262a46a3f4
compiler.cfg.loop-detection: more use of hash-sets.
2013-03-08 12:38:50 -08:00
9d43e21563
compiler.cfg.write-barrier: more use of hash-sets.
2013-03-08 12:30:37 -08:00
88626d38ca
compiler.cfg.ssa.construction: more use of hash-sets.
2013-03-08 12:24:06 -08:00
b97a1a0753
compiler: fix stack effects to indicate visited is same object.
2013-03-08 12:16:02 -08:00
bf35114fec
compiler: use hash-sets to track visited sets.
2013-03-08 11:04:47 -08:00
5908186301
compiler.cfg.alias-analysis: use clear-set.
2013-03-07 21:35:00 -08:00
63e55ac813
compiler.cfg.liveness: Break the infinite loop--add the base-pointer as f to start with, and replace it with the real one if there's not a circular chain. Fixes #22 again.
2013-01-01 14:47:54 -08:00
125e4befd3
compiler.cfg.liveness: Add parallel-copy method for lookup-base-pointer*. Adjust lookup-base-pointer* stack effect to take two parameters. Still retain stack overflows, but more correct now...
2013-01-01 14:16:51 -08:00
8b44af4717
compiler.cfg.renaming.functor: fix ##parallel-copy methods
2012-12-31 11:26:46 -08:00
e90ce0865c
compiler.cfg.ssa.destruction.leaders: break off leader-map to resolve circular dependencies (oops)
2012-12-31 11:20:57 -08:00
dcb27c51e4
compiler.cfg.liveness: still kill defs by leaders, but keep original
...
vregs in the live-sets so accurate info (e.g., representations) can
still be looked up (issue #22 )
2012-12-31 11:05:05 -08:00
fc91dcf530
compiler.cfg.ssa.destruction: alien-call-insns are too hairy to coalesce (issue #22 )
2012-12-31 11:04:57 -08:00
86a15ad47b
compiler.cfg.liveness: look up leaders (if possible) so liveness analysis doesn't break after leaving SSA (issue #22 )
2012-12-31 11:04:52 -08:00
b08c077023
compiler.cfg.parallel-copy: make sure new vregs used to break cycles in parallel-copy-rep update the leader-map (issue #22 )
2012-12-31 11:04:44 -08:00
ff1c2c293b
compiler.cfg.ssa.interference.live-ranges: fix kill-indices for ##parallel-copy (issue #22 )
2012-12-31 11:04:38 -08:00
ae0b77a0c4
compiler.cfg.ssa.cssa: split the definitions of ##phis so live-ranges don't interfere (issue #22 )
...
See the explanation after the proof of Lemma 1 in "Revisiting Out-of-SSA
Translation for Correctness, Code Quality, and Efficiency" (Boissinot et
al.) for why this is necessary.
2012-12-31 11:04:26 -08:00
a0dd953e33
starting point - Fix bootstrap; I broke it in merging. Disable environment variables because the compiler isn't loaded yet and can't call C ffi functions, and tools.test loads this vocab.
...
Disable intrinsics in cpu.x86.64 for now, since they invoke the compiler
Fix ##branch renaming to ##branch,
To work on this branch:
./factor -include=math -i=boot.image
./factor -run=listener
USE: compiler USE: tools.test save
enable-optimizer
Error message is now:
T{ vregs-shouldn't-interfere f 409 424 }
2012-12-31 11:03:45 -08:00
07227f22f9
WIP
2012-12-31 11:03:33 -08:00
013160df9a
change some "2drop 2drop" to "4drop".
2012-10-23 12:21:30 -07:00
f75f99ecc9
compiler.cfg.copy-prop: kludge to make sure ##phis get updated
2012-09-12 15:14:11 -07:00
4008edfdf8
move compiler.cfg.graphviz & compiler.cfg.gvn from basis to extra, just to keep organized
2012-09-12 15:14:07 -07:00
51b3b1fd87
compiler.cfg.graphviz: tweak code & output a little
2012-09-12 15:14:06 -07:00
d6849da7ec
compiler.cfg.graphviz: refactoring
2012-09-12 15:14:06 -07:00
64b541759e
add crappy experimental code for compiler.cfg.graphviz & compiler.cfg.gvn
2012-09-12 15:14:06 -07:00
b6c069f99a
compiler.cfg.representations.selection: removing 10^ in favor of 10^.
