[Git][ghc/ghc][wip/T22428] Fix contification with stable unfoldings (#22428)
Sebastian Graf (@sgraf812)
gitlab at gitlab.haskell.org
Thu Jan 12 12:23:50 UTC 2023
Sebastian Graf pushed to branch wip/T22428 at Glasgow Haskell Compiler / GHC
Commits:
0f67ccdd by Sebastian Graf at 2023-01-12T13:23:32+01:00
Fix contification with stable unfoldings (#22428)
Many functions now return a `TailUsageDetails` that adorns a `UsageDetails` with
a `JoinArity` that reflects the number of join point binders around the body
for which the `UsageDetails` was computed. `TailUsageDetails` is now returned by
`occAnalLamTail` as well as `occAnalUnfolding` and `occAnalRules`.
I adjusted `Note [Join points and unfoldings/rules]` and
`Note [Adjusting right-hand sides]` to account for the new machinery.
I also wrote a new `Note [Join arity prediction based on joinRhsArity]`
and refer to it when we combine `TailUsageDetails` for a recursive RHS.
I also renamed
* `occAnalLam` to `occAnalLamTail`
* `adjustRhsUsage` to `adjustTailUsage`
* a few other less important functions
and properly documented the that each call of `occAnalLamTail` must pair up with
`adjustTailUsage`.
I removed `Note [Unfoldings and join points]` because it was redundant with
`Note [Occurrences in stable unfoldings]`.
While in town, I refactored `mkLoopBreakerNodes` so that it returns a condensed
`NodeDetails` called `SimpleNodeDetails`.
Fixes #22428.
The refactoring seems to have quite beneficial effect on ghc/alloc performance:
```
CoOpt_Read(normal) ghc/alloc 784,778,420 768,091,176 -2.1% GOOD
T12150(optasm) ghc/alloc 77,762,270 75,986,720 -2.3% GOOD
T12425(optasm) ghc/alloc 85,740,186 84,641,712 -1.3% GOOD
T13056(optasm) ghc/alloc 306,104,656 299,811,632 -2.1% GOOD
T13253(normal) ghc/alloc 350,233,952 346,004,008 -1.2%
T14683(normal) ghc/alloc 2,800,514,792 2,754,651,360 -1.6%
T15304(normal) ghc/alloc 1,230,883,318 1,215,978,336 -1.2%
T15630(normal) ghc/alloc 153,379,590 151,796,488 -1.0%
T16577(normal) ghc/alloc 7,356,797,056 7,244,194,416 -1.5%
T17516(normal) ghc/alloc 1,718,941,448 1,692,157,288 -1.6%
T19695(normal) ghc/alloc 1,485,794,632 1,458,022,112 -1.9%
T21839c(normal) ghc/alloc 437,562,314 431,295,896 -1.4% GOOD
T21839r(normal) ghc/alloc 446,927,580 440,615,776 -1.4% GOOD
geo. mean -0.6%
minimum -2.4%
maximum -0.0%
```
Metric Decrease:
CoOpt_Read
T10421
T12150
T12425
T13056
T18698a
T18698b
T21839c
T21839r
T9961
- - - - -
7 changed files:
- compiler/GHC/Core/Opt/Arity.hs
- compiler/GHC/Core/Opt/OccurAnal.hs
- compiler/GHC/Data/Graph/Directed.hs
- compiler/GHC/Utils/Misc.hs
- + testsuite/tests/simplCore/should_compile/T22428.hs
- + testsuite/tests/simplCore/should_compile/T22428.stderr
- testsuite/tests/simplCore/should_compile/all.T
Changes:
=====================================
compiler/GHC/Core/Opt/Arity.hs
=====================================
@@ -132,6 +132,9 @@ joinRhsArity :: CoreExpr -> JoinArity
-- Join points are supposed to have manifestly-visible
-- lambdas at the top: no ticks, no casts, nothing
-- Moreover, type lambdas count in JoinArity
+-- NB: For non-recursive bindings, the join arity of the binding may actually be
+-- less that the number of manifestly-visible lambdas.
+-- See Note [Join arity prediction based on joinRhsArity] in GHC.Core.Opt.OccurAnal
joinRhsArity (Lam _ e) = 1 + joinRhsArity e
joinRhsArity _ = 0
=====================================
compiler/GHC/Core/Opt/OccurAnal.hs
=====================================
@@ -59,7 +59,7 @@ import GHC.Builtin.Names( runRWKey )
import GHC.Unit.Module( Module )
import Data.List (mapAccumL, mapAccumR)
-import Data.List.NonEmpty (NonEmpty (..), nonEmpty)
+import Data.List.NonEmpty (NonEmpty (..))
import qualified Data.List.NonEmpty as NE
{-
@@ -510,6 +510,16 @@ of the file.
at any of the definitions. This is done by Simplify.simplRecBind,
when it calls addLetIdInfo.
+Note [TailUsageDetails when forming Rec groups]
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+The `TailUsageDetails` stored in the `nd_uds` field of a `NodeDetails` is
+computed by `occAnalLamTail` applied to the RHS, not `occAnalExpr`.
+That is because the binding might still become a *non-recursive join point* in
+the AcyclicSCC case of dependency analysis!
+Hence we do the delayed `adjustTailUsage` in `occAnalRec`/`tagRecBinders` to get
+a regular, adjusted UsageDetails.
+See Note [Join points and unfoldings/rules] for more details on the contract.
+
Note [Stable unfoldings]
~~~~~~~~~~~~~~~~~~~~~~~~
None of the above stuff about RULES applies to a stable unfolding
@@ -608,6 +618,65 @@ tail call with `n` arguments (counting both value and type arguments). Otherwise
'occ_tail' will be 'NoTailCallInfo'. The tail call info flows bottom-up with the
rest of 'OccInfo' until it goes on the binder.
+Note [Join arity prediction based on joinRhsArity]
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+In general, the join arity from tail occurrences of a join point (O) may be
+higher or lower than the manifest join arity of the join body (M). E.g.,
+
+ -- M > O:
+ let f x y = x + y -- M = 2
+ in if b then f 1 else f 2 -- O = 1
+ ==> { Contify for join arity 1 }
+ join f x = \y -> x + y
+ in if b then jump f 1 else jump f 2
+
+ -- M < O
+ let f = id -- M = 0
+ in if ... then f 12 else f 13 -- O = 1
+ ==> { Contify for join arity 1, eta-expand f }
+ join f x = id x
+ in if b then jump f 12 else jump f 13
+
+But for *recursive* let, it is crucial that both arities match up, consider
+
+ letrec f x y = if ... then f x else True
+ in f 42
+
+Here, M=2 but O=1. If we settled for a joinrec arity of 1, the recursive jump
+would not happen in a tail context! Contification is invalid here.
+So indeed it is crucial to demand that M=O.
+
+(Side note: Actually, we could be more specific: Let O1 be the join arity of
+occurrences from the letrec RHS and O2 the join arity from the let body. Then
+we need M=O1 and M<=O2 and could simply eta-expand the RHS to match O2 later.
+M=O is the specific case where we don't want to eta-expand. Neither the join
+points paper nor GHC does this at the moment.)
+
+We can capitalise on this observation and conclude that *if* f could become a
+joinrec (without eta-expansion), it will have join arity M.
+Now, M is just the result of 'joinRhsArity', a rather simple, local analysis.
+It is also the join arity inside the 'TailUsageDetails' returned by
+'occAnalLamTail', so we can predict join arity without doing any fixed-point
+iteration or really doing any deep traversal of let body or RHS at all.
+We check for M in the 'adjustTailUsage' call inside 'tagRecBinders'.
+
+All this is quite apparent if you look at the contification transformation in
+Fig. 5 of "Compiling without Continuations" (which does not account for
+eta-expansion at all, mind you). The letrec case looks like this
+
+ letrec f = /\as.\xs. L[us] in L'[es]
+ ... and a bunch of conditions establishing that f only occurs
+ in app heads of join arity (len as + len xs) inside us and es ...
+
+The syntactic form `/\as.\xs. L[us]` forces M=O iff `f` occurs in `us`. However,
+for non-recursive functions, this is the definition of contification from the
+paper:
+
+ let f = /\as.\xs.u in L[es] ... conditions ...
+
+Note that u could be a lambda itself, as we have seen. No relationship between M
+and O to exploit here.
