summaryrefslogtreecommitdiff
path: root/compiler/specialise/SpecConstr.lhs
diff options
context:
space:
mode:
Diffstat (limited to 'compiler/specialise/SpecConstr.lhs')
-rw-r--r--compiler/specialise/SpecConstr.lhs4
1 files changed, 2 insertions, 2 deletions
diff --git a/compiler/specialise/SpecConstr.lhs b/compiler/specialise/SpecConstr.lhs
index afc70f9b9e..ae578bddd4 100644
--- a/compiler/specialise/SpecConstr.lhs
+++ b/compiler/specialise/SpecConstr.lhs
@@ -587,7 +587,7 @@ We get two specialisations:
But perhaps the first one isn't good. After all, we know that tpl_B2 is
a T (I# x) really, because T is strict and Int has one constructor. (We can't
-unbox the strict fields, becuase T is polymorphic!)
+unbox the strict fields, because T is polymorphic!)
%************************************************************************
%* *
@@ -867,7 +867,7 @@ specialiations. Consider
If we specialise $j1 then in each specialisation (as well as the original)
we can specialise $j2, and similarly $j3. Even if we make just *one*
-specialisation of each, becuase we also have the original we'll get 2^n
+specialisation of each, because we also have the original we'll get 2^n
copies of $j3, which is not good.
So when recursively specialising we divide the sc_count by the number of