2012-09-05 09:25:26 -07:00
6a183718fa
compiler.cfg: don't need to escape the symbol here.
2012-08-02 15:14:42 -07:00
4e72d80256
Using "same?" in more places.
2012-07-21 10:22:44 -07:00
559b5bfa5b
using the new H{ } make.
2012-07-19 09:50:09 -07:00
42f4dc36b2
namespaces: Rename ``bind`` to ``with-variables``. Update a few places that called ``global [ ] with-variables`` to use ``with-global``.
2012-07-19 00:02:47 -07:00
7c77597b09
issue #358 : fix a couple compiler errors
2012-06-21 08:35:47 -07:00
80385770ef
compiler: The non-optimizing compiler is too slow with smart combinators in the compiler. Speed up bootstrap by not using them.
2011-12-05 14:50:51 -08:00
943596575a
use radix literals
2011-11-23 19:03:40 -08:00
f225bf7f55
compiler.cfg.finalization: reenable scheduling
...
Prolog safepoint appears to work with it enabled
2011-11-23 11:11:27 -08:00
6498bc65ce
compiler: emit safepoints in optimized prologs
2011-11-23 11:11:25 -08:00
248066c710
code beautification
2011-11-13 16:10:26 -08:00
873208f986
compiler: more peephole tests for shl/sar
2011-11-13 16:10:24 -08:00
d79b462f75
compiler: add intrinsic for PMOVMSKB/MOVMSKP[SD]
...
Combined with a fast bit-count this will let us rice byte-counting.
2011-11-13 16:10:20 -08:00
910748819d
Now that #foo and ##foo are symbols we can remove a bunch of \
2011-11-12 22:04:26 -08:00
6cca0ea468
compiler.cfg: Rename <##foo> to ##foo, in the low-level IR
2011-11-11 19:48:38 -08:00
eb2a0c611b
compiler.cfg: Change low-level IR constructors from ##foo to <##foo>
2011-11-06 23:02:46 -08:00
c731dc6edb
Rename lookup to lookup-word.
2011-11-06 16:00:00 -08:00
70056d756b
compiler.cfg.builder: add test to ensure that fast union predicates do not contain a conditional
2011-10-29 17:11:04 -07:00
c06be95816
compiler: no spurious safepoint at start of loops
2011-10-27 21:14:48 -07:00
41c36752fc
compiler.cfg.tco: clean up commented-out old code
2011-10-27 21:14:47 -07:00
6f75e84d93
compiler.cfg.tco: teach TCO about safepoints
2011-10-27 21:14:46 -07:00
e449fef4e3
compiler.cfg.builder: safepoint loops and epilogs
2011-10-27 21:14:46 -07:00
0c022f2454
compiler: add ##safepoint IR insn
2011-10-27 21:14:46 -07:00
51c033a1f6
Fix all failures in the linux64 build email
2011-10-24 19:33:09 -07:00
f9257959fd
Rename class to class-of
2011-10-24 06:47:42 -05:00
f4a6e10818
More cleanup of unused stuff.
2011-10-19 11:01:16 -07:00
378786599d
Cleanup lint warnings.
2011-10-14 11:47:24 -07:00
1987deb359
Fix M: stack effects.
2011-10-13 16:41:17 -07:00
4baf14972f
sequences: adding ?last for symmetry.
2011-10-13 13:04:24 -07:00
dc36fd712d
compiler.cfg.representations: fix unit test, it depended on hashing order which changed recently
2011-10-10 12:15:41 -07:00
06218d5d12
compiler.cfg.builder.alien: fix compilation of callbacks which unconditionally throw errors
2011-10-02 19:52:12 -07:00
671f19d70f
Remove all non-core uses of (scan-token)
...
