6d04bf9c4a
Linear scan: build live ranges
2009-06-03 03:22:27 -05:00
9da8cfe942
Merge branch 'master' of git://factorcode.org/git/factor
2009-06-02 18:30:25 -05:00
2d231f066a
GC checks now save and restore registers
2009-06-02 18:23:47 -05:00
19d6285035
fix load error
2009-06-01 09:34:28 -05:00
7dd08892b2
Fix for recent word renaming
2009-06-01 03:19:35 -05:00
9e987e8642
Merge branch 'master' into global_optimization
2009-06-01 03:12:32 -05:00
d558545b65
Remove obsolete optimization
2009-06-01 03:05:49 -05:00
3a9922d161
Fix compiler errors
2009-06-01 03:00:10 -05:00
2c8223fdaf
Fix loop handling in stack-analysis
2009-05-31 19:04:26 -05:00
32f17f3e14
Fix scoping issue in compiler.cfg.linear-scan.assignment
2009-05-31 18:21:23 -05:00
1a52414bb1
Rename _gc to ##gc
2009-05-31 18:21:11 -05:00
692b479302
Split off local-optimization combinator into compiler.cfg.local, factor out CFG -> MR into compiler.cfg.mr, split off GC check insertion into a new compiler.cfg.gc-checks pass
2009-05-31 12:20:46 -05:00
3e00dc8c8d
Start cleaning up stack analysis
2009-05-30 13:22:30 -05:00
e04df76f60
Various codegen improvements:
...
- new-insn word to construct instructions
- cache RPO in the CFG
- re-organize low-level optimizer so that MR is built after register allocation
- register allocation now stores instruction numbers in the instructions themselves
- split defs-vregs into defs-vregs and temp-vregs
2009-05-29 13:11:34 -05:00
280736ab00
On second thought, linearization will be done after SSA destruction so live-in is accurate
2009-05-29 05:45:40 -05:00
743550f19c
Put GC checks in the right place when linearizing, and generate _dispatch-labels
2009-05-29 05:36:04 -05:00
76d74c16af
Fixing various bugs; alias analysis wasn't handling ##phi nodes, stack analysis incorrectly handled height-changing back edges and ##fixnum-*, clean up ##dispatch generation
2009-05-29 01:39:14 -05:00
ceb332f596
compiler.cfg.alias-analysis: ##peek needs to set alias class of output value
2009-05-28 18:19:01 -05:00
dbf18927b9
compiler.cfg.liveness: correct handling of phi nodes
2009-05-28 17:54:27 -05:00
fc5587bda3
Test updates
2009-05-28 02:50:07 -05:00
ecece1d08b
CFG checker now checks consistency of successors and predecessors lists; fix long-standing bug in useless-blocks optimization
2009-05-28 02:49:51 -05:00
1fa465d77f
Fix alias analysis
2009-05-28 02:48:58 -05:00
117cb78d6b
Fixing local optimizations
2009-05-27 18:58:41 -05:00
f3688c9373
Better CFG checker
2009-05-27 18:58:14 -05:00
dadb9a2c50
Add phi elimination pass
2009-05-27 18:58:01 -05:00
3b79d61496
Add a new ##allocation union to remove some code duplication
2009-05-27 18:55:49 -05:00
2025823ba6
Remove old local DCE pass
2009-05-27 17:55:15 -05:00
ce25e0ad8d
New local-optimization combinator removes some boilerplate
2009-05-26 19:56:56 -05:00
1db81da264
Refactoring low-level optimizer to support stack analysis pass
2009-05-26 19:31:19 -05:00
eda44f28a6
Move maybe-set-at to assocs
2009-05-26 18:45:37 -05:00
ccb662c60e
Fix another bug in stack-analysis
2009-05-26 03:42:39 -05:00
7ea4e255fb
compiler.cfg.stack-analysis: make it pass more tests
2009-05-26 02:59:14 -05:00
8b022f926c
compiler.cfg.dce: new global dead code elimination pass
2009-05-26 02:58:57 -05:00
e0841fa695
compiler.cfg.checker: new check-rpo word
2009-05-26 02:58:40 -05:00
a08bbde2e7
compiler.cfg.stack-analysis: progress
2009-05-25 19:18:13 -05:00
dead771b3f
compiler.cfg.checker: new vocabulary for checking CFG invariants
2009-05-25 19:16:58 -05:00
a2b982e247
compiler.cfg.builder: don't make basic blocks after terminating calls
2009-05-25 19:16:36 -05:00
407377fc98
rename peek -> last and update all usages
2009-05-25 16:38:33 -05:00
145f1dbeef
compiler.cfg.stack-analysis: change how inc-d/inc-r work
2009-05-23 15:47:08 -05:00
6af61656f3
CFG optimizer work in progress - adding phi nodes