+
Note [Join points and unfoldings/rules]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Consider
@@ -618,8 +687,10 @@ Consider
Before j is inlined, we'll have occurrences of j2 in
both j's RHS and in its stable unfolding. We want to discover
-j2 as a join point. So we must do the adjustRhsUsage thing
-on j's RHS. That's why we pass mb_join_arity to calcUnfolding.
+j2 as a join point. So 'occAnalUnfolding' returns an unadjusted
+'TailUsageDetails', like 'occAnalLamTail'. We adjust the usage details of the
+unfolding to the actual join arity using the same 'adjustTailArity' as for
+the RHS, see Note [Adjusting right-hand sides].
Same with rules. Suppose we have:
@@ -636,14 +707,31 @@ up. So provided the join-point arity of k matches the args of the
rule we can allow the tail-call info from the RHS of the rule to
propagate.
-* Wrinkle for Rec case. In the recursive case we don't know the
- join-point arity in advance, when calling occAnalUnfolding and
- occAnalRules. (See makeNode.) We don't want to pass Nothing,
- because then a recursive joinrec might lose its join-poin-hood
- when SpecConstr adds a RULE. So we just make do with the
- *current* join-poin-hood, stored in the Id.
+* Note that the join arity of the RHS and that of the unfolding or RULE might
+ mismatch:
+
+ let j x y = j2 (x+x)
+ {-# INLINE[2] j = \x. g #-}
+ {-# RULE forall x y z. j x y z = h 17 #-}
+ in j 1 2
- In the non-recursive case things are simple: see occAnalNonRecBind
+ So it is crucial that we adjust each TailUsageDetails individually
+ with the actual join arity 2 here before we combine with `andUDs`.
+ Here, that means losing tail call info on `g` and `h`.
+
+* Wrinkle for Rec case: We store one TailUsageDetails in the node Details for
+ RHS, unfolding and RULE combined. Clearly, if they don't agree on their join
+ arity, we have to do some adjusting. We choose to adjust to the join arity
+ of the RHS, because that is likely the join arity that the join point will
+ have; see Note [Join arity prediction based on joinRhsArity].
+
+ If the guess is correct, then tail calls in the RHS are preserved; a necessary
+ condition for the whole binding becoming a joinrec.
+ The guess can only be incorrect in the 'AcyclicSCC' case when the binding
+ becomes a non-recursive join point with a different join arity. But then the
+ eventual call to 'adjustTailUsage' in 'tagRecBinders'/'occAnalRec' will
+ be with a different join arity and destroy unsound tail call info with
+ 'markNonTail'.
* Wrinkle for RULES. Suppose the example was a bit different:
let j :: Int -> Int
@@ -669,28 +757,21 @@ propagate.
This appears to be very rare in practice. TODO Perhaps we should gather
statistics to be sure.
-Note [Unfoldings and join points]
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-We assume that anything in an unfolding occurs multiple times, since
-unfoldings are often copied (that's the whole point!). But we still
-need to track tail calls for the purpose of finding join points.
-
-
------------------------------------------------------------
Note [Adjusting right-hand sides]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
There's a bit of a dance we need to do after analysing a lambda expression or
a right-hand side. In particular, we need to
- a) call 'markAllInsideLam' *unless* the binding is for a thunk, a one-shot
- lambda, or a non-recursive join point; and
- b) call 'markAllNonTail' *unless* the binding is for a join point, and
- the RHS has the right arity; e.g.
+ a) call 'markAllNonTail' *unless* the binding is for a join point, and
+ the TailUsageDetails from the RHS has the right join arity; e.g.
join j x y = case ... of
A -> j2 p
B -> j2 q
in j a b
Here we want the tail calls to j2 to be tail calls of the whole expression
+ b) call 'markAllInsideLam' *unless* the binding is for a thunk, a one-shot
+ lambda, or a non-recursive join point
Some examples, with how the free occurrences in e (assumed not to be a value
lambda) get marked:
@@ -707,26 +788,39 @@ lambda) get marked:
There are a few other caveats; most importantly, if we're marking a binding as
'AlwaysTailCalled', it's *going* to be a join point, so we treat it as one so
that the effect cascades properly. Consequently, at the time the RHS is
-analysed, we won't know what adjustments to make; thus 'occAnalLamOrRhs' must
-return the unadjusted 'UsageDetails', to be adjusted by 'adjustRhsUsage' once
-join-point-hood has been decided.
-
-Thus the overall sequence taking place in 'occAnalNonRecBind' and
-'occAnalRecBind' is as follows:
-
- 1. Call 'occAnalLamOrRhs' to find usage information for the RHS.
- 2. Call 'tagNonRecBinder' or 'tagRecBinders', which decides whether to make
+analysed, we won't know what adjustments to make; thus 'occAnalLamTail' must
+return the unadjusted 'TailUsageDetails', to be adjusted by 'adjustTailUsage'
+once join-point-hood has been decided and eventual one-shot annotations have
+been added through 'markNonRecJoinOneShots'.
+
+It is not so simple to see that 'occAnalNonRecBind' and 'occAnalRecBind' indeed
+perform a similar sequence of steps. Thus, here is an interleaving of events
+of both functions, serving as a specification:
+
+ 1. Call 'occAnalLamTail' to find usage information for the RHS.
+ Recursive case: 'makeNode'
+ Non-recursive case: 'occAnalNonRecBind'
+ 2. (Analyse the binding's scope. Done in 'occAnalBind'/`occAnal Let{}`.
+ Same whether recursive or not.)
+ 3. Call 'tagNonRecBinder' or 'tagRecBinders', which decides whether to make
the binding a join point.
- 3. Call 'adjustRhsUsage' accordingly. (Done as part of 'tagRecBinders' when
- recursive.)
-
-(In the recursive case, this logic is spread between 'makeNode' and
-'occAnalRec'.)
+ Cyclic Recursive case: 'mkLoopBreakerNodes'
+ Acyclic Recursive case: `occAnalRec AcyclicSCC{}`
+ Non-recursive case: 'occAnalNonRecBind'
+ 4. Non-recursive join point: Call 'markNonRecJoinOneShots' so that e.g.,
+ FloatOut sees one-shot annotations on lambdas
+ Acyclic Recursive case: `occAnalRec AcyclicSCC{}` calls 'adjustNonRecRhs'
+ Non-recursive case: 'occAnalNonRecBind' calls 'adjustNonRecRhs'
+ 5. Call 'adjustTailUsage' accordingly.
+ Cyclic Recursive case: 'tagRecBinders'
+ Acyclic Recursive case: 'adjustNonRecRhs'
+ Non-recursive case: 'adjustNonRecRhs'
-}
-
data WithUsageDetails a = WithUsageDetails !UsageDetails !a
+data WithTailUsageDetails a = WithTailUsageDetails !TailUsageDetails !a
+
------------------------------------------------------------------
-- occAnalBind
------------------------------------------------------------------
@@ -750,19 +844,17 @@ occAnalNonRecBind !env lvl imp_rule_edges bndr rhs body_usage
| isTyVar bndr -- A type let; we don't gather usage info
= WithUsageDetails body_usage [NonRec bndr rhs]
- | not (bndr `usedIn` body_usage) -- It's not mentioned
- = WithUsageDetails body_usage []
+ | not (bndr `usedIn` body_usage)
+ = WithUsageDetails body_usage [] -- See Note [Dead code]
| otherwise -- It's mentioned in the body
- = WithUsageDetails (body_usage' `andUDs` rhs_usage) [NonRec final_bndr rhs']
+ = WithUsageDetails (body_usage' `andUDs` rhs_usage) [NonRec final_bndr final_rhs]
where
- (body_usage', tagged_bndr) = tagNonRecBinder lvl body_usage bndr
- final_bndr = tagged_bndr `setIdUnfolding` unf'
- `setIdSpecialisation` mkRuleInfo rules'
- rhs_usage = rhs_uds `andUDs` unf_uds `andUDs` rule_uds
+ WithUsageDetails body_usage' tagged_bndr = tagNonRecBinder lvl body_usage bndr
-- Get the join info from the *new* decision
-- See Note [Join points and unfoldings/rules]
+ -- => join arity O of Note [Join arity prediction based on joinRhsArity]
mb_join_arity = willBeJoinId_maybe tagged_bndr
is_join_point = isJust mb_join_arity
@@ -773,17 +865,28 @@ occAnalNonRecBind !env lvl imp_rule_edges bndr rhs body_usage
-- See Note [Sources of one-shot information]
rhs_env = env1 { occ_one_shots = argOneShots dmd }
- (WithUsageDetails rhs_uds rhs') = occAnalRhs rhs_env NonRecursive mb_join_arity rhs
+ -- See Note [Join arity prediction based on joinRhsArity]
+ -- Match join arity O from mb_join_arity with manifest join arity M as
+ -- returned by of occAnalLamTail. It's totally OK for them to mismatch;
+ -- hence adjust the UDs from the RHS
+ WithUsageDetails adj_rhs_uds final_rhs
+ = adjustNonRecRhs mb_join_arity $ occAnalLamTail rhs_env rhs
+ rhs_usage = adj_rhs_uds `andUDs` adj_unf_uds `andUDs` adj_rule_uds
+ final_bndr = tagged_bndr `setIdSpecialisation` mkRuleInfo rules'
+ `setIdUnfolding` unf2
--------- Unfolding ---------
- -- See Note [Unfoldings and join points]
+ -- See Note [Join points and unfoldings/rules]
unf | isId bndr = idUnfolding bndr
| otherwise = NoUnfolding
- (WithUsageDetails unf_uds unf') = occAnalUnfolding rhs_env NonRecursive mb_join_arity unf
+ WithTailUsageDetails unf_uds unf1 = occAnalUnfolding rhs_env unf
+ unf2 = markNonRecUnfoldingOneShots mb_join_arity unf1
+ adj_unf_uds = adjustTailArity mb_join_arity unf_uds
--------- Rules ---------
-- See Note [Rules are extra RHSs] and Note [Rule dependency info]
- rules_w_uds = occAnalRules rhs_env mb_join_arity bndr
+ -- and Note [Join points and unfoldings/rules]
+ rules_w_uds = occAnalRules rhs_env bndr
rules' = map fstOf3 rules_w_uds
imp_rule_uds = impRulesScopeUsage (lookupImpRules imp_rule_edges bndr)
-- imp_rule_uds: consider
@@ -794,8 +897,9 @@ occAnalNonRecBind !env lvl imp_rule_edges bndr rhs body_usage
-- that g is (since the RULE might turn g into h), so
-- we make g mention h.