Add scan-datum
Add scan-number
Document more words
Fixes #225 .
2011-10-02 12:00:08 -07:00
76580da5d5
Refactor the lexer/parser to expose friendlier words for scanning tokens. The preferred top-level words now throw an exception on EOF.
...
CREATE -> scan-new
CREATE-CLASS -> scan-new-class
CREATE-WORD -> scan-new-word
CREATE-GENERIC -> scan-new-generic
scan -> (scan-token)
scan-token now throws on eof
(scan-word) returns word/number/f
scan-word now throws on eof
scan-word-name expects a non-number
Fixes #183 .
Fixes #209 .
2011-09-29 11:28:28 -07:00
3826febfae
Fixes the symptom but not the problem. See bug #201 .
2011-09-26 17:32:34 -07:00
2bd6de9a02
slots: pre-create the "at", "nth", and "global" slots to make deterministic the currently nondeterministic ambiguity between accessors:change-* and assocs:change-at, sequences:change-nth, and namespaces:change-global
2011-09-19 10:03:43 -07:00
d6668caf7d
compiler.cfg.builder.alien: fix caller-linkage to return the DLL and not the library tuple again; fixes issue #121
2011-09-13 15:10:48 -07:00
5703e8d7a1
alien.libraries, compiler.cfg.builder.alien: if `dlopen` fails during `<library>`, call `dlerror` and store the error message in the library object; put the dlerror message in the linkage-error when a word from the library is compiled
2011-09-12 23:25:59 -07:00
b6e0f0180b
alien.libraries, compiler.cfg.builder.alien: include the result of dlerror/GetLastError in the linkage-error for missing symbols
2011-09-12 23:25:59 -07:00
edddffdadd
compiler.cfg.dependence: Add ##allot to control dependency chain, so that we don't re-order an allocation and initialization with an alien call. The alien call might GC, exposing uninitialized memory to the GC. Fixes #53 , fixes #58 , fixes #20 .
2011-08-28 21:25:39 -07:00
34f71ea042
compiler.cfg.debugger: add random-scheduling word.
...
Run 'random-scheduling "compiler" test' to find bugs in compiler.cfg.dependencies.
2011-08-28 21:25:37 -07:00
3ed1910ab1
compiler.cfg.scheduling: remove old scheduling test code, adding a better way of testing scheduling
2011-08-28 21:25:35 -07:00
64252dbdbc
32 and 64 bit Linux PPC support
2011-05-23 23:36:14 -04:00
5a3c5c7749
compiler.cfg.branch-splitting was totally broken
2011-02-27 16:43:26 -08:00
6f7e4e32d0
compiler.cfg.ssa.construction: update unit tests for nths change
2011-02-25 21:05:38 -08:00
b052aa61cb
compiler.cfg.linear-scan: extend lifetime intervals of base pointers correctly (reported by Anton Gorenko)
2010-09-30 20:49:03 -07:00
6b5fffc026
compiler: preliminary implementation of tracking derived pointers in GC maps
2010-09-27 20:12:44 -07:00
53aed0805a
compiler.cfg.alias-analysis: fix bug in lazy alias class instantiation
2010-09-27 20:12:43 -07:00
546b81b697
compiler.cfg.scheduling: always run scheduling because heuristic was broken
2010-09-25 16:31:42 -07:00
80d7aab40b
compiler.cfg.alias-analysis: don't need to do a local live-in calculation anymore
2010-09-25 16:13:17 -07:00
ec2f42fd40
compiler.cfg.liveness: merge in compiler.cfg.liveness.ssa and simplify the code, since we don't compute live sets before SSA construction anymore
2010-09-25 14:36:58 -07:00
8bc2ea7a5c
compiler.cfg.ssa.construction: use the baller method for pruned SSA
2010-09-20 21:36:22 -07:00
a67c7c5739
compiler.cfg.write-barrier: fix soundness issue with ##copy that never came up but probably should've been fixed anyway
