compiler.cfg: Major restructuring -- do not compute liveness before local optimization, and instead change local optimizations to be more permissive of undefined values. Now, liveness is only computed once, after phi elimination and before register allocation. This means liveness analysis does not need to take phi nodes into account and can now use the new compiler.cfg.dataflow-analysis framework

basis/compiler/cfg/alias-analysis/alias-analysis.factor

@@ -3,8 +3,7 @@
 USING: kernel math namespaces assocs hashtables sequences arrays
 accessors vectors combinators sets classes compiler.cfg
 compiler.cfg.registers compiler.cfg.instructions
-compiler.cfg.copy-prop compiler.cfg.rpo
+compiler.cfg.copy-prop compiler.cfg.rpo compiler.cfg.def-use ;
-compiler.cfg.liveness compiler.cfg.local ;
 IN: compiler.cfg.alias-analysis
 
 ! We try to eliminate redundant slot operations using some simple heuristics.
@@ -197,7 +196,10 @@ M: ##set-slot insn-object obj>> resolve ;
 M: ##set-slot-imm insn-object obj>> resolve ;
 M: ##alien-global insn-object drop \ ##alien-global ;
 
-: init-alias-analysis ( live-in -- )
+: inputs ( insns -- seq )
+    [ [ ##phi? not ] filter gen-set ] [ kill-set ] bi assoc-diff keys ;
+
+: init-alias-analysis ( insns -- insns' )
     H{ } clone histories set
     H{ } clone vregs>acs set
     H{ } clone acs>vregs set
@@ -208,7 +210,7 @@ M: ##alien-global insn-object drop \ ##alien-global ;
     0 ac-counter set
     next-ac heap-ac set
 
-    [ set-heap-ac ] each ;
+    dup inputs [ set-heap-ac ] each ;
 
 GENERIC: analyze-aliases* ( insn -- insn' )
 
@@ -280,9 +282,10 @@ M: insn eliminate-dead-stores* ;
     [ insn# set eliminate-dead-stores* ] map-index sift ;
 
 : alias-analysis-step ( insns -- insns' )
+    init-alias-analysis
     analyze-aliases
     compute-live-stores
     eliminate-dead-stores ;
 
 : alias-analysis ( cfg -- cfg' )
-    [ init-alias-analysis ] [ alias-analysis-step ] local-optimization ;
+    [ alias-analysis-step ] local-optimization ;

basis/compiler/cfg/checker/checker.factor

@@ -1,7 +1,7 @@
 ! Copyright (C) 2009 Slava Pestov.
 ! See http://factorcode.org/license.txt for BSD license.
 USING: kernel compiler.cfg.instructions compiler.cfg.rpo
-compiler.cfg.def-use compiler.cfg.linearization compiler.cfg.liveness
+compiler.cfg.def-use compiler.cfg.linearization
 combinators.short-circuit accessors math sequences sets assocs ;
 IN: compiler.cfg.checker
 
@@ -54,8 +54,6 @@ ERROR: undefined-values uses defs ;
     2dup subset? [ 2drop ] [ undefined-values ] if ;
 
 : check-cfg ( cfg -- )
-    compute-liveness
-    [ entry>> live-in assoc-empty? [ bad-live-in ] unless ]
     [ [ check-basic-block ] each-basic-block ]
     [ flatten-cfg check-mr ]
-    tri ;
+    bi ;

basis/compiler/cfg/dataflow-analysis/dataflow-analysis.factor

@@ -20,7 +20,7 @@ MIXIN: dataflow-analysis
 
 GENERIC# compute-in-set 2 ( bb out-sets dfa -- set )
 
-M: kill-block compute-in-set 3drop f ;
+! M: kill-block compute-in-set 3drop f ;
 
 M:: basic-block compute-in-set ( bb out-sets dfa -- set )
     bb dfa predecessors [ out-sets at ] map dfa join-sets ;
@@ -31,7 +31,7 @@ M:: basic-block compute-in-set ( bb out-sets dfa -- set )
 
 GENERIC# compute-out-set 2 ( bb out-sets dfa -- set )
 
-M: kill-block compute-out-set 3drop f ;
+! M: kill-block compute-out-set 3drop f ;
 
 M:: basic-block compute-out-set ( bb in-sets dfa -- set )
     bb in-sets at bb dfa transfer-set ;

basis/compiler/cfg/debugger/debugger.factor

@@ -7,7 +7,7 @@ parser compiler.tree.builder compiler.tree.optimizer
 compiler.cfg.builder compiler.cfg.linearization
 compiler.cfg.registers compiler.cfg.stack-frame
 compiler.cfg.linear-scan compiler.cfg.two-operand
-compiler.cfg.liveness compiler.cfg.optimizer
+compiler.cfg.optimizer
 compiler.cfg.mr compiler.cfg ;
 IN: compiler.cfg.debugger
 

basis/compiler/cfg/def-use/def-use.factor

@@ -1,6 +1,7 @@
 ! Copyright (C) 2008, 2009 Slava Pestov.
 ! See http://factorcode.org/license.txt for BSD license.
-USING: accessors arrays kernel assocs compiler.cfg.instructions ;
+USING: accessors arrays kernel assocs sequences
+sets compiler.cfg.instructions ;
 IN: compiler.cfg.def-use
 