2009-05-21 16:49:28 -05:00
e58fcd485c
Working on global optimizations
2009-05-19 17:28:13 -05:00
fb6df472a2
Cleaning up USING: lists for new strict semantics
2009-05-14 16:54:16 -05:00
fc4894fbdf
Replace ratio and complex built-in types with tuples defined in the library. This frees up two lo-tags, so move array and quotation over to these tags and update compiler for new tags
2009-04-30 00:27:35 -05:00
7f766ab355
Update compiler tests for tag number changes
2009-04-28 21:43:30 -05:00
057f75e9a1
Refactor compiler.tree.builder to fix various regressions
2009-04-21 23:02:00 -05:00
05f3f9dcb9
Fixing unit tests for stack effect inference changes
2009-04-20 21:15:19 -05:00
37e278ed02
Use [ ] [ ] map-reduce instead of unclip [ ] reduce
2009-04-06 17:50:54 -05:00
ea60f8ae93
Changing : foo ; parsing to SYNTAX: foo ;
2009-03-21 01:27:50 -05:00
ca7a155b72
Merge branch 'master' into new_ui
2009-03-01 20:13:05 -06:00
03cb32ad9c
Make execute( faster, add execute-unsafe( and make effect tuple slots read only
2009-03-01 20:12:35 -06:00
ce1bc1d6ed
Fixing conflicts from stack checker changes
2009-02-24 01:21:10 -06:00
421ee726b8
Merge branch 'master' into new_ui
2009-02-16 21:22:12 -06:00
6b25e99470
Add summary for heaps more vocabs
2009-02-16 21:05:13 -06:00
634435d816
Merge branch 'master' into new_ui
2009-02-15 04:07:51 -06:00
cddb1f6133
Fix spelling
2009-02-15 04:07:05 -06:00
669548e62e
Clean up some usages of tuck, and add Joe's curried cleave/spread/apply combinators to kernel vocabulary
2009-02-02 13:43:54 -06:00
3c408342ef
renaming: contain? -> any?, deep-contains? -> deep-any?, pad-left -> pad-head, pad-right -> pad-tail, trim-left -> trim-head, trim-right -> trim-tail
2009-01-29 22:19:07 -06:00
44bfff7c7b
Rename ##load-indirect to ##load-reference since this is more descriptive; value numbering doesn't assign expressions to ##load-reference nodes since this would end up folding literals which were eq? but not =
2009-01-29 01:44:58 -06:00
65a53e1fa5
Don't keep compiled-effect around anymore
2009-02-23 23:55:16 -06:00
ca2f01e0d0
Updating non-core libraries for monotonic? change
2009-01-16 17:02:54 -06:00
34792a9f23
Remove >r/r>
2008-12-17 19:17:37 -06:00
c5160d76e9
Merge qualified, alias, symbols, constants into core
2008-12-17 18:10:01 -06:00
0021cde743
Fix more compile errors
2008-12-15 22:21:56 -06:00
daf2857d7b
define-inline now takes an effect parameter
...
C: now sets the stack effect of the new word to slot names
add new basis/constants vocab; CONSTANT: foo bar is equivalent to : foo bar ; inline
(going to fold CONSTANT:, ALIAS: and qualified into core soon perhaps)
tighten the screws: inline words now need a stack declaration too
update some usages of inline words
2008-12-15 19:44:56 -06:00
eb79c6ab71
Fix conflict
2008-12-09 17:52:45 -06:00
4f0a9f311e
Untangling some dependencies
2008-12-08 14:58:00 -06:00
8a8f0c925c
Use BSR instruction to implement fixnum-log2 intrinsic
2008-12-06 15:31:17 -06:00
145b635eb6
More optimization intended to reduce compile time. Another 10% speedup on compiling empty PEG parser
...
- new map-flat combinator replaces usages of 'map flatten' in compiler
- compiler.tree.def-use.simplified uses an explicit accumulator instead of flatten
- compiler.tree.tuple-unboxing uses an explicit accumulator instead of flatten
- fix inlining regression from last time: custom inlining results would sometimes be discarded
- compiler.tree's 3each and 3map combinators rewritten to not use flip
- rewrite math.partial-dispatch without locals (purely stylistic, no performance increase)
- hand-optimize flip for common arrays-of-arrays case
- don't run escape analysis and tuple unboxing if there are no allocations in the IR
2008-12-06 11:17:19 -06:00
a56d480aa6
Various optimizations leading to a 10% speedup on compiling empty EBNF parser:
...
