[Git][ghc/ghc][wip/T23923-mikolaj-take-2] Touch up the new Note

[Mikolaj Konarski (@Mikolaj)]

Mikolaj Konarski pushed to branch wip/T23923-mikolaj-take-2 at Glasgow Haskell Compiler / GHC


Commits:
6eedc476 by Mikolaj Konarski at 2024-04-05T16:01:14+02:00
Touch up the new Note

- - - - -


1 changed file:

- compiler/GHC/Core/TyCo/Rep.hs


Changes:

=====================================
compiler/GHC/Core/TyCo/Rep.hs
=====================================
@@ -1796,7 +1796,6 @@ In particular, given
                        | tv `elemVarSet` acc = acc
                        | otherwise = acc `extendVarSet` tv
 
-
 we want to end up with
    fvs ty = go emptyVarSet ty emptyVarSet
      where
@@ -1832,9 +1831,9 @@ Note [Use explicit recursion in foldTyCo]
 In foldTyCo you'll see things like:
     go_tys _   []     = mempty
     go_tys env (t:ts) = go_ty env t `mappend` go_tys env ts
-where we use /explicit recursion/.  You might wonder about using foldr instead:
+where we use /explicit recursion/.  You might wonder about using foldl instead:
     go_tys env = foldl (\t acc -> go_ty env t `mappend` acc) mempty
-Or maybe or foldl', or foldr.
+Or maybe foldl', or foldr.
 
 But don't do that for two reasons (see #24591)
 
@@ -1848,7 +1847,7 @@ But don't do that for two reasons (see #24591)
 
   But in the foldl form that is /much/ less obvious, and the strictness
   analyser fails utterly.  Result: lots and lots of thunks get built.  In
-  !12037, Mikolaj found that GHC allocated allocated /six times/ as much heap
+  !12037, Mikolaj found that GHC allocated /six times/ as much heap
   on test perf/compiler/T9198 as a result of this single problem!
 
 * Second, while I think that using `foldr` would be fine (simple experiments in



View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6eedc47691c7078c731a3e51db92b9c68bf35b3a

-- 
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/6eedc47691c7078c731a3e51db92b9c68bf35b3a
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/20240405/ec4affa4/attachment-0001.html>