 GENERIC: defs-vregs ( insn -- seq )
@@ -62,3 +63,12 @@ UNION: vreg-insn
 _conditional-branch
 _compare-imm-branch
 _dispatch ;
+
+: map-unique ( seq quot -- assoc )
+    map concat unique ; inline
+
+: gen-set ( instructions -- seq )
+    [ uses-vregs ] map-unique ;
+
+: kill-set ( instructions -- seq )
+    [ defs-vregs ] map-unique ;

basis/compiler/cfg/linear-scan/assignment/assignment.factor

@@ -4,14 +4,15 @@ USING: accessors kernel math assocs namespaces sequences heaps
 fry make combinators sets locals
 cpu.architecture
 compiler.cfg
+compiler.cfg.rpo
 compiler.cfg.def-use
-compiler.cfg.liveness
 compiler.cfg.registers
 compiler.cfg.instructions
 compiler.cfg.linear-scan.mapping
 compiler.cfg.linear-scan.allocation
 compiler.cfg.linear-scan.allocation.state
-compiler.cfg.linear-scan.live-intervals ;
+compiler.cfg.linear-scan.live-intervals
+compiler.cfg.linear-scan.liveness ;
 IN: compiler.cfg.linear-scan.assignment
 
 ! This contains both active and inactive intervals; any interval
@@ -185,6 +186,6 @@ ERROR: bad-vreg vreg ;
         ] V{ } make
     ] change-instructions drop ;
 
-: assign-registers ( live-intervals rpo -- )
+: assign-registers ( live-intervals cfg -- )
     [ init-assignment ] dip
-    [ assign-registers-in-block ] each ;
+    [ assign-registers-in-block ] each-basic-block ;

basis/compiler/cfg/linear-scan/linear-scan-tests.factor

@@ -7,7 +7,6 @@ compiler.cfg
 compiler.cfg.optimizer
 compiler.cfg.instructions
 compiler.cfg.registers
-compiler.cfg.liveness
 compiler.cfg.predecessors
 compiler.cfg.rpo
 compiler.cfg.linearization
@@ -1507,9 +1506,7 @@ SYMBOL: linear-scan-result
     [
         cfg new 0 get >>entry
         compute-predecessors
-        compute-liveness
+        dup { { int-regs regs } } (linear-scan)
-        dup reverse-post-order
-        { { int-regs regs } } (linear-scan)
         cfg-changed
         flatten-cfg 1array mr.
     ] with-scope ;
@@ -2331,9 +2328,6 @@ test-diamond
 ! early in bootstrap on x86-32
 [ t ] [
     [
-        H{ } clone live-ins set
-        H{ } clone live-outs set
-        H{ } clone phi-live-ins set
         T{ basic-block
            { id 12345 }
            { instructions
@@ -2353,7 +2347,8 @@ test-diamond
                  T{ ##replace f V int-regs 5 D 0 }
              }
            }
-        } dup 1array { { int-regs V{ 0 1 2 3 } } } (linear-scan)
+        } cfg new over >>entry
+        { { int-regs V{ 0 1 2 3 } } } (linear-scan)
         instructions>> first
         live-values>> assoc-empty?
     ] with-scope

basis/compiler/cfg/linear-scan/linear-scan.factor

@@ -6,6 +6,7 @@ compiler.cfg
 compiler.cfg.rpo
 compiler.cfg.instructions
 compiler.cfg.linear-scan.numbering
+compiler.cfg.linear-scan.liveness
 compiler.cfg.linear-scan.live-intervals
 compiler.cfg.linear-scan.allocation
 compiler.cfg.linear-scan.allocation.state
@@ -28,17 +29,18 @@ IN: compiler.cfg.linear-scan
 ! by Omri Traub, Glenn Holloway, Michael D. Smith
 ! http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.34.8435
 
-:: (linear-scan) ( rpo machine-registers -- )
+:: (linear-scan) ( cfg machine-registers -- )
-    rpo number-instructions
+    cfg compute-live-sets
-    rpo compute-live-intervals machine-registers allocate-registers
+    cfg number-instructions
-    rpo assign-registers
+    cfg compute-live-intervals machine-registers allocate-registers
-    rpo resolve-data-flow
+    cfg assign-registers
-    rpo check-numbering ;
+    cfg resolve-data-flow
+    cfg check-numbering ;
 
 : linear-scan ( cfg -- cfg' )
     [
         init-mapping
-        dup reverse-post-order machine-registers (linear-scan)
+        dup machine-registers (linear-scan)
         spill-counts get >>spill-counts
         cfg-changed
     ] with-scope ;

basis/compiler/cfg/linear-scan/live-intervals/live-intervals.factor

@@ -2,7 +2,8 @@
 ! See http://factorcode.org/license.txt for BSD license.
 USING: namespaces kernel assocs accessors sequences math math.order fry
 combinators binary-search compiler.cfg.instructions compiler.cfg.registers
-compiler.cfg.def-use compiler.cfg.liveness compiler.cfg ;
+compiler.cfg.def-use compiler.cfg.linear-scan.liveness compiler.cfg.rpo
+compiler.cfg ;
 IN: compiler.cfg.linear-scan.live-intervals
 
 TUPLE: live-range from to ;
@@ -144,10 +145,10 @@ ERROR: bad-live-interval live-interval ;
         } cleave
     ] each ;
 
-: compute-live-intervals ( rpo -- live-intervals )
+: compute-live-intervals ( cfg -- live-intervals )
     H{ } clone [
         live-intervals set
-        <reversed> [ compute-live-intervals-step ] each
+        post-order [ compute-live-intervals-step ] each
     ] keep values dup finish-live-intervals ;
 