- rule_uds = foldr add_rule_uds imp_rule_uds rules_w_uds
- add_rule_uds (_, l, r) uds = l `andUDs` r `andUDs` uds
+ adj_rule_uds = foldr add_rule_uds imp_rule_uds rules_w_uds
+ add_rule_uds (_, l, r) uds
+ = l `andUDs` adjustTailArity mb_join_arity r `andUDs` uds
----------
occ = idOccInfo tagged_bndr
@@ -820,7 +924,7 @@ occAnalRecBind :: OccEnv -> TopLevelFlag -> ImpRuleEdges -> [(Var,CoreExpr)]
occAnalRecBind !env lvl imp_rule_edges pairs body_usage
= foldr (occAnalRec rhs_env lvl) (WithUsageDetails body_usage []) sccs
where
- sccs :: [SCC Details]
+ sccs :: [SCC NodeDetails]
sccs = {-# SCC "occAnalBind.scc" #-}
stronglyConnCompFromEdgedVerticesUniq nodes
@@ -832,48 +936,62 @@ occAnalRecBind !env lvl imp_rule_edges pairs body_usage
bndr_set = mkVarSet bndrs
rhs_env = env `addInScope` bndrs
+adjustNonRecRhs :: Maybe JoinArity -> WithTailUsageDetails CoreExpr -> WithUsageDetails CoreExpr
+-- ^ This function concentrates shared logic between occAnalNonRecBind and the
+-- AcyclicSCC case of occAnalRec.
+-- * It applies 'markNonRecJoinOneShots' to the RHS
+-- * and returns the adjusted rhs UsageDetails combined with the body usage
+adjustNonRecRhs mb_join_arity (WithTailUsageDetails rhs_tuds rhs)
+ = WithUsageDetails rhs_uds' rhs'
+ where
+ --------- Marking (non-rec) join binders one-shot ---------
+ !rhs' | Just ja <- mb_join_arity = markNonRecJoinOneShots ja rhs
+ | otherwise = rhs
+ --------- Adjusting right-hand side usage ---------
+ rhs_uds' = adjustTailUsage mb_join_arity rhs' rhs_tuds
+
+bindersOfSCC :: SCC NodeDetails -> [Var]
+bindersOfSCC (AcyclicSCC nd) = [nd_bndr nd]
+bindersOfSCC (CyclicSCC ds) = map nd_bndr ds
-----------------------------
occAnalRec :: OccEnv -> TopLevelFlag
- -> SCC Details
+ -> SCC NodeDetails
-> WithUsageDetails [CoreBind]
-> WithUsageDetails [CoreBind]
- -- The NonRec case is just like a Let (NonRec ...) above
-occAnalRec !_ lvl (AcyclicSCC (ND { nd_bndr = bndr, nd_rhs = rhs
- , nd_uds = rhs_uds }))
- (WithUsageDetails body_uds binds)
- | not (bndr `usedIn` body_uds)
- = WithUsageDetails body_uds binds -- See Note [Dead code]
+-- Check for Note [Dead code]
+-- NB: Only look at body_uds, ignoring uses in the SCC
+occAnalRec !_ _ scc (WithUsageDetails body_uds binds)
+ | not (any (`usedIn` body_uds) (bindersOfSCC scc))
+ = WithUsageDetails body_uds binds
- | otherwise -- It's mentioned in the body
- = WithUsageDetails (body_uds' `andUDs` rhs_uds')
- (NonRec tagged_bndr rhs : binds)
+-- The NonRec case is just like a Let (NonRec ...) above
+occAnalRec !_ lvl
+ (AcyclicSCC (ND { nd_bndr = bndr, nd_rhs = wtuds }))
+ (WithUsageDetails body_uds binds)
+ = WithUsageDetails (body_uds' `andUDs` rhs_uds') (NonRec bndr' rhs' : binds)
where
- (body_uds', tagged_bndr) = tagNonRecBinder lvl body_uds bndr
- rhs_uds' = adjustRhsUsage mb_join_arity rhs rhs_uds
+ WithUsageDetails body_uds' tagged_bndr = tagNonRecBinder lvl body_uds bndr
mb_join_arity = willBeJoinId_maybe tagged_bndr
+ WithUsageDetails rhs_uds' rhs' = adjustNonRecRhs mb_join_arity wtuds
+ !unf' = markNonRecUnfoldingOneShots mb_join_arity (idUnfolding tagged_bndr)
+ !bndr' = tagged_bndr `setIdUnfolding` unf'
- -- The Rec case is the interesting one
- -- See Note [Recursive bindings: the grand plan]
- -- See Note [Loop breaking]
+-- The Rec case is the interesting one
+-- See Note [Recursive bindings: the grand plan]
+-- See Note [Loop breaking]
occAnalRec env lvl (CyclicSCC details_s) (WithUsageDetails body_uds binds)
- | not (any (`usedIn` body_uds) bndrs) -- NB: look at body_uds, not total_uds
- = WithUsageDetails body_uds binds -- See Note [Dead code]
-
- | otherwise -- At this point we always build a single Rec
= -- pprTrace "occAnalRec" (ppr loop_breaker_nodes)
WithUsageDetails final_uds (Rec pairs : binds)
-
where
- bndrs = map nd_bndr details_s
all_simple = all nd_simple details_s
------------------------------
-- Make the nodes for the loop-breaker analysis
-- See Note [Choosing loop breakers] for loop_breaker_nodes
final_uds :: UsageDetails
- loop_breaker_nodes :: [LetrecNode]
+ loop_breaker_nodes :: [LoopBreakerNode]
(WithUsageDetails final_uds loop_breaker_nodes) = mkLoopBreakerNodes env lvl body_uds details_s
------------------------------
@@ -1102,7 +1220,7 @@ type Binding = (Id,CoreExpr)
loopBreakNodes :: Int
-> VarSet -- Binders whose dependencies may be "missing"
-- See Note [Weak loop breakers]
- -> [LetrecNode]
+ -> [LoopBreakerNode]
-> [Binding] -- Append these to the end
-> [Binding]
@@ -1121,7 +1239,7 @@ loopBreakNodes depth weak_fvs nodes binds
CyclicSCC nodes -> reOrderNodes depth weak_fvs nodes binds
----------------------------------
-reOrderNodes :: Int -> VarSet -> [LetrecNode] -> [Binding] -> [Binding]
+reOrderNodes :: Int -> VarSet -> [LoopBreakerNode] -> [Binding] -> [Binding]
-- Choose a loop breaker, mark it no-inline,
-- and call loopBreakNodes on the rest
reOrderNodes _ _ [] _ = panic "reOrderNodes"
@@ -1133,7 +1251,7 @@ reOrderNodes depth weak_fvs (node : nodes) binds
(map (nodeBinding mk_loop_breaker) chosen_nodes ++ binds)
where
(chosen_nodes, unchosen) = chooseLoopBreaker approximate_lb
- (nd_score (node_payload node))
+ (snd_score (node_payload node))
[node] [] nodes
approximate_lb = depth >= 2
@@ -1142,8 +1260,8 @@ reOrderNodes depth weak_fvs (node : nodes) binds
-- After two iterations (d=0, d=1) give up
-- and approximate, returning to d=0
-nodeBinding :: (Id -> Id) -> LetrecNode -> Binding
-nodeBinding set_id_occ (node_payload -> ND { nd_bndr = bndr, nd_rhs = rhs})
+nodeBinding :: (Id -> Id) -> LoopBreakerNode -> Binding
+nodeBinding set_id_occ (node_payload -> SND { snd_bndr = bndr, snd_rhs = rhs})
= (set_id_occ bndr, rhs)
mk_loop_breaker :: Id -> Id
@@ -1163,13 +1281,13 @@ mk_non_loop_breaker weak_fvs bndr
tail_info = tailCallInfo (idOccInfo bndr)
----------------------------------
-chooseLoopBreaker :: Bool -- True <=> Too many iterations,
- -- so approximate
- -> NodeScore -- Best score so far
- -> [LetrecNode] -- Nodes with this score
- -> [LetrecNode] -- Nodes with higher scores
- -> [LetrecNode] -- Unprocessed nodes
- -> ([LetrecNode], [LetrecNode])
+chooseLoopBreaker :: Bool -- True <=> Too many iterations,
+ -- so approximate
+ -> NodeScore -- Best score so far
+ -> [LoopBreakerNode] -- Nodes with this score
+ -> [LoopBreakerNode] -- Nodes with higher scores
+ -> [LoopBreakerNode] -- Unprocessed nodes
+ -> ([LoopBreakerNode], [LoopBreakerNode])
-- This loop looks for the bind with the lowest score
-- to pick as the loop breaker. The rest accumulate in
chooseLoopBreaker _ _ loop_nodes acc []
@@ -1189,7 +1307,7 @@ chooseLoopBreaker approx_lb loop_sc loop_nodes acc (node : nodes)
| otherwise -- Worse score so don't pick it
= chooseLoopBreaker approx_lb loop_sc loop_nodes (node : acc) nodes
where
- sc = nd_score (node_payload node)
+ sc = snd_score (node_payload node)
{-
Note [Complexity of loop breaking]
@@ -1322,16 +1440,21 @@ ToDo: try using the occurrence info for the inline'd binder.
************************************************************************
-}
-type LetrecNode = Node Unique Details -- Node comes from Digraph
- -- The Unique key is gotten from the Id
-data Details
- = ND { nd_bndr :: Id -- Binder
+-- | Digraph node as constructed by 'makeNode' and consumed by 'occAnalRec'.
+-- The Unique key is gotten from the Id.
+type LetrecNode = Node Unique NodeDetails
- , nd_rhs :: CoreExpr -- RHS, already occ-analysed
+-- | Node details as consumed by 'occAnalRec'.
+data NodeDetails
+ = ND { nd_bndr :: Id -- Binder
- , nd_uds :: UsageDetails -- Usage from RHS, and RULES, and stable unfoldings
- -- ignoring phase (ie assuming all are active)
- -- See Note [Forming Rec groups]
+ , nd_rhs :: !(WithTailUsageDetails CoreExpr)
+ -- ^ RHS, already occ-analysed
+ -- With TailUsageDetails from RHS, and RULES, and stable unfoldings,
+ -- ignoring phase (ie assuming all are active).
+ -- NB: Unadjusted TailUsageDetails, as if this Node becomes a
+ -- non-recursive join point!
+ -- See Note [TailUsageDetails when forming Rec groups]
, nd_inl :: IdSet -- Free variables of the stable unfolding and the RHS
-- but excluding any RULES
@@ -1348,18 +1471,33 @@ data Details
, nd_active_rule_fvs :: IdSet -- Variables bound in this Rec group that are free
-- in the RHS of an active rule for this bndr
-- See Note [Rules and loop breakers]
-
- , nd_score :: NodeScore
}
-instance Outputable Details where
+instance Outputable NodeDetails where
ppr nd = text "ND" <> braces
(sep [ text "bndr =" <+> ppr (nd_bndr nd)
- , text "uds =" <+> ppr (nd_uds nd)
+ , text "uds =" <+> ppr uds
, text "inl =" <+> ppr (nd_inl nd)
, text "simple =" <+> ppr (nd_simple nd)
, text "active_rule_fvs =" <+> ppr (nd_active_rule_fvs nd)
- , text "score =" <+> ppr (nd_score nd)
+ ])
+ where WithTailUsageDetails uds _ = nd_rhs nd
+
+-- | Digraph with simplified and completely occurrence analysed
+-- 'SimpleNodeDetails', retaining just the info we need for breaking loops.
+type LoopBreakerNode = Node Unique SimpleNodeDetails
+
+-- | Condensed variant of 'NodeDetails' needed during loop breaking.
+data SimpleNodeDetails
+ = SND { snd_bndr :: IdWithOccInfo -- OccInfo accurate
+ , snd_rhs :: CoreExpr -- properly occur-analysed
+ , snd_score :: NodeScore
+ }
+
+instance Outputable SimpleNodeDetails where
+ ppr nd = text "SND" <> braces
+ (sep [ text "bndr =" <+> ppr (snd_bndr nd)
+ , text "score =" <+> ppr (snd_score nd)
])
-- The NodeScore is compared lexicographically;
@@ -1387,52 +1525,59 @@ makeNode !env imp_rule_edges bndr_set (bndr, rhs)
-- explained in Note [Deterministic SCC] in GHC.Data.Graph.Directed.
where
details = ND { nd_bndr = bndr'
- , nd_rhs = rhs'
- , nd_uds = scope_uds
+ , nd_rhs = WithTailUsageDetails scope_uds rhs'
, nd_inl = inl_fvs
, nd_simple = null rules_w_uds && null imp_rule_info
, nd_weak_fvs = weak_fvs
- , nd_active_rule_fvs = active_rule_fvs
- , nd_score = pprPanic "makeNodeDetails" (ppr bndr) }
+ , nd_active_rule_fvs = active_rule_fvs }
bndr' = bndr `setIdUnfolding` unf'
`setIdSpecialisation` mkRuleInfo rules'
- inl_uds = rhs_uds `andUDs` unf_uds
- scope_uds = inl_uds `andUDs` rule_uds
+ -- NB: Both adj_unf_uds and adj_rule_uds have been adjusted to match the
+ -- JoinArity rhs_ja of unadj_rhs_uds.
+ unadj_inl_uds = unadj_rhs_uds `andUDs` adj_unf_uds
+ unadj_scope_uds = unadj_inl_uds `andUDs` adj_rule_uds
+ scope_uds = TUD rhs_ja unadj_scope_uds
-- Note [Rules are extra RHSs]
-- Note [Rule dependency info]
- scope_fvs = udFreeVars bndr_set scope_uds
+ scope_fvs = udFreeVars bndr_set unadj_scope_uds
-- scope_fvs: all occurrences from this binder: RHS, unfolding,
-- and RULES, both LHS and RHS thereof, active or inactive
- inl_fvs = udFreeVars bndr_set inl_uds
+ inl_fvs = udFreeVars bndr_set unadj_inl_uds
-- inl_fvs: vars that would become free if the function was inlined.
-- We conservatively approximate that by thefree vars from the RHS
-- and the unfolding together.
-- See Note [inl_fvs]
- mb_join_arity = isJoinId_maybe bndr
- -- Get join point info from the *current* decision
- -- We don't know what the new decision will be!
- -- Using the old decision at least allows us to
- -- preserve existing join point, even RULEs are added
- -- See Note [Join points and unfoldings/rules]
--------- Right hand side ---------
-- Constructing the edges for the main Rec computation
-- See Note [Forming Rec groups]
- -- Do not use occAnalRhs because we don't yet know the final
- -- answer for mb_join_arity; instead, do the occAnalLam call from
- -- occAnalRhs, and postpone adjustRhsUsage until occAnalRec
- rhs_env = rhsCtxt env
- (WithUsageDetails rhs_uds rhs') = occAnalLam rhs_env rhs
+ -- and Note [TailUsageDetails when forming Rec groups]
+ -- Compared to occAnalNonRecBind, we can't yet adjust the RHS because
+ -- (a) we don't yet know the final joinpointhood. It might not become a
+ -- join point after all!
