diff options
author | Bartosz Nitka <niteria@gmail.com> | 2016-05-18 10:36:49 -0700 |
---|---|---|
committer | Bartosz Nitka <niteria@gmail.com> | 2016-05-18 16:43:49 -0700 |
commit | 6282bc31808e335cd8386dd20d469bc2457f84de (patch) | |
tree | e719c05fed4ea0aa9bab174c25571fe34687c597 /testsuite/tests/parser/should_fail | |
parent | f18e8d8db6246340c8840bd0ca2c43767009f14c (diff) | |
download | haskell-6282bc31808e335cd8386dd20d469bc2457f84de.tar.gz |
Kill varSetElems in tidyFreeTyCoVars
I haven't observed this to have an effect on nondeterminism,
but tidyOccName appears to modify the TidyOccEnv in a
way dependent on the order of inputs.
It's easy enough to change it to be deterministic to be on the
safe side.
Test Plan: ./validate
Reviewers: simonmar, austin, bgamari, simonpj
Reviewed By: simonpj
Subscribers: thomie
Differential Revision: https://phabricator.haskell.org/D2238
GHC Trac Issues: #4012
Diffstat (limited to 'testsuite/tests/parser/should_fail')
-rw-r--r-- | testsuite/tests/parser/should_fail/T7848.stderr | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/testsuite/tests/parser/should_fail/T7848.stderr b/testsuite/tests/parser/should_fail/T7848.stderr index f7617ee606..e1a64e4668 100644 --- a/testsuite/tests/parser/should_fail/T7848.stderr +++ b/testsuite/tests/parser/should_fail/T7848.stderr @@ -1,9 +1,9 @@ T7848.hs:6:1: error: • Occurs check: cannot construct the infinite type: - t ~ t0 -> t1 -> A -> A -> A -> A -> t2 -> t + t ~ t2 -> t1 -> A -> A -> A -> A -> t0 -> t • When checking that: - t0 -> t1 -> A -> A -> A -> A -> forall t2. t2 -> t + t2 -> t1 -> A -> A -> A -> A -> forall t4. t4 -> t is more polymorphic than: t • Relevant bindings include x :: t (bound at T7848.hs:6:1) |