 : relevant-ranges ( interval1 interval2 -- ranges1 ranges2 )

basis/compiler/cfg/linear-scan/liveness/liveness.factor

@@ -0,0 +1,17 @@
+! Copyright (C) 2009 Slava Pestov.
+! See http://factorcode.org/license.txt for BSD license.
+USING: kernel accessors assocs compiler.cfg.def-use
+compiler.cfg.dataflow-analysis ;
+IN: compiler.cfg.linear-scan.liveness
+
+! See http://en.wikipedia.org/wiki/Liveness_analysis
+
+BACKWARD-ANALYSIS: live
+
+M: live-analysis transfer-set
+    drop instructions>>
+    [ gen-set assoc-union ] keep
+    kill-set assoc-diff ;
+
+M: live-analysis join-sets
+    drop assoc-combine ;

basis/compiler/cfg/linear-scan/numbering/numbering.factor

@@ -1,6 +1,7 @@
 ! Copyright (C) 2009 Slava Pestov.
 ! See http://factorcode.org/license.txt for BSD license.
-USING: kernel accessors math sequences grouping namespaces ;
+USING: kernel accessors math sequences grouping namespaces
+compiler.cfg.rpo ;
 IN: compiler.cfg.linear-scan.numbering
 
 : number-instructions ( rpo -- )
@@ -8,7 +9,7 @@ IN: compiler.cfg.linear-scan.numbering
         instructions>> [
             [ (>>insn#) ] [ drop 2 + ] 2bi
         ] each
-    ] each drop ;
+    ] each-basic-block drop ;
 
 SYMBOL: check-numbering?
 
@@ -18,5 +19,5 @@ ERROR: bad-numbering bb ;
     dup instructions>> [ insn#>> ] map sift [ <= ] monotonic?
     [ drop ] [ bad-numbering ] if ;
 
-: check-numbering ( rpo -- )
+: check-numbering ( cfg -- )
-    check-numbering? get [ [ check-block-numbering ] each ] [ drop ] if ;
+    check-numbering? get [ [ check-block-numbering ] each-basic-block ] [ drop ] if ;

basis/compiler/cfg/linear-scan/resolve/resolve.factor

@@ -3,10 +3,12 @@
 USING: accessors arrays assocs combinators
 combinators.short-circuit fry kernel locals
 make math sequences
+compiler.cfg.rpo
 compiler.cfg.utilities
 compiler.cfg.instructions
 compiler.cfg.linear-scan.assignment
-compiler.cfg.linear-scan.mapping compiler.cfg.liveness ;
+compiler.cfg.linear-scan.mapping
+compiler.cfg.linear-scan.liveness ;
 IN: compiler.cfg.linear-scan.resolve
 
 : add-mapping ( from to reg-class -- )
@@ -43,5 +45,5 @@ IN: compiler.cfg.linear-scan.resolve
 : resolve-block-data-flow ( bb -- )
     dup successors>> [ resolve-edge-data-flow ] with each ;
 
-: resolve-data-flow ( rpo -- )
+: resolve-data-flow ( cfg -- )
-    [ resolve-block-data-flow ] each ;
+    [ resolve-block-data-flow ] each-basic-block ;

basis/compiler/cfg/linearization/linearization.factor

@@ -4,7 +4,6 @@ USING: kernel math accessors sequences namespaces make
 combinators assocs arrays locals cpu.architecture
 compiler.cfg
 compiler.cfg.rpo
-compiler.cfg.liveness
 compiler.cfg.comparisons
 compiler.cfg.stack-frame
 compiler.cfg.instructions ;

basis/compiler/cfg/liveness/authors.txt

@@ -1 +0,0 @@
-Slava Pestov

basis/compiler/cfg/liveness/liveness-tests.factor

@@ -1,15 +0,0 @@
-USING: compiler.cfg compiler.cfg.instructions compiler.cfg.registers
-compiler.cfg.liveness accessors tools.test cpu.architecture ;
-IN: compiler.cfg.liveness.tests
-
-[
-    H{
-        { "A" H{ { V int-regs 1 V int-regs 1 } { V int-regs 4 V int-regs 4 } } }
-        { "B" H{ { V int-regs 3 V int-regs 3 } { V int-regs 2 V int-regs 2 } } }
-    }
-] [
-    <basic-block> V{
-        T{ ##phi f V int-regs 0 { { "A" V int-regs 1 } { "B" V int-regs 2 } } }
-        T{ ##phi f V int-regs 1 { { "B" V int-regs 3 } { "A" V int-regs 4 } } }
-    } >>instructions compute-phi-live-in
-] unit-test

