![]() Deeply normalize when computing implied outlives bounds
r? lcnr
Unfortunately resolving regions is still slightly scuffed (though in an unrelated way). Specifically, we should be normalizing our param-env outlives when constructing the `OutlivesEnv`; otherwise, these assumptions (
|
||
---|---|---|
.. | ||
error_reporting | ||
errors | ||
solve | ||
traits | ||
errors.rs | ||
infer.rs | ||
lib.rs | ||
regions.rs | ||
solve.rs |