- open-code getenv primitive
- inline tuple predicates in finalization
- faster partial dispatch
- faster built-in type predicates
- faster tuple predicates
- faster lo-tag dispatch
- compile V{ } clone and H{ } clone more efficiently
- add fixnum fast-path to =; avoid indirect branch if two fixnums not eq
- faster >alist on hashtables
2008-12-06 09:16:29 -06:00
908644ee7a
O(1) <sbuf> and new-sequence on byte-arrays (work in progress)
2008-12-05 07:28:52 -06:00
e256846acd
Tweak string representation; high bit indicates if character has high bits in aux vector. Avoids memory access in common case. Split set-string-nth into two primitives; set-string-nth-fast is open-coded by optimizing compiler. 13% improvement on reverse-complement
2008-12-05 06:38:51 -06:00
e7f4563374
fixnum* intrinsic for x86
2008-11-30 07:26:49 -06:00
1c0789e616
Didn't generate ##branch after ##alien-invoke and ##alien-indirect
2008-11-29 03:46:57 -06:00
67471dedac
Forgot to emit a ##branch after ##fixnum-{add,sub,mul} ops; tools.annotations test faild
2008-11-29 00:36:43 -06:00
68dd6cc35a
Working on PowerPC overflow checks
2008-11-28 20:02:50 -06:00
eb8c621b6f
Faster generic arithmetic on fiixnums: both-fixnums? sub-primitive performs a check if the top two stack items are both fixnums with a single conditional branch
2008-11-28 08:35:02 -06:00
f44506089d
More work on overflow instructions: don't need temp register anymore, add -tail variants which don't need stack frame
2008-11-28 06:36:30 -06:00
5634becda1
##fixnum-add, ##fixnum-sub instructions open-code overflow check
2008-11-28 05:33:58 -06:00
ab689c098b
Clean up direct literal code and make a first attempt at PowerPC support
2008-11-24 08:16:14 -06:00
eb05dd3a12
Optimize a ##dispatch that is applied to the result of a ##sub-imm or ##add-imm; this eliminates an instruction from the common 1 fixnum-fast { ... } dispatch and 8 fixnum-fast { ... } dispatch code sequences appearing in generic word expansions
2008-11-13 04:16:08 -06:00
b1aa3697cb
Emit branches in the same order they're written in code
2008-11-13 03:52:01 -06:00
26f309d2ae
Trying to make PEGs compile faster by reducing the number of low level IR nodes: merge functionality of #>r and #r> into #shuffle, and generate 1 node instead of 3 for calls to get-local
2008-11-11 18:46:31 -06:00
5b7d40d9b4
We need to end the basic block after the ##prologue in the dispatch branch so that the GC check can go after the prologue
2008-11-10 02:58:05 -06:00
d1f248dac6
Fixing PowerPC backend: prolog register clobberage, spilling, and general stack frame usage. Add some lame tests for spilling
2008-11-06 19:00:56 -06:00
4e55cd973b
If a #dispatch branch is a call to another word which is not an intrinsic, we avoid generating the dispatch branch and just jump to the word directly
2008-11-06 11:48:55 -06:00
64cbf619a9
Add more algebraic simplifications: comparison of a comparison, comparison where first is immediate
2008-11-06 09:27:52 -06:00
1c1333fbe9
Compile not and >boolean as branchless intrinsics by having the CFG builder detect certain code patterns
2008-11-06 09:09:21 -06:00
d2ec46e38f
PowerPC backend almost functional; some new compiler unit tests added,
...
better compilation of 'f eq?'; f becomes an immediate operand
move aux-offset to compiler.constants
2008-11-06 06:27:27 -06:00
53cd75b06c
Add string-nth intrinsic
2008-11-06 01:11:28 -06:00
cc879fa9b7
Tuple layouts are now arrays, instead of built-in types. The superclass
...
array is now part of the tuple layout object itself, and class hashcodes
are stored alongside class words there. This removes 2 indirections when
reading a superclass, and 3 when reading a superclass hashcode.