basis/compiler/cfg/liveness/liveness.factor

@@ -1,79 +0,0 @@
-! Copyright (C) 2009 Slava Pestov.
-! See http://factorcode.org/license.txt for BSD license.
-USING: kernel namespaces deques accessors sets sequences assocs fry
-hashtables dlists compiler.cfg.def-use compiler.cfg.instructions
-compiler.cfg.rpo ;
-IN: compiler.cfg.liveness
-
-! This is a backward dataflow analysis. See http://en.wikipedia.org/wiki/Liveness_analysis
-
-! Assoc mapping basic blocks to sets of vregs
-SYMBOL: live-ins
-
-: live-in ( basic-block -- set ) live-ins get at ;
-
-! Assoc mapping basic blocks to sequences of sets of vregs; each sequence
-! is in conrrespondence with a predecessor
-SYMBOL: phi-live-ins
-
-: phi-live-in ( predecessor basic-block -- set ) phi-live-ins get at at ;
-
-! Assoc mapping basic blocks to sets of vregs
-SYMBOL: live-outs
-
-: live-out ( basic-block -- set ) live-outs get at ;
-
-SYMBOL: work-list
-
-: add-to-work-list ( basic-blocks -- )
-    work-list get '[ _ push-front ] each ;
-
-: map-unique ( seq quot -- assoc )
-    map concat unique ; inline
-
-: gen-set ( instructions -- seq )
-    [ ##phi? not ] filter [ uses-vregs ] map-unique ;
-
-: kill-set ( instructions -- seq )
-    [ [ defs-vregs ] [ temp-vregs ] bi append ] map-unique ;
-
-: compute-live-in ( basic-block -- live-in )
-    dup instructions>>
-    [ [ live-out ] [ gen-set ] bi* assoc-union ]
-    [ nip kill-set ]
-    2bi assoc-diff ;
-
-: compute-phi-live-in ( basic-block -- phi-live-in )
-    instructions>> [ ##phi? ] filter [ f ] [
-        H{ } clone [
-            '[ inputs>> [ swap _ conjoin-at ] assoc-each ] each
-        ] keep
-    ] if-empty ;
-
-: update-live-in ( basic-block -- changed? )
-    [ [ compute-live-in ] keep live-ins get maybe-set-at ]
-    [ [ compute-phi-live-in ] keep phi-live-ins get maybe-set-at ]
-    bi and ; 
-
-: compute-live-out ( basic-block -- live-out )
-    [ successors>> [ live-in ] map ]
-    [ dup successors>> [ phi-live-in ] with map ] bi
-    append assoc-combine ;
-
-: update-live-out ( basic-block -- changed? )
-    [ compute-live-out ] keep
-    live-outs get maybe-set-at ;
-
-: liveness-step ( basic-block -- )
-    dup update-live-out [
-        dup update-live-in
-        [ predecessors>> add-to-work-list ] [ drop ] if
-    ] [ drop ] if ;
-
-: compute-liveness ( cfg -- cfg' )
-    <hashed-dlist> work-list set
-    H{ } clone live-ins set
-    H{ } clone phi-live-ins set
-    H{ } clone live-outs set
-    dup post-order add-to-work-list
-    work-list get [ liveness-step ] slurp-deque ;

basis/compiler/cfg/local/authors.txt

@@ -1 +0,0 @@
-Slava Pestov

basis/compiler/cfg/local/local.factor

@@ -1,14 +0,0 @@
-! Copyright (C) 2009 Slava Pestov.
-! See http://factorcode.org/license.txt for BSD license.
-USING: locals accessors kernel assocs namespaces
-compiler.cfg compiler.cfg.liveness compiler.cfg.rpo ;
-IN: compiler.cfg.local
-
-:: optimize-basic-block ( bb init-quot insn-quot -- )
-    bb basic-block set
-    bb live-in keys init-quot call
-    bb insn-quot change-instructions drop ; inline
-
-:: local-optimization ( cfg init-quot: ( live-in -- ) insn-quot: ( insns -- insns' ) -- cfg' )
-    cfg [ init-quot insn-quot optimize-basic-block ] each-basic-block
-    cfg ; inline

basis/compiler/cfg/mr/mr.factor

@@ -1,13 +1,12 @@
 ! Copyright (C) 2009 Slava Pestov.
 ! See http://factorcode.org/license.txt for BSD license.
 USING: compiler.cfg.linearization compiler.cfg.two-operand
-compiler.cfg.liveness compiler.cfg.gc-checks compiler.cfg.linear-scan
+compiler.cfg.gc-checks compiler.cfg.linear-scan
 compiler.cfg.build-stack-frame compiler.cfg.rpo ;
 IN: compiler.cfg.mr
 
 : build-mr ( cfg -- mr )
     convert-two-operand
-    compute-liveness
     insert-gc-checks
     linear-scan
     flatten-cfg

basis/compiler/cfg/optimizer/optimizer.factor

@@ -11,7 +11,6 @@ compiler.cfg.alias-analysis
 compiler.cfg.value-numbering
 compiler.cfg.dce
 compiler.cfg.write-barrier
-compiler.cfg.liveness
 compiler.cfg.rpo
 compiler.cfg.phi-elimination
 compiler.cfg.checker ;
@@ -35,7 +34,6 @@ SYMBOL: check-optimizer?
         join-blocks
         compute-predecessors
         stack-analysis
-        compute-liveness
         alias-analysis
         value-numbering
         compute-predecessors

basis/compiler/cfg/renaming/renaming.factor

@@ -6,7 +6,7 @@ IN: compiler.cfg.renaming
 
 SYMBOL: renamings
 
-: rename-value ( vreg -- vreg' ) renamings get at ;
+: rename-value ( vreg -- vreg' ) renamings get ?at drop ;
 