+ -- (b) we don't even know whether it stays a recursive RHS after the SCC
+ -- analysis we are about to seed! So we can't markAllInsideLam in
+ -- advance, because if it ends up as a non-recursive join point we'll
+ -- consider it as one-shot and don't need to markAllInsideLam.
+ -- Instead, do the occAnalLamTail call here and postpone adjustTailUsage
+ -- until occAnalRec. In effect, we pretend that the RHS becomes a
+ -- non-recursive join point and fix up later with adjustTailUsage.
+ rhs_env = rhsCtxt env
+ WithTailUsageDetails (TUD rhs_ja unadj_rhs_uds) rhs' = occAnalLamTail rhs_env rhs
+ -- corresponding call to adjustTailUsage in occAnalRec and tagRecBinders
--------- Unfolding ---------
- -- See Note [Unfoldings and join points]
+ -- See Note [Join points and unfoldings/rules]
unf = realIdUnfolding bndr -- realIdUnfolding: Ignore loop-breaker-ness
-- here because that is what we are setting!
- (WithUsageDetails unf_uds unf') = occAnalUnfolding rhs_env Recursive mb_join_arity unf
+ WithTailUsageDetails unf_tuds unf' = occAnalUnfolding rhs_env unf
+ adj_unf_uds = adjustTailArity (Just rhs_ja) unf_tuds
+ -- `rhs_ja` is `joinRhsArity rhs` and is the prediction for source M
+ -- of Note [Join arity prediction based on joinRhsArity]
--------- IMP-RULES --------
is_active = occ_rule_act env :: Activation -> Bool
@@ -1441,11 +1586,15 @@ makeNode !env imp_rule_edges bndr_set (bndr, rhs)
imp_rule_fvs = impRulesActiveFvs is_active bndr_set imp_rule_info
--------- All rules --------
+ -- See Note [Join points and unfoldings/rules]
+ -- `rhs_ja` is `joinRhsArity rhs'` and is the prediction for source M
+ -- of Note [Join arity prediction based on joinRhsArity]
rules_w_uds :: [(CoreRule, UsageDetails, UsageDetails)]
- rules_w_uds = occAnalRules rhs_env mb_join_arity bndr
+ rules_w_uds = [ (r,l,adjustTailArity (Just rhs_ja) rhs_tuds)
+ | (r,l,rhs_tuds) <- occAnalRules rhs_env bndr ]
rules' = map fstOf3 rules_w_uds
- rule_uds = foldr add_rule_uds imp_rule_uds rules_w_uds
+ adj_rule_uds = foldr add_rule_uds imp_rule_uds rules_w_uds
add_rule_uds (_, l, r) uds = l `andUDs` r `andUDs` uds
-------- active_rule_fvs ------------
@@ -1463,8 +1612,8 @@ makeNode !env imp_rule_edges bndr_set (bndr, rhs)
mkLoopBreakerNodes :: OccEnv -> TopLevelFlag
-> UsageDetails -- for BODY of let
- -> [Details]
- -> WithUsageDetails [LetrecNode] -- adjusted
+ -> [NodeDetails]
+ -> WithUsageDetails [LoopBreakerNode] -- with OccInfo up-to-date
-- See Note [Choosing loop breakers]
-- This function primarily creates the Nodes for the
-- loop-breaker SCC analysis. More specifically:
@@ -1477,10 +1626,10 @@ mkLoopBreakerNodes :: OccEnv -> TopLevelFlag
mkLoopBreakerNodes !env lvl body_uds details_s
= WithUsageDetails final_uds (zipWithEqual "mkLoopBreakerNodes" mk_lb_node details_s bndrs')
where
- (final_uds, bndrs') = tagRecBinders lvl body_uds details_s
+ WithUsageDetails final_uds bndrs' = tagRecBinders lvl body_uds details_s
mk_lb_node nd@(ND { nd_bndr = old_bndr, nd_inl = inl_fvs }) new_bndr
- = DigraphNode { node_payload = new_nd
+ = DigraphNode { node_payload = simple_nd
, node_key = varUnique old_bndr
, node_dependencies = nonDetKeysUniqSet lb_deps }
-- It's OK to use nonDetKeysUniqSet here as
@@ -1488,7 +1637,8 @@ mkLoopBreakerNodes !env lvl body_uds details_s
-- in nondeterministic order as explained in
-- Note [Deterministic SCC] in GHC.Data.Graph.Directed.
where
- new_nd = nd { nd_bndr = new_bndr, nd_score = score }
+ WithTailUsageDetails _ rhs = nd_rhs nd
+ simple_nd = SND { snd_bndr = new_bndr, snd_rhs = rhs, snd_score = score }
score = nodeScore env new_bndr lb_deps nd
lb_deps = extendFvs_ rule_fv_env inl_fvs
-- See Note [Loop breaker dependencies]
@@ -1524,10 +1674,10 @@ group { f1 = e1; ...; fn = en } are:
nodeScore :: OccEnv
-> Id -- Binder with new occ-info
-> VarSet -- Loop-breaker dependencies
- -> Details
+ -> NodeDetails
-> NodeScore
nodeScore !env new_bndr lb_deps
- (ND { nd_bndr = old_bndr, nd_rhs = bind_rhs })
+ (ND { nd_bndr = old_bndr, nd_rhs = WithTailUsageDetails _ bind_rhs })
| not (isId old_bndr) -- A type or coercion variable is never a loop breaker
= (100, 0, False)
@@ -1748,7 +1898,7 @@ lambda and casts, e.g.
* Occurrence analyser: we just mark each binder in the lambda-group
(here: x,y,z) with its occurrence info in the *body* of the
- lambda-group. See occAnalLam.
+ lambda-group. See occAnalLamTail.
* Simplifier. The simplifier is careful when partially applying
lambda-groups. See the call to zapLambdaBndrs in
@@ -1804,25 +1954,31 @@ zapLambdaBndrs fun arg_count
zap_bndr b | isTyVar b = b
| otherwise = zapLamIdInfo b
-occAnalLam :: OccEnv -> CoreExpr -> (WithUsageDetails CoreExpr)
--- See Note [Occurrence analysis for lambda binders]
+occAnalLamTail :: OccEnv -> CoreExpr -> WithTailUsageDetails CoreExpr
+-- ^ See Note [Occurrence analysis for lambda binders].
-- It does the following:
-- * Sets one-shot info on the lambda binder from the OccEnv, and
-- removes that one-shot info from the OccEnv
-- * Sets the OccEnv to OccVanilla when going under a value lambda
-- * Tags each lambda with its occurrence information
-- * Walks through casts
+-- * Package up the analysed lambda with its manifest join arity
+--
-- This function does /not/ do
-- markAllInsideLam or
-- markAllNonTail
--- The caller does that, either in occAnal (Lam {}), or in adjustRhsUsage
+-- The caller does that, via adjustTailUsage (mostly calls go through
+-- adjustNonRecRhs). Every call to occAnalLamTail must ultimately call
+-- adjustTailUsage to discharge the assumed join arity.
+--
+-- In effect, the analysis result is for a non-recursive join point with
+-- manifest arity and adjustTailUsage does the fixup.