2008-11-05 22:20:29 -06:00
c8521ad826
Add tool to dump live intervals
2008-11-04 17:23:20 -06:00
1af3b8ed65
If a register is spilled and the reload location is also a copy, we chicken out and don't coalesce. This is suboptimal but it's not clear to me how to implement it cleanly, and SSA graph coloring will solve this problem eventually anyway
2008-11-04 00:32:56 -06:00
efcb916e35
Handle a jump to a jump by cloning the block, in the same way we optimize jumps to returns
2008-11-03 21:02:34 -06:00
82e9eedd39
Forgot about float= intrinsic
2008-11-03 07:15:09 -06:00
59f4f25b91
Loop alignment: appears to be a small win for reverse-complement
2008-11-03 06:20:51 -06:00
0250d7c1d2
Forgot a drop in unit test
2008-11-03 01:53:27 -06:00
08b277e892
Fix alias anlaysis bug: result of ##box-alien and ##box-float did not get an AC, so calling hi-tag would not compile
2008-11-03 01:52:55 -06:00
f253e74942
Inline subprimitives in optimizing compiler
2008-11-03 01:18:54 -06:00
445e353337
Optimize away useless jumps
2008-11-02 23:09:31 -06:00
d11a779fdd
RicE
2008-11-02 03:58:32 -06:00
3834eaeb05
Don't generate GC checks if the allocation instruction is optimized out
2008-11-02 01:50:48 -06:00
8df1aba71d
Coalescing
2008-11-02 01:49:57 -06:00
804c6f93ea
Fix x86.32
2008-10-31 21:07:41 -05:00
db4db19cd9
Start working on coalescing
2008-10-28 02:38:37 -07:00
492a15e345
Move insn class to compiler.cfg.instructions
2008-10-24 09:17:06 -05:00
885adc8dc9
Various improvements
2008-10-23 05:27:54 -05:00
52967c5bf1
Improved value numbering
2008-10-23 02:49:26 -05:00
d723b69990
Value numbering
2008-10-22 21:59:07 -05:00
084e64d0bc
Tweak inheritance
2008-10-22 21:58:46 -05:00
ba0ed21e1d
Better error message
2008-10-22 21:58:37 -05:00
9a07760542
CFG optimizer
2008-10-22 18:41:37 -05:00
3711aa3bca
Stack height normalization
2008-10-22 18:41:26 -05:00
6d812aaa52
Write barrier elimination
2008-10-22 18:41:10 -05:00
1f693b50b3
Massive focused action
2008-10-22 18:39:41 -05:00
73d01452cb
Replace ##gc with a gc flag in the basic block
2008-10-22 18:38:30 -05:00
f09813f6fd
Alias analysis
2008-10-22 18:37:47 -05:00
1b06ab1b39
Fixing various bugs
2008-10-21 23:17:32 -05:00
94a2bfa2ea
Working on comparison operations, clearing out remaining dead wood
2008-10-21 03:20:48 -05:00
df498c21a3
Trim USING:
2008-10-20 20:46:47 -05:00
e92f795a76
More work on intrinsics; memory allocation and slot access now expands correctly
2008-10-20 20:40:15 -05:00
37cf7d9a9c
Add SSA comparison instructions, fix various problems
2008-10-20 05:55:20 -05:00
f092622fac
CFG IR is now pure SSA
2008-10-20 01:56:28 -05:00
14d8696f40
Oops, don't mix register classes in active set
2008-10-19 03:34:42 -05:00
c0d89b061e
Fixing register allocator prspilling
2008-10-19 01:10:21 -05:00
627dfd1ff5
Finish vreg simplification
2008-10-17 20:03:59 -05:00
239578353f
Simplifying vregs work in progress
2008-10-17 15:35:04 -05:00
ae3c4ae1b6
Fix some problems with callbacks
2008-10-12 23:32:14 -05:00
749c77d6b1
Fix linear scan test
2008-10-12 18:40:10 -05:00
b2ade7f556
Fix callbacks and non-tailcalls to dispatch
2008-10-12 17:37:26 -05:00
5f93ab74e4
Fix #dispatch generation
2008-10-12 16:46:59 -05:00
68c9b22cef
Don't need known-tag templates any more
2008-10-11 14:08:00 -05:00
f979ae5b82
Expand slot accessors further to avoid having to use complex template decision
2008-10-11 14:05:15 -05:00
3844cb62d8
Fix %write-barrier
2008-10-10 03:16:26 -05:00
3e29808f17
Fix loop compilation
2008-10-10 02:33:32 -05:00
cf46a832e7
Debugging register allocator and inline allocation
2008-10-08 23:42:53 -05:00
0e4e05d5cd
Debugging new codegen
2008-10-08 03:51:44 -05:00
7b6d9c4c4f
Debugging new codegen
2008-10-07 20:00:38 -05:00
f436fd0c0f
Merging in new codegen
2008-10-07 16:16:50 -05:00
Slava Pestov
Doug Coleman
Doug Coleman
sheeple