 GENERIC: rename-insn-defs ( insn -- )
 

basis/compiler/cfg/rpo/rpo.factor

@@ -33,3 +33,10 @@ SYMBOL: visited
 
 : each-basic-block ( cfg quot -- )
     [ reverse-post-order ] dip each ; inline
+
+: optimize-basic-block ( bb quot -- )
+    [ drop basic-block set ]
+    [ change-instructions drop ] 2bi ; inline
+
+: local-optimization ( cfg quot: ( insns -- insns' ) -- cfg' )
+    dupd '[ _ optimize-basic-block ] each-basic-block ; inline

basis/compiler/cfg/two-operand/two-operand.factor

@@ -1,7 +1,7 @@
 ! Copyright (C) 2008, 2009 Slava Pestov.
 ! See http://factorcode.org/license.txt for BSD license.
 USING: accessors kernel sequences make compiler.cfg.instructions
-compiler.cfg.local cpu.architecture ;
+compiler.cfg.rpo cpu.architecture ;
 IN: compiler.cfg.two-operand
 
 ! On x86, instructions take the form x = x op y
@@ -54,7 +54,6 @@ M: insn convert-two-operand* , ;
 
 : convert-two-operand ( cfg -- cfg' )
     two-operand? [
-        [ drop ]
         [ [ [ convert-two-operand* ] each ] V{ } make ]
         local-optimization
     ] when ;

basis/compiler/cfg/value-numbering/expressions/expressions.factor

@@ -6,7 +6,6 @@ compiler.cfg.value-numbering.graph ;
 IN: compiler.cfg.value-numbering.expressions
 
 ! Referentially-transparent expressions
-TUPLE: expr op ;
 TUPLE: unary-expr < expr in ;
 TUPLE: binary-expr < expr in1 in2 ;
 TUPLE: commutative-expr < binary-expr ;
@@ -37,17 +36,6 @@ M: reference-expr equal?
         } cond
     ] [ 2drop f ] if ;
 
-! Expressions whose values are inputs to the basic block. We
-! can eliminate a second computation having the same 'n' as
-! the first one; we can also eliminate input-exprs whose
-! result is not used.
-TUPLE: input-expr < expr n ;
-
-SYMBOL: input-expr-counter
-
-: next-input-expr ( class -- expr )
-    input-expr-counter [ dup 1 + ] change input-expr boa ;
-
 : constant>vn ( constant -- vn ) <constant> expr>vn ; inline
 
 GENERIC: >expr ( insn -- expr )
@@ -97,7 +85,7 @@ M: ##compare-imm >expr compare-imm>expr ;
 
 M: ##compare-float >expr compare>expr ;
 
-M: ##flushable >expr class next-input-expr ;
+M: ##flushable >expr drop next-input-expr ;
 
 : init-expressions ( -- )
     0 input-expr-counter set ;

basis/compiler/cfg/value-numbering/graph/graph.factor

@@ -10,13 +10,24 @@ SYMBOL: vn-counter
 ! biassoc mapping expressions to value numbers
 SYMBOL: exprs>vns
 
+TUPLE: expr op ;
+
 : expr>vn ( expr -- vn ) exprs>vns get [ drop next-vn ] cache ;
 
 : vn>expr ( vn -- expr ) exprs>vns get value-at ;
 
+! Expressions whose values are inputs to the basic block.
+TUPLE: input-expr < expr n ;
+
+SYMBOL: input-expr-counter
+
+: next-input-expr ( -- expr )
+    f input-expr-counter counter input-expr boa ;
+
 SYMBOL: vregs>vns
 
-: vreg>vn ( vreg -- vn ) vregs>vns get at ;
+: vreg>vn ( vreg -- vn )
+    vregs>vns get [ drop next-input-expr expr>vn ] cache ;
 
 : vn>vreg ( vn -- vreg ) vregs>vns get value-at ;
 

basis/compiler/cfg/value-numbering/value-numbering-tests.factor

@@ -3,7 +3,7 @@ USING: compiler.cfg.value-numbering compiler.cfg.instructions
 compiler.cfg.registers compiler.cfg.debugger compiler.cfg.comparisons
 cpu.architecture tools.test kernel math combinators.short-circuit
 accessors sequences compiler.cfg.predecessors locals
-compiler.cfg.phi-elimination compiler.cfg.dce compiler.cfg.liveness
+compiler.cfg.phi-elimination compiler.cfg.dce
 compiler.cfg assocs vectors arrays layouts namespaces ;
 
 : trim-temps ( insns -- insns )
@@ -15,10 +15,6 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         } 1|| [ f >>temp ] when
     ] map ;
 
-: test-value-numbering ( insns -- insns )
-    { } init-value-numbering
-    value-numbering-step ;
-
 ! Folding constants together
 [
     {
@@ -33,7 +29,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##load-reference f V int-regs 1 -0.0 }
         T{ ##replace f V int-regs 0 D 0 }
         T{ ##replace f V int-regs 1 D 1 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -49,7 +45,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##load-reference f V int-regs 1 0.0 }
         T{ ##replace f V int-regs 0 D 0 }
         T{ ##replace f V int-regs 1 D 1 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -65,7 +61,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##load-reference f V int-regs 1 t }
         T{ ##replace f V int-regs 0 D 0 }
         T{ ##replace f V int-regs 1 D 1 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 ! Copy propagation
@@ -80,7 +76,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 45 D 1 }
         T{ ##copy f V int-regs 48 V int-regs 45 }
         T{ ##compare-imm-branch f V int-regs 48 7 cc/= }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 ! Compare propagation
@@ -99,7 +95,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##compare f V int-regs 4 V int-regs 2 V int-regs 1 cc> }
         T{ ##compare-imm f V int-regs 6 V int-regs 4 5 cc/= }
         T{ ##replace f V int-regs 6 D 0 }
-    } test-value-numbering trim-temps
+    } value-numbering-step trim-temps
 ] unit-test
 
