SI-7226 Fix inference regression caused by TypeVar equality.

TypeVars, being mutable creatures, mustn't have structural
equality/hashing, otherwise TypeRefs that differ only by
having distinct TypeVars as components get wrongly uniqued
together.

The reported bug showed the disaterous consequences: constraints
from the `C?[Int]` in the return type applied to the `?C[?A]` in
the parameter list.

This commit overrides `equals` and `hashCode` in `TypeVar`
to use reference equality. An alternative fix would be to drop
the `case`-ness of the class, as was the case before 0cde930b
when this regressed.

1 files changed, 26 insertions, 0 deletions

diff --git a/test/files/pos/t7226.scala b/test/files/pos/t7226.scala
new file mode 100644
index 0000000000..06f0c95dc4
--- /dev/null
+++ b/test/files/pos/t7226.scala
@@ -0,0 +1,26 @@
+trait HK {
+  type Rep[X]
+
+  // okay
+  def unzip2[A, B](ps: Rep[List[(A, B)]])
+  unzip2(null.asInstanceOf[Rep[List[(Int, String)]]])
+
+  // okay
+  def unzipHK[A, B, C[_]](ps: Rep[C[(A, B)]])
+  unzipHK(null.asInstanceOf[Rep[List[(Int, String)]]])
+
+  def unzipHKRet0[A, C[_]](ps: C[A]): C[Int]
+  def ls: List[String]
+  unzipHKRet0(ls)
+
+  // fail
+  def unzipHKRet[A, C[_]](ps: Rep[C[A]]): Rep[C[Int]]
+  def rls: Rep[List[String]]
+  unzipHKRet(rls)
+}
+
+trait HK1 {
+  type Rep[A]
+  def unzip1[A, B, C[_]](ps: Rep[C[(A, B)]]): (Rep[C[A]], Rep[C[B]])
+  def doUnzip1[A, B](ps: Rep[List[(A, B)]]) = unzip1(ps)
+}