-- See Note [Adjusting right-hand sides]
-
-occAnalLam env (Lam bndr expr)
+occAnalLamTail env (Lam bndr expr)
| isTyVar bndr
- = let env1 = addOneInScope env bndr
- WithUsageDetails usage expr' = occAnalLam env1 expr
- in WithUsageDetails usage (Lam bndr expr')
+ , let env1 = addOneInScope env bndr
+ , WithTailUsageDetails (TUD ja usage) expr' <- occAnalLamTail env1 expr
+ = WithTailUsageDetails (TUD (ja+1) usage) (Lam bndr expr')
-- Important: Keep the 'env' unchanged so that with a RHS like
-- \(@ x) -> K @x (f @x)
-- we'll see that (K @x (f @x)) is in a OccRhs, and hence refrain
@@ -1840,14 +1996,14 @@ occAnalLam env (Lam bndr expr)
env1 = env { occ_encl = OccVanilla, occ_one_shots = env_one_shots' }
env2 = addOneInScope env1 bndr
- (WithUsageDetails usage expr') = occAnalLam env2 expr
+ WithTailUsageDetails (TUD ja usage) expr' = occAnalLamTail env2 expr
(usage', bndr2) = tagLamBinder usage bndr1
- in WithUsageDetails usage' (Lam bndr2 expr')
+ in WithTailUsageDetails (TUD (ja+1) usage') (Lam bndr2 expr')
-- For casts, keep going in the same lambda-group
-- See Note [Occurrence analysis for lambda binders]
-occAnalLam env (Cast expr co)
- = let (WithUsageDetails usage expr') = occAnalLam env expr
+occAnalLamTail env (Cast expr co)
+ = let WithTailUsageDetails (TUD ja usage) expr' = occAnalLamTail env expr
-- usage1: see Note [Gather occurrences of coercion variables]
usage1 = addManyOccs usage (coVarsOfCo co)
@@ -1857,15 +2013,16 @@ occAnalLam env (Cast expr co)
_ -> usage1
-- usage3: you might think this was not necessary, because of
- -- the markAllNonTail in adjustRhsUsage; but not so! For a
- -- join point, adjustRhsUsage doesn't do this; yet if there is
+ -- the markAllNonTail in adjustTailUsage; but not so! For a
+ -- join point, adjustTailUsage doesn't do this; yet if there is
-- a cast, we must! Also: why markAllNonTail? See
-- GHC.Core.Lint: Note Note [Join points and casts]
usage3 = markAllNonTail usage2
- in WithUsageDetails usage3 (Cast expr' co)
+ in WithTailUsageDetails (TUD ja usage3) (Cast expr' co)
-occAnalLam env expr = occAnal env expr
+occAnalLamTail env expr = case occAnal env expr of
+ WithUsageDetails usage expr' -> WithTailUsageDetails (TUD 0 usage) expr'
{- Note [Occ-anal and cast worker/wrapper]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@@ -1885,8 +2042,8 @@ RHS. So it'll get a Many occ-info. (Maybe Cast w/w should create a stable
unfolding, which would obviate this Note; but that seems a bit of a
heavyweight solution.)
-We only need to this in occAnalLam, not occAnal, because the top leve
-of a right hand side is handled by occAnalLam.
+We only need to this in occAnalLamTail, not occAnal, because the top leve
+of a right hand side is handled by occAnalLamTail.
-}
@@ -1896,57 +2053,23 @@ of a right hand side is handled by occAnalLam.
* *
********************************************************************* -}
-occAnalRhs :: OccEnv -> RecFlag -> Maybe JoinArity
- -> CoreExpr -- RHS
- -> WithUsageDetails CoreExpr
-occAnalRhs !env is_rec mb_join_arity rhs
- = let (WithUsageDetails usage rhs1) = occAnalLam env rhs
- -- We call occAnalLam here, not occAnalExpr, so that it doesn't
- -- do the markAllInsideLam and markNonTailCall stuff before
- -- we've had a chance to help with join points; that comes next
- rhs2 = markJoinOneShots is_rec mb_join_arity rhs1
- rhs_usage = adjustRhsUsage mb_join_arity rhs2 usage
- in WithUsageDetails rhs_usage rhs2
-
-
-
-markJoinOneShots :: RecFlag -> Maybe JoinArity -> CoreExpr -> CoreExpr
--- For a /non-recursive/ join point we can mark all
--- its join-lambda as one-shot; and it's a good idea to do so
-markJoinOneShots NonRecursive (Just join_arity) rhs
- = go join_arity rhs
- where
- go 0 rhs = rhs
- go n (Lam b rhs) = Lam (if isId b then setOneShotLambda b else b)
- (go (n-1) rhs)
- go _ rhs = rhs -- Not enough lambdas. This can legitimately happen.
- -- e.g. let j = case ... in j True
- -- This will become an arity-1 join point after the
- -- simplifier has eta-expanded it; but it may not have
- -- enough lambdas /yet/. (Lint checks that JoinIds do
- -- have enough lambdas.)
-markJoinOneShots _ _ rhs
- = rhs
-
occAnalUnfolding :: OccEnv
- -> RecFlag
- -> Maybe JoinArity -- See Note [Join points and unfoldings/rules]
-> Unfolding
- -> WithUsageDetails Unfolding
+ -> WithTailUsageDetails Unfolding
-- Occurrence-analyse a stable unfolding;
--- discard a non-stable one altogether.
-occAnalUnfolding !env is_rec mb_join_arity unf
+-- discard a non-stable one altogether and return empty usage details.
+occAnalUnfolding !env unf
= case unf of
unf@(CoreUnfolding { uf_tmpl = rhs, uf_src = src })
| isStableSource src ->
let
- (WithUsageDetails usage rhs') = occAnalRhs env is_rec mb_join_arity rhs
+ WithTailUsageDetails (TUD rhs_ja usage) rhs' = occAnalLamTail env rhs
unf' | noBinderSwaps env = unf -- Note [Unfoldings and rules]
| otherwise = unf { uf_tmpl = rhs' }
- in WithUsageDetails (markAllMany usage) unf'
+ in WithTailUsageDetails (TUD rhs_ja (markAllMany usage)) unf'
-- markAllMany: see Note [Occurrences in stable unfoldings]
- | otherwise -> WithUsageDetails emptyDetails unf
+ | otherwise -> WithTailUsageDetails (TUD 0 emptyDetails) unf
-- For non-Stable unfoldings we leave them undisturbed, but
-- don't count their usage because the simplifier will discard them.
-- We leave them undisturbed because nodeScore uses their size info
@@ -1955,29 +2078,26 @@ occAnalUnfolding !env is_rec mb_join_arity unf
-- scope remain in scope; there is no cloning etc.
unf@(DFunUnfolding { df_bndrs = bndrs, df_args = args })
- -> WithUsageDetails final_usage (unf { df_args = args' })
+ -> WithTailUsageDetails (TUD 0 final_usage) (unf { df_args = args' })
where
env' = env `addInScope` bndrs
(WithUsageDetails usage args') = occAnalList env' args
- final_usage = markAllManyNonTail (delDetailsList usage bndrs)
- `addLamCoVarOccs` bndrs
- `delDetailsList` bndrs
+ final_usage = usage `addLamCoVarOccs` bndrs `delDetailsList` bndrs
-- delDetailsList; no need to use tagLamBinders because we
-- never inline DFuns so the occ-info on binders doesn't matter
- unf -> WithUsageDetails emptyDetails unf
+ unf -> WithTailUsageDetails (TUD 0 emptyDetails) unf
occAnalRules :: OccEnv
- -> Maybe JoinArity -- See Note [Join points and unfoldings/rules]
-> Id -- Get rules from here
-> [(CoreRule, -- Each (non-built-in) rule
UsageDetails, -- Usage details for LHS
- UsageDetails)] -- Usage details for RHS
-occAnalRules !env mb_join_arity bndr
+ TailUsageDetails)] -- Usage details for RHS
+occAnalRules !env bndr
= map occ_anal_rule (idCoreRules bndr)
where
occ_anal_rule rule@(Rule { ru_bndrs = bndrs, ru_args = args, ru_rhs = rhs })
- = (rule', lhs_uds', rhs_uds')
+ = (rule', lhs_uds', TUD rhs_ja rhs_uds')
where
env' = env `addInScope` bndrs
rule' | noBinderSwaps env = rule -- Note [Unfoldings and rules]
@@ -1990,14 +2110,11 @@ occAnalRules !env mb_join_arity bndr
(WithUsageDetails rhs_uds rhs') = occAnal env' rhs
-- Note [Rules are extra RHSs]
-- Note [Rule dependency info]
- rhs_uds' = markAllNonTailIf (not exact_join) $
- markAllMany $
+ rhs_uds' = markAllMany $
rhs_uds `delDetailsList` bndrs
+ rhs_ja = length args -- See Note [Join points and unfoldings/rules]
- exact_join = exactJoin mb_join_arity args
- -- See Note [Join points and unfoldings/rules]
-
- occ_anal_rule other_rule = (other_rule, emptyDetails, emptyDetails)
+ occ_anal_rule other_rule = (other_rule, emptyDetails, TUD 0 emptyDetails)
{- Note [Join point RHSs]
~~~~~~~~~~~~~~~~~~~~~~~~~
@@ -2032,6 +2149,8 @@ Another way to think about it: if we inlined g as-is into multiple
call sites, now there's be multiple calls to f.