2010-09-06 17:01:44 -07:00
a605d5c9c9
compiler.cfg.alias-analysis: dead store elimination was too aggressive, can't eliminate dead stores across a GC call even for fresh allocations because GC will see uninitialized data
2010-09-06 16:57:56 -07:00
add75411a0
compiler.cfg.write-barrier: handle ##copy instructions
2010-09-05 23:07:30 -07:00
54f97557e6
compiler.cfg.dependence: compute dependency graph for write barrier instructions
2010-09-05 22:51:17 -07:00
c677c35de4
compiler.cfg: fix major facepalm with write barrier elimination
2010-09-05 21:39:45 -07:00
1985705413
compiler: now that FFI has been deconcatenatized, we no longer need the special ##unary-float-function and ##binary-float-function fastpaths
2010-08-13 23:59:19 -07:00
b5fc39c198
compiler.cfg: Reading the return value of a float-returning function on x86-32 had a side effect of popping the x87 stack, so it was not correct for DCE to just eliminate this if the return value was not used. Fix this by adding a new dead-outputs slot to alien-call-insns and having DCE move dead returns there
2010-08-13 23:19:56 -07:00
42b858b3f0
math.libm: fix regression: fsqrt intrinsic was not working ever since change was made to inline FUNCTION: bodies
2010-08-12 21:41:57 -07:00
94f6ac98de
compiler.cfg.linear-scan: fix bad interaction between inactive intervals and sync points
2010-07-29 21:06:48 -04:00
168dd1f825
FFI rewrite part 7: compile callback bodies with the optimizing compiler
2010-07-28 00:49:26 -04:00
5fc9aa05b0
Revert "compiler.cfg: change linear ordering to place GC call blocks at the end"
...
This reverts commit 544a288192
2010-07-27 23:58:41 -04:00
8adde5360a
compiler.cfg: change linear ordering to place GC call blocks at the end
2010-07-27 13:00:28 -04:00
355d89e8e8
compiler.cfg: now that kill-blocks cannot contain instructions that define vregs we can skip them all
2010-07-27 12:40:31 -04:00
b6fe62299e
compiler.cfg.alias-analysis: factor-call-insns which defined values were not handled properly
2010-07-19 19:56:00 -04:00
b23aac1beb
compiler.cfg: open-code parameter boxing and unboxing for certain C types
2010-07-19 10:25:13 -04:00
8e46305288
compiler.cfg.save-contexts: don't insert ##save-context in front of ##phi
2010-07-19 09:27:10 -04:00
adc9db3ea8
compiler.cfg.linear-scan.allocation: fix wrong order
2010-07-19 06:46:50 -04:00
b776a92546
compiler.cfg.linear-scan: fix bad interaction between split position calculation and register-clobbering instructions
2010-07-18 17:38:52 -04:00
e3edb2653d
compiler.cfg.intrinsics.allot: clean up
2010-07-16 20:04:03 -04:00
e5dd21b0b9
compiler.cfg: nuke ##allot-byte-array instruction
2010-07-16 19:57:45 -04:00
60ddbd9d9b
compiler.cfg.scheduling: ##dispatch must go at the end since its a control flow transfer
2010-07-15 20:55:31 -04:00
0fd636b4b9
compiler.cfg: ##unbox-long-long can have multiple outputs now, clean up long long parameter passing code using this
2010-07-15 19:49:29 -04:00
48e96ef032
compiler.cfg.scheduling: update to support multiple-output instructions
2010-07-15 17:38:34 -04:00
e27adb2830
compiler: re-architect low-level optimizer to allow more than one output value per instruction
2010-07-13 07:40:14 -04:00
a55c8ee671
FFI rewrite part 6: deconcatenatize
2010-07-02 15:44:12 -04:00
John Benediktsson
Doug Coleman
Alex Vondrak
Slava Pestov
Alex Vondrak
Joe Groff
Slava Pestov
Erik Charlebois
Slava Pestov
Slava Pestov
Slava Pestov