 [
@@ -117,7 +113,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##compare f V int-regs 4 V int-regs 2 V int-regs 1 cc<= }
         T{ ##compare-imm f V int-regs 6 V int-regs 4 5 cc= }
         T{ ##replace f V int-regs 6 D 0 }
-    } test-value-numbering trim-temps
+    } value-numbering-step trim-temps
 ] unit-test
 
 [
@@ -139,7 +135,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##compare-float f V int-regs 12 V double-float-regs 10 V double-float-regs 11 cc< }
         T{ ##compare-imm f V int-regs 14 V int-regs 12 5 cc= }
         T{ ##replace f V int-regs 14 D 0 }
-    } test-value-numbering trim-temps
+    } value-numbering-step trim-temps
 ] unit-test
 
 [
@@ -155,7 +151,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 30 D -2 }
         T{ ##compare f V int-regs 33 V int-regs 29 V int-regs 30 cc<= }
         T{ ##compare-imm-branch f V int-regs 33 5 cc/= }
-    } test-value-numbering trim-temps
+    } value-numbering-step trim-temps
 ] unit-test
 
 ! Immediate operand conversion
@@ -170,7 +166,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##add f V int-regs 2 V int-regs 0 V int-regs 1 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -184,7 +180,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##add f V int-regs 2 V int-regs 1 V int-regs 0 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -198,7 +194,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##sub f V int-regs 2 V int-regs 0 V int-regs 1 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -210,7 +206,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
     {
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##sub f V int-regs 1 V int-regs 0 V int-regs 0 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -224,7 +220,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##mul f V int-regs 2 V int-regs 0 V int-regs 1 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -238,7 +234,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##mul f V int-regs 2 V int-regs 1 V int-regs 0 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -250,7 +246,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
     {
         T{ ##peek f V int-regs 1 D 0 }
         T{ ##mul-imm f V int-regs 2 V int-regs 1 8 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -264,7 +260,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##and f V int-regs 2 V int-regs 0 V int-regs 1 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -278,7 +274,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##and f V int-regs 2 V int-regs 1 V int-regs 0 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -292,7 +288,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##or f V int-regs 2 V int-regs 0 V int-regs 1 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -306,7 +302,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##or f V int-regs 2 V int-regs 1 V int-regs 0 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -320,7 +316,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##xor f V int-regs 2 V int-regs 0 V int-regs 1 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -334,7 +330,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##xor f V int-regs 2 V int-regs 1 V int-regs 0 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -348,7 +344,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##compare f V int-regs 2 V int-regs 0 V int-regs 1 cc<= }
-    } test-value-numbering trim-temps
+    } value-numbering-step trim-temps
 ] unit-test
 
 [
@@ -362,7 +358,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##compare f V int-regs 2 V int-regs 1 V int-regs 0 cc<= }
-    } test-value-numbering trim-temps
+    } value-numbering-step trim-temps
 ] unit-test
 
 [
@@ -376,7 +372,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##compare-branch f V int-regs 0 V int-regs 1 cc<= }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -390,7 +386,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 100 }
         T{ ##compare-branch f V int-regs 1 V int-regs 0 cc<= }
-    } test-value-numbering trim-temps
+    } value-numbering-step trim-temps
 ] unit-test
 