Bottom line: treat all occurrences in a stable unfolding as "Many".
+We still leave tail call information intact, though, as to not spoil
+potential join points.
Note [Unfoldings and rules]
~~~~~~~~~~~~~~~~~~~~~~~~~~~
@@ -2209,10 +2328,7 @@ occAnal env app@(App _ _)
= occAnalApp env (collectArgsTicks tickishFloatable app)
occAnal env expr@(Lam {})
- = let (WithUsageDetails usage expr') = occAnalLam env expr
- final_usage = markAllInsideLamIf (not (isOneShotFun expr')) $
- markAllNonTail usage
- in WithUsageDetails final_usage expr'
+ = adjustNonRecRhs Nothing $ occAnalLamTail env expr -- mb_join_arity == Nothing <=> markAllManyNonTail
occAnal env (Case scrut bndr ty alts)
= let
@@ -2287,7 +2403,7 @@ occAnalApp !env (Var fun, args, ticks)
-- This caused #18296
| fun `hasKey` runRWKey
, [t1, t2, arg] <- args
- , let (WithUsageDetails usage arg') = occAnalRhs env NonRecursive (Just 1) arg
+ , WithUsageDetails usage arg' <- adjustNonRecRhs (Just 1) $ occAnalLamTail env arg
= WithUsageDetails usage (mkTicks ticks $ mkApps (Var fun) [t1, t2, arg'])
occAnalApp env (Var fun_id, args, ticks)
@@ -2872,7 +2988,6 @@ lookupBndrSwap env@(OccEnv { occ_bs_env = bs_env }) bndr
case lookupBndrSwap env bndr1 of
(fun, fun_id) -> (mkCastMCo fun mco, fun_id) }
-
{- Historical note [Proxy let-bindings]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
We used to do the binder-swap transformation by introducing
@@ -2998,6 +3113,19 @@ data UsageDetails
instance Outputable UsageDetails where
ppr ud = ppr (ud_env (flattenUsageDetails ud))
+-- | Captures the result of applying 'occAnalLamTail' to a function `\xyz.body`.
+-- The TailUsageDetails records
+-- * the number of lambdas (including type lambdas: a JoinArity)
+-- * UsageDetails for the `body`, unadjusted by `adjustTailUsage`.
+-- If the binding turns out to be a join point with the indicated join
+-- arity, this unadjusted usage details is just what we need; otherwise we
+-- need to discard tail calls. That's what `adjustTailUsage` does.
+data TailUsageDetails = TUD !JoinArity !UsageDetails
+
+instance Outputable TailUsageDetails where
+ ppr (TUD ja uds) = lambda <> ppr ja <> ppr uds
+
+
-------------------
-- UsageDetails API
@@ -3139,24 +3267,49 @@ flattenUsageDetails ud@(UD { ud_env = env })
-------------------
-- See Note [Adjusting right-hand sides]
-adjustRhsUsage :: Maybe JoinArity
- -> CoreExpr -- Rhs, AFTER occ anal
- -> UsageDetails -- From body of lambda
+adjustTailUsage :: Maybe JoinArity
+ -> CoreExpr -- Rhs, AFTER occAnalLamTail
+ -> TailUsageDetails -- From body of lambda
-> UsageDetails
-adjustRhsUsage mb_join_arity rhs usage
+adjustTailUsage mb_join_arity rhs (TUD rhs_ja usage)
= -- c.f. occAnal (Lam {})
markAllInsideLamIf (not one_shot) $
markAllNonTailIf (not exact_join) $
usage
where
one_shot = isOneShotFun rhs
- exact_join = exactJoin mb_join_arity bndrs
- (bndrs,_) = collectBinders rhs
+ exact_join = mb_join_arity == Just rhs_ja
+
+adjustTailArity :: Maybe JoinArity -> TailUsageDetails -> UsageDetails
+adjustTailArity mb_rhs_ja (TUD ud_ja usage) =
+ markAllNonTailIf (mb_rhs_ja /= Just ud_ja) usage
+
+markNonRecJoinOneShots :: JoinArity -> CoreExpr -> CoreExpr
+-- For a /non-recursive/ join point we can mark all
+-- its join-lambda as one-shot; and it's a good idea to do so
+markNonRecJoinOneShots join_arity rhs
+ = go join_arity rhs
+ where
+ go 0 rhs = rhs
+ go n (Lam b rhs) = Lam (if isId b then setOneShotLambda b else b)
+ (go (n-1) rhs)
+ go _ rhs = rhs -- Not enough lambdas. This can legitimately happen.
+ -- e.g. let j = case ... in j True
+ -- This will become an arity-1 join point after the
+ -- simplifier has eta-expanded it; but it may not have
+ -- enough lambdas /yet/. (Lint checks that JoinIds do
+ -- have enough lambdas.)
-exactJoin :: Maybe JoinArity -> [a] -> Bool
-exactJoin Nothing _ = False
-exactJoin (Just join_arity) args = args `lengthIs` join_arity
- -- Remember join_arity includes type binders
+markNonRecUnfoldingOneShots :: Maybe JoinArity -> Unfolding -> Unfolding
+-- ^ Apply 'markNonRecJoinOneShots' to a stable unfolding
+markNonRecUnfoldingOneShots mb_join_arity unf
+ | Just ja <- mb_join_arity
+ , CoreUnfolding{uf_src=src,uf_tmpl=tmpl} <- unf
+ , isStableSource src
+ , let !tmpl' = markNonRecJoinOneShots ja tmpl
+ = unf{uf_tmpl=tmpl'}
+ | otherwise
+ = unf
type IdWithOccInfo = Id
@@ -3192,8 +3345,8 @@ tagLamBinder usage bndr
tagNonRecBinder :: TopLevelFlag -- At top level?
-> UsageDetails -- Of scope
-> CoreBndr -- Binder
- -> (UsageDetails, -- Details with binder removed
- IdWithOccInfo) -- Tagged binder
+ -> WithUsageDetails -- Details with binder removed
+ IdWithOccInfo -- Tagged binder
tagNonRecBinder lvl usage binder
= let
@@ -3205,37 +3358,34 @@ tagNonRecBinder lvl usage binder
binder' = setBinderOcc occ' binder
usage' = usage `delDetails` binder
in
- usage' `seq` (usage', binder')
+ WithUsageDetails usage' binder'
tagRecBinders :: TopLevelFlag -- At top level?
-> UsageDetails -- Of body of let ONLY
- -> [Details]
- -> (UsageDetails, -- Adjusted details for whole scope,
+ -> [NodeDetails]
+ -> WithUsageDetails -- Adjusted details for whole scope,
-- with binders removed
- [IdWithOccInfo]) -- Tagged binders
+ [IdWithOccInfo] -- Tagged binders
-- Substantially more complicated than non-recursive case. Need to adjust RHS
-- details *before* tagging binders (because the tags depend on the RHSes).
tagRecBinders lvl body_uds details_s
= let
bndrs = map nd_bndr details_s
- rhs_udss = map nd_uds details_s
-
- -- 1. Determine join-point-hood of whole group, as determined by
- -- the *unadjusted* usage details
- unadj_uds = foldr andUDs body_uds rhs_udss
- -- This is only used in `mb_join_arity`, to adjust each `Details` in `details_s`, thus,
- -- when `bndrs` is non-empty. So, we only write `maybe False` as `decideJoinPointHood`
- -- takes a `NonEmpty CoreBndr`; the default value `False` won't affect program behavior.
- will_be_joins = maybe False (decideJoinPointHood lvl unadj_uds) (nonEmpty bndrs)
+ -- 1. See Note [Join arity prediction based on joinRhsArity]
+ -- Determine possible join-point-hood of whole group, by testing for
+ -- manifest join arity M.
+ -- This (re-)asserts that makeNode had made tuds for that same arity M!
+ unadj_uds = foldr (andUDs . test_manifest_arity) body_uds details_s
+ test_manifest_arity ND{nd_rhs=WithTailUsageDetails tuds rhs}
+ = adjustTailArity (Just (joinRhsArity rhs)) tuds
- -- 2. Adjust usage details of each RHS, taking into account the
- -- join-point-hood decision
- rhs_udss' = [ adjustRhsUsage (mb_join_arity bndr) rhs rhs_uds
- | ND { nd_bndr = bndr, nd_uds = rhs_uds
- , nd_rhs = rhs } <- details_s ]
+ bndr_ne = expectNonEmpty "List of binders is never empty" bndrs
+ will_be_joins = decideJoinPointHood lvl unadj_uds bndr_ne
mb_join_arity :: Id -> Maybe JoinArity
+ -- mb_join_arity: See Note [Join arity prediction based on joinRhsArity]
+ -- This is the source O
mb_join_arity bndr
-- Can't use willBeJoinId_maybe here because we haven't tagged
-- the binder yet (the tag depends on these adjustments!)
@@ -3247,6 +3397,12 @@ tagRecBinders lvl body_uds details_s
= assert (not will_be_joins) -- Should be AlwaysTailCalled if
Nothing -- we are making join points!
+ -- 2. Adjust usage details of each RHS, taking into account the
+ -- join-point-hood decision
+ rhs_udss' = [ adjustTailUsage (mb_join_arity bndr) rhs rhs_tuds -- matching occAnalLamTail in makeNode
+ | ND { nd_bndr = bndr, nd_rhs = WithTailUsageDetails rhs_tuds rhs }
+ <- details_s ]
+
-- 3. Compute final usage details from adjusted RHS details
adj_uds = foldr andUDs body_uds rhs_udss'
@@ -3257,7 +3413,7 @@ tagRecBinders lvl body_uds details_s
-- 5. Drop the binders from the adjusted details and return
usage' = adj_uds `delDetailsList` bndrs
in
- (usage', bndrs')
+ WithUsageDetails usage' bndrs'
setBinderOcc :: OccInfo -> CoreBndr -> CoreBndr
setBinderOcc occ_info bndr
@@ -3271,12 +3427,13 @@ setBinderOcc occ_info bndr
| otherwise = setIdOccInfo bndr occ_info
--- | Decide whether some bindings should be made into join points or not.
+-- | Decide whether some bindings should be made into join points or not, based
+-- on its occurrences. This is
-- Returns `False` if they can't be join points. Note that it's an
-- all-or-nothing decision, as if multiple binders are given, they're
-- assumed to be mutually recursive.
--
--- It must, however, be a final decision. If we say "True" for 'f',
+-- It must, however, be a final decision. If we say `True` for 'f',
-- and then subsequently decide /not/ make 'f' into a join point, then
-- the decision about another binding 'g' might be invalidated if (say)
-- 'f' tail-calls 'g'.
=====================================
compiler/GHC/Data/Graph/Directed.hs
=====================================
@@ -4,6 +4,7 @@
{-# LANGUAGE GeneralizedNewtypeDeriving #-}
{-# LANGUAGE ScopedTypeVariables #-}
{-# LANGUAGE ViewPatterns #-}
+{-# LANGUAGE DeriveFunctor #-}
module GHC.Data.Graph.Directed (
Graph, graphFromEdgedVerticesOrd, graphFromEdgedVerticesUniq,
@@ -108,7 +109,7 @@ data Node key payload = DigraphNode {
node_payload :: payload, -- ^ User data
node_key :: key, -- ^ User defined node id
node_dependencies :: [key] -- ^ Dependencies/successors of the node
- }
+ } deriving Functor
instance (Outputable a, Outputable b) => Outputable (Node a b) where
=====================================
compiler/GHC/Utils/Misc.hs
=====================================
@@ -35,7 +35,7 @@ module GHC.Utils.Misc (
equalLength, compareLength, leLength, ltLength,
isSingleton, only, expectOnly, GHC.Utils.Misc.singleton,
- notNull, snocView,
+ notNull, expectNonEmpty, snocView,
chunkList,
@@ -481,7 +481,6 @@ expectOnly _ (a:_) = a
#endif
expectOnly msg _ = panic ("expectOnly: " ++ msg)
-
-- | Split a list into chunks of /n/ elements
chunkList :: Int -> [a] -> [[a]]
chunkList _ [] = []
@@ -500,6 +499,16 @@ changeLast [] _ = panic "changeLast"
changeLast [_] x = [x]
changeLast (x:xs) x' = x : changeLast xs x'
+-- | Like @expectJust msg . nonEmpty@; a better alternative to 'NE.fromList'.
+expectNonEmpty :: HasCallStack => String -> [a] -> NonEmpty a
+{-# INLINE expectNonEmpty #-}
+expectNonEmpty _ (x:xs) = x:|xs
+expectNonEmpty msg [] = expectNonEmptyPanic msg
+
+expectNonEmptyPanic :: String -> a
+expectNonEmptyPanic msg = panic ("expectNonEmpty: " ++ msg)
+{-# NOINLINE expectNonEmptyPanic #-}
+
-- | Apply an effectful function to the last list element.
mapLastM :: Functor f => (a -> f a) -> NonEmpty a -> f (NonEmpty a)
mapLastM f (x:|[]) = NE.singleton <$> f x
=====================================
testsuite/tests/simplCore/should_compile/T22428.hs
=====================================
@@ -0,0 +1,9 @@
+module T22428 where
+
+f :: Integer -> Integer -> Integer
+f x y = go y
+ where
+ go :: Integer -> Integer
+ go 0 = x
+ go n = go (n-1)
+ {-# INLINE go #-}
=====================================
testsuite/tests/simplCore/should_compile/T22428.stderr
=====================================
@@ -0,0 +1,45 @@
+
+==================== Tidy Core ====================
+Result size of Tidy Core
+ = {terms: 32, types: 14, coercions: 0, joins: 1/1}
+
+-- RHS size: {terms: 2, types: 0, coercions: 0, joins: 0/0}
+T22428.f1 :: Integer
+[GblId,
+ Unf=Unf{Src=<vanilla>, TopLvl=True, Value=True, ConLike=True,
+ WorkFree=True, Expandable=True, Guidance=IF_ARGS [] 10 10}]
+T22428.f1 = GHC.Num.Integer.IS 1#
+
+-- RHS size: {terms: 28, types: 10, coercions: 0, joins: 1/1}
+f :: Integer -> Integer -> Integer
+[GblId,
+ Arity=2,
+ Str=<SL><1L>,
+ Unf=Unf{Src=<vanilla>, TopLvl=True, Value=True, ConLike=True,
+ WorkFree=True, Expandable=True, Guidance=IF_ARGS [0 0] 156 0}]
+f = \ (x :: Integer) (y :: Integer) ->
+ joinrec {
+ go [InlPrag=INLINE (sat-args=1), Occ=LoopBreaker, Dmd=SC(S,L)]
+ :: Integer -> Integer
+ [LclId[JoinId(1)(Just [!])],
+ Arity=1,
+ Str=<1L>,
+ Unf=Unf{Src=StableUser, TopLvl=False, Value=True, ConLike=True,
+ WorkFree=True, Expandable=True,
+ Guidance=ALWAYS_IF(arity=1,unsat_ok=False,boring_ok=False)}]
+ go (ds :: Integer)
+ = case ds of wild {
+ GHC.Num.Integer.IS x1 ->
+ case x1 of {
+ __DEFAULT -> jump go (GHC.Num.Integer.integerSub wild T22428.f1);
+ 0# -> x
+ };
+ GHC.Num.Integer.IP x1 ->
+ jump go (GHC.Num.Integer.integerSub wild T22428.f1);
+ GHC.Num.Integer.IN x1 ->
+ jump go (GHC.Num.Integer.integerSub wild T22428.f1)
+ }; } in
+ jump go y
+
+
+
=====================================
testsuite/tests/simplCore/should_compile/all.T
=====================================
@@ -459,6 +459,10 @@ test('T22494', [grep_errmsg(r'case') ], compile, ['-O -ddump-simpl -dsuppress-un
test('T22491', normal, compile, ['-O2'])
test('T21476', normal, compile, [''])
test('T22272', normal, multimod_compile, ['T22272', '-O -fexpose-all-unfoldings -fno-omit-interface-pragmas -fno-ignore-interface-pragmas'])
+
+# go should become a join point
+test('T22428', [grep_errmsg(r'jump go') ], compile, ['-O -ddump-simpl -dsuppress-uniques -dno-typeable-binds -dsuppress-unfoldings'])
+
test('T22459', normal, compile, [''])
test('T22623', normal, multimod_compile, ['T22623', '-O -v0'])
test('T22662', normal, compile, [''])
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0f67ccdd1dc5aa9e04f9fd777705bc76ddd1ae05
--
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/0f67ccdd1dc5aa9e04f9fd777705bc76ddd1ae05
You're receiving this email because of your account on gitlab.haskell.org.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.haskell.org/pipermail/ghc-commits/attachments/20230112/65e21821/attachment-0001.html>
More information about the ghc-commits
mailing list