Make DataCon workers strict in strict fields (#20749)wip/T20749

This patch tweaks `exprIsConApp_maybe`, `exprIsHNF` and friends, and Demand
Analysis so that they exploit and maintain strictness of DataCon workers. See
`Note [Strict fields in Core]` for details.

Very little needed to change, and it puts field seq insertion done by Tag
Inference into a new perspective: That of *implementing* strict field semantics.
Before Tag Inference, DataCon workers are strict. Afterwards they are
effectively lazy and field seqs happen around use sites. History has shown
that there is no other way to guarantee taggedness and thus the STG Strict Field
Invariant.

Knock-on changes:

  * `exprIsHNF` previously used `exprOkForSpeculation` on unlifted arguments
    instead of recursing into `exprIsHNF`. That regressed the termination
    analysis in CPR analysis (which simply calls out to `exprIsHNF`), so I made
    it call `exprOkForSpeculation`, too.

  * There's a small regression in Demand Analysis, visible in the changed test
    output of T16859: Previously, a field seq on a variable would give that
    variable a "used exactly once" demand, now it's "used at least once",
    because `dmdTransformDataConSig` accounts for future uses of the field
    that actually all go through the case binder (and hence won't re-enter the
    potential thunk). The difference should hardly be observable.

  * The Simplifier's fast path for data constructors only applies to lazy
    data constructors now. I observed regressions involving Data.Binary.Put's
    `Pair` data type.

  * Unfortunately, T21392 does no longer reproduce after this patch, so I marked
    it as "not broken" in order to track whether we regress again in the future.

Fixes #20749, the satisfying conclusion of an annoying saga (cf. the ideas
in #21497 and #22475).

1 files changed, 3 insertions, 3 deletions

diff --git a/compiler/GHC/Core/Opt/ConstantFold.hs b/compiler/GHC/Core/Opt/ConstantFold.hs
index 86fdc5cdb5..3c2c1348d1 100644
--- a/compiler/GHC/Core/Opt/ConstantFold.hs
+++ b/compiler/GHC/Core/Opt/ConstantFold.hs
@@ -46,7 +46,7 @@ import GHC.Core
 import GHC.Core.Make
 import GHC.Core.SimpleOpt (  exprIsConApp_maybe, exprIsLiteral_maybe )
 import GHC.Core.DataCon ( DataCon,dataConTagZ, dataConTyCon, dataConWrapId, dataConWorkId )
-import GHC.Core.Utils  ( cheapEqExpr, exprIsHNF
+import GHC.Core.Utils  ( cheapEqExpr, exprOkForSpeculation
                        , stripTicksTop, stripTicksTopT, mkTicks )
 import GHC.Core.Multiplicity
 import GHC.Core.Rules.Config
@@ -1932,7 +1932,7 @@ Things to note
 
 Implementing seq#.  The compiler has magic for SeqOp in
 
-- GHC.Core.Opt.ConstantFold.seqRule: eliminate (seq# <whnf> s)
+- GHC.Core.Opt.ConstantFold.seqRule: eliminate (seq# <ok-for-spec> s)
 
 - GHC.StgToCmm.Expr.cgExpr, and cgCase: special case for seq#
 
@@ -1947,7 +1947,7 @@ Implementing seq#.  The compiler has magic for SeqOp in
 seqRule :: RuleM CoreExpr
 seqRule = do
   [Type _ty_a, Type _ty_s, a, s] <- getArgs
-  guard $ exprIsHNF a
+  guard $ exprOkForSpeculation a
   return $ mkCoreUnboxedTuple [s, a]
 
 -- spark# :: forall a s . a -> State# s -> (# State# s, a #)