 ! Reassociation
@@ -409,7 +405,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##add f V int-regs 2 V int-regs 0 V int-regs 1 }
         T{ ##load-immediate f V int-regs 3 50 }
         T{ ##add f V int-regs 4 V int-regs 2 V int-regs 3 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -427,7 +423,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##add f V int-regs 2 V int-regs 1 V int-regs 0 }
         T{ ##load-immediate f V int-regs 3 50 }
         T{ ##add f V int-regs 4 V int-regs 3 V int-regs 2 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -445,7 +441,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##add f V int-regs 2 V int-regs 0 V int-regs 1 }
         T{ ##load-immediate f V int-regs 3 50 }
         T{ ##sub f V int-regs 4 V int-regs 2 V int-regs 3 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -463,7 +459,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##sub f V int-regs 2 V int-regs 0 V int-regs 1 }
         T{ ##load-immediate f V int-regs 3 50 }
         T{ ##sub f V int-regs 4 V int-regs 2 V int-regs 3 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -481,7 +477,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##mul f V int-regs 2 V int-regs 0 V int-regs 1 }
         T{ ##load-immediate f V int-regs 3 50 }
         T{ ##mul f V int-regs 4 V int-regs 2 V int-regs 3 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -499,7 +495,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##mul f V int-regs 2 V int-regs 1 V int-regs 0 }
         T{ ##load-immediate f V int-regs 3 50 }
         T{ ##mul f V int-regs 4 V int-regs 3 V int-regs 2 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -517,7 +513,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##and f V int-regs 2 V int-regs 0 V int-regs 1 }
         T{ ##load-immediate f V int-regs 3 50 }
         T{ ##and f V int-regs 4 V int-regs 2 V int-regs 3 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -535,7 +531,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##and f V int-regs 2 V int-regs 1 V int-regs 0 }
         T{ ##load-immediate f V int-regs 3 50 }
         T{ ##and f V int-regs 4 V int-regs 3 V int-regs 2 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -553,7 +549,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##or f V int-regs 2 V int-regs 0 V int-regs 1 }
         T{ ##load-immediate f V int-regs 3 50 }
         T{ ##or f V int-regs 4 V int-regs 2 V int-regs 3 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -571,7 +567,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##or f V int-regs 2 V int-regs 1 V int-regs 0 }
         T{ ##load-immediate f V int-regs 3 50 }
         T{ ##or f V int-regs 4 V int-regs 3 V int-regs 2 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -589,7 +585,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##xor f V int-regs 2 V int-regs 0 V int-regs 1 }
         T{ ##load-immediate f V int-regs 3 50 }
         T{ ##xor f V int-regs 4 V int-regs 2 V int-regs 3 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -607,7 +603,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##xor f V int-regs 2 V int-regs 1 V int-regs 0 }
         T{ ##load-immediate f V int-regs 3 50 }
         T{ ##xor f V int-regs 4 V int-regs 3 V int-regs 2 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 ! Simplification
@@ -626,7 +622,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##sub f V int-regs 2 V int-regs 1 V int-regs 1 }
         T{ ##add f V int-regs 3 V int-regs 0 V int-regs 2 }
         T{ ##replace f V int-regs 3 D 0 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -644,7 +640,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##sub f V int-regs 2 V int-regs 1 V int-regs 1 }
         T{ ##sub f V int-regs 3 V int-regs 0 V int-regs 2 }
         T{ ##replace f V int-regs 3 D 0 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -662,7 +658,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##sub f V int-regs 2 V int-regs 1 V int-regs 1 }
         T{ ##or f V int-regs 3 V int-regs 0 V int-regs 2 }
         T{ ##replace f V int-regs 3 D 0 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -680,7 +676,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##sub f V int-regs 2 V int-regs 1 V int-regs 1 }
         T{ ##xor f V int-regs 3 V int-regs 0 V int-regs 2 }
         T{ ##replace f V int-regs 3 D 0 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -696,7 +692,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##load-immediate f V int-regs 1 1 }
         T{ ##mul f V int-regs 2 V int-regs 0 V int-regs 1 }
         T{ ##replace f V int-regs 2 D 0 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 ! Constant folding
@@ -713,7 +709,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##load-immediate f V int-regs 1 1 }
         T{ ##load-immediate f V int-regs 2 3 }
         T{ ##add f V int-regs 3 V int-regs 1 V int-regs 2 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -729,7 +725,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##load-immediate f V int-regs 1 1 }
         T{ ##load-immediate f V int-regs 2 3 }
         T{ ##sub f V int-regs 3 V int-regs 1 V int-regs 2 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -745,7 +741,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##load-immediate f V int-regs 1 2 }
         T{ ##load-immediate f V int-regs 2 3 }
         T{ ##mul f V int-regs 3 V int-regs 1 V int-regs 2 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -761,7 +757,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##load-immediate f V int-regs 1 2 }
         T{ ##load-immediate f V int-regs 2 1 }
         T{ ##and f V int-regs 3 V int-regs 1 V int-regs 2 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -777,7 +773,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##load-immediate f V int-regs 1 2 }
         T{ ##load-immediate f V int-regs 2 1 }
         T{ ##or f V int-regs 3 V int-regs 1 V int-regs 2 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -793,7 +789,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##load-immediate f V int-regs 1 2 }
         T{ ##load-immediate f V int-regs 2 3 }
         T{ ##xor f V int-regs 3 V int-regs 1 V int-regs 2 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -807,7 +803,7 @@ compiler.cfg assocs vectors arrays layouts namespaces ;
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 1 }
         T{ ##shl-imm f V int-regs 3 V int-regs 1 3 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 cell 8 = [
@@ -822,7 +818,7 @@ cell 8 = [
             T{ ##peek f V int-regs 0 D 0 }
             T{ ##load-immediate f V int-regs 1 -1 }
             T{ ##shr-imm f V int-regs 3 V int-regs 1 16 }
-        } test-value-numbering
+        } value-numbering-step
     ] unit-test
 ] when
 
@@ -837,7 +833,7 @@ cell 8 = [
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##load-immediate f V int-regs 1 -8 }
         T{ ##sar-imm f V int-regs 3 V int-regs 1 1 }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 cell 8 = [
@@ -854,7 +850,7 @@ cell 8 = [
             T{ ##load-immediate f V int-regs 1 65536 }
             T{ ##shl-imm f V int-regs 2 V int-regs 1 31 }
             T{ ##add f V int-regs 3 V int-regs 0 V int-regs 2 }
-        } test-value-numbering
+        } value-numbering-step
     ] unit-test
 
     [
@@ -868,7 +864,7 @@ cell 8 = [
             T{ ##peek f V int-regs 0 D 0 }
             T{ ##load-immediate f V int-regs 2 140737488355328 }
             T{ ##add f V int-regs 3 V int-regs 0 V int-regs 2 }
-        } test-value-numbering
+        } value-numbering-step
     ] unit-test
 
     [
@@ -884,7 +880,7 @@ cell 8 = [
             T{ ##load-immediate f V int-regs 2 2147483647 }
             T{ ##add f V int-regs 3 V int-regs 0 V int-regs 2 }
             T{ ##add f V int-regs 4 V int-regs 3 V int-regs 2 }
-        } test-value-numbering
+        } value-numbering-step
     ] unit-test
 ] when
 
