Commit 9f71f697 authored by Simon Peyton Jones's avatar Simon Peyton Jones Committed by Marge Bot

Add two bangs to improve perf of flattening

This tiny patch improves the compile time of flatten-heavy
programs by 1-2%, by adding two bangs.

Addresses (somewhat) #18502

This reduces allocation by
   T9872b   -1.1%
   T9872d   -3.3%

   T5321Fun -0.2%
   T5631    -0.2%
   T5837    +0.1%
   T6048    +0.1%

Metric Decrease:
    T9872b
    T9872d
parent ebe2cf45
Pipeline #22815 failed with stages
in 400 minutes and 47 seconds
......@@ -1891,7 +1891,9 @@ substForAllCoBndrUsingLC sym sco (LC subst lc_env) tv co
--
-- For the inverse operation, see 'liftCoMatch'
ty_co_subst :: LiftingContext -> Role -> Type -> Coercion
ty_co_subst lc role ty
ty_co_subst !lc role ty
-- !lc: making this function strict in lc allows callers to
-- pass its two components separately, rather than boxing them
= go role ty
where
go :: Role -> Type -> Coercion
......@@ -2864,9 +2866,9 @@ simplifyArgsWorker orig_ki_binders orig_inner_ki orig_fvs
-- need a coercion (kind_co :: old_kind ~ new_kind).
--
-- The bangs here have been observed to improve performance
-- significantly in optimized builds.
let kind_co = mkSymCo $
liftCoSubst Nominal lc (tyCoBinderType binder)
-- significantly in optimized builds; see #18502
let !kind_co = mkSymCo $
liftCoSubst Nominal lc (tyCoBinderType binder)
!casted_xi = xi `mkCastTy` kind_co
casted_co = mkCoherenceLeftCo role xi kind_co co
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment