-
Notifications
You must be signed in to change notification settings - Fork 12.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Allow printing TraitRef
and TraitPredicate
with Infcx
information
#125262
Conversation
@@ -16,7 +16,7 @@ LL | / fn foo() -> Bar | |||
LL | | where | |||
LL | | Bar: Send, | |||
| |______________^ | |||
= note: ...which requires revealing opaque types in `[Binder { value: TraitPredicate(<Bar as core::marker::Send>, polarity:Positive), bound_vars: [] }]`... | |||
= note: ...which requires revealing opaque types in `[Binder { value: (Alias(Opaque, AliasTy { args: [], def_id: DefId(0:7 ~ in_where_clause[cb1b]::Bar::{opaque#0}) }