@@ -900,7 +896,7 @@ cell 8 = [
         T{ ##load-immediate f V int-regs 1 1 }
         T{ ##load-immediate f V int-regs 2 2 }
         T{ ##compare f V int-regs 3 V int-regs 1 V int-regs 2 cc= }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -914,7 +910,7 @@ cell 8 = [
         T{ ##load-immediate f V int-regs 1 1 }
         T{ ##load-immediate f V int-regs 2 2 }
         T{ ##compare f V int-regs 3 V int-regs 1 V int-regs 2 cc/= }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -928,7 +924,7 @@ cell 8 = [
         T{ ##load-immediate f V int-regs 1 1 }
         T{ ##load-immediate f V int-regs 2 2 }
         T{ ##compare f V int-regs 3 V int-regs 1 V int-regs 2 cc< }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -942,7 +938,7 @@ cell 8 = [
         T{ ##load-immediate f V int-regs 1 1 }
         T{ ##load-immediate f V int-regs 2 2 }
         T{ ##compare f V int-regs 3 V int-regs 2 V int-regs 1 cc< }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -954,7 +950,7 @@ cell 8 = [
     {
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##compare f V int-regs 1 V int-regs 0 V int-regs 0 cc< }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -966,7 +962,7 @@ cell 8 = [
     {
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##compare f V int-regs 1 V int-regs 0 V int-regs 0 cc<= }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -978,7 +974,7 @@ cell 8 = [
     {
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##compare f V int-regs 1 V int-regs 0 V int-regs 0 cc> }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -990,7 +986,7 @@ cell 8 = [
     {
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##compare f V int-regs 1 V int-regs 0 V int-regs 0 cc>= }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -1002,7 +998,7 @@ cell 8 = [
     {
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##compare f V int-regs 1 V int-regs 0 V int-regs 0 cc/= }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 [
@@ -1014,12 +1010,12 @@ cell 8 = [
     {
         T{ ##peek f V int-regs 0 D 0 }
         T{ ##compare f V int-regs 1 V int-regs 0 V int-regs 0 cc= }
-    } test-value-numbering
+    } value-numbering-step
 ] unit-test
 
 : test-branch-folding ( insns -- insns' n )
     <basic-block>
-    [ V{ 0 1 } clone >>successors basic-block set test-value-numbering ] keep
+    [ V{ 0 1 } clone >>successors basic-block set value-numbering-step ] keep
     successors>> first ;
 
 [
@@ -1208,7 +1204,6 @@ test-diamond
 
 [ ] [
     cfg new 0 get >>entry
-    compute-liveness
     value-numbering
     compute-predecessors
     eliminate-phis drop
@@ -1253,7 +1248,6 @@ test-diamond
 [ ] [
     cfg new 0 get >>entry
     compute-predecessors
-    compute-liveness
     value-numbering
     compute-predecessors
     eliminate-dead-code
@@ -1324,7 +1318,7 @@ V{
 
 [ ] [
     cfg new 0 get >>entry
-    compute-liveness value-numbering eliminate-dead-code drop
+    value-numbering eliminate-dead-code drop
 ] unit-test
 
 [ f ] [ 1 get instructions>> [ ##peek? ] any? ] unit-test

basis/compiler/cfg/value-numbering/value-numbering.factor

@@ -3,8 +3,7 @@
 USING: namespaces assocs biassocs classes kernel math accessors
 sorting sets sequences fry
 compiler.cfg
-compiler.cfg.local
+compiler.cfg.rpo
-compiler.cfg.liveness
 compiler.cfg.renaming
 compiler.cfg.value-numbering.graph
 compiler.cfg.value-numbering.expressions
@@ -13,15 +12,6 @@ compiler.cfg.value-numbering.rewrite ;
 IN: compiler.cfg.value-numbering
 
 ! Local value numbering. Predecessors must be recomputed after this
-
-: number-input-values ( live-in -- )
-    [ [ f next-input-expr simplify ] dip set-vn ] each ;
-
-: init-value-numbering ( live-in -- )
-    init-value-graph
-    init-expressions
-    number-input-values ;
-
 : vreg>vreg-mapping ( -- assoc )
     vregs>vns get [ keys ] keep
     '[ dup _ [ at ] [ value-at ] bi ] H{ } map>assoc ;
@@ -32,8 +22,10 @@ IN: compiler.cfg.value-numbering
     ] with-variable ;
 
 : value-numbering-step ( insns -- insns' )
-    [ rewrite ] map dup rename-uses ;
+    init-value-graph
+    init-expressions
+    [ rewrite ] map
+    dup rename-uses ;
 
 : value-numbering ( cfg -- cfg' )
-    [ init-value-numbering ] [ value-numbering-step ] local-optimization
+    [ value-numbering-step ] local-optimization cfg-changed ;
-    cfg-changed ;

basis/compiler/cfg/write-barrier/write-barrier.factor

@@ -2,7 +2,7 @@
 ! See http://factorcode.org/license.txt for BSD license.
 USING: kernel accessors namespaces assocs sets sequences locals
 compiler.cfg compiler.cfg.instructions compiler.cfg.copy-prop
-compiler.cfg.liveness compiler.cfg.local ;
+compiler.cfg.rpo ;
 IN: compiler.cfg.write-barrier
 
 ! Eliminate redundant write barrier hits.
@@ -43,4 +43,4 @@ M: insn eliminate-write-barrier ;
     [ eliminate-write-barrier ] map sift ;
 
 : eliminate-write-barriers ( cfg -- cfg' )
-    [ drop ] [ write-barriers-step ] local-optimization ;
+    [ write-barriers-step ] local-optimization ;