
| 936 : | <arch>-<os>.cm -> smlnj/cmb/<arch>-<os>.cm | ||
| 937 : | <arch>-compiler.cm -> smlnj/compiler/<arch>.cm | ||
| 938 : | |||
| 939 : | * Bug fixes in CM. | ||
| 940 : | - exceptions in user code are being passed through (i.e., reach top level) | ||
| 941 : | - more bugs in paranoia mode fixed | ||
| 942 : | - bug related to checking group owners fixed | ||
| 943 : | |||
| 944 : | * New install.sh script that automagically fetches archive files: | ||
| 945 : | The new file config/srcarchiveurl must contain the URL of the | ||
| 946 : | (remote) directory that contains bin files (or other source archives). | ||
| 947 : | If install.sh does not find the archive locally, it tries to get | ||
| 948 : | it from that remote directory. | ||
| 949 : | This should simplify installation further: For machines that have | ||
| 950 : | access to the internet, just fetch <version>-config.tgz, unpack it, | ||
| 951 : | edit config/targets, and go (run config/install.sh). The scipt will | ||
| 952 : | fetch everything else that it might need all by itself. | ||
| 953 : | |||
| 954 : | For CVS users, this mechanism is not relevant for source archives, but | ||
| 955 : | it is convenient for getting new sets of binfiles. | ||
| 956 : | |||
| 957 : | Archives should be tar files compressed with either gzip, compress, or | ||
| 958 : | bzip2. The script recognizes .tgz, .tar, tar.gz, tz, .tar.Z, and .tar.bz2. | ||
| 959 : | |||
| 960 : | ---------------------------------------------------------------------- | ||
| 961 : | Name: Matthias Blume | ||
| 962 : | blume | 572 | Date: 2000/03/07 04:01:04 |
| 963 : | Tag: blume_main_v110_26_2 | ||
| 964 : | dbm | 570 | Description: |
| 965 : | blume | 572 | - size info in BOOTLIST |
| 966 : | * no fixed upper limits for number of bootfiles or length of | ||
| 967 : | bootfile names in runtime | ||
| 968 : | * falling back to old behavior if no BOOTLIST size info found | ||
| 969 : | - allocation size heuristics in .run-sml | ||
| 970 : | * tries to read cache size from /proc/cpuinfo (this is important for | ||
| 971 : | small-cache Celeron systems!) | ||
| 972 : | - install.sh robustified | ||
| 973 : | - CM manual updates | ||
| 974 : | - paranoid mode | ||
| 975 : | * no more CMB.deliver() (i.e., all done by CMB.make()) | ||
| 976 : | * can re-use existing sml.boot.* files | ||
| 977 : | * init.cmi now treated as library | ||
| 978 : | * library stamps for consistency checks | ||
| 979 : | - sml.boot.<arch>-<os>/PIDMAP file | ||
| 980 : | * This file is read by the CM startup code. This is used to minimize | ||
| 981 : | the amount of dynamic state that needs to be stowed away for the | ||
| 982 : | purpose of sharing between interactive system and user code. | ||
| 983 : | - CM.Anchor.anchor instead of CM.Anchor.{set,cancel} | ||
| 984 : | * Upon request by Elsa. Anchors now controlled by get-set-pair | ||
| 985 : | like most other CM state variables. | ||
| 986 : | - Compiler.CMSA eliminated | ||
| 987 : | * No longer supported by CM anyway. | ||
| 988 : | - fixed bugs in pickler that kept biting Stefan | ||
| 989 : | * past refs to past refs (was caused by the possibility that | ||
| 990 : | ad-hoc sharing is more discriminating than hash-cons sharing) | ||
| 991 : | * integer overflow on LargeInt.minInt | ||
| 992 : | - ml-{lex,yacc} build scripts now use new mechanism | ||
| 993 : | for building standalone programs | ||
| 994 : | - fixed several gcc -Wall warnings that were caused by missing header | ||
| 995 : | files, missing initializations, etc., in runtime (not all warnings | ||
| 996 : | eliminated, though) |
| root@smlnj-gforge.cs.uchicago.edu | ViewVC Help |
| Powered by ViewVC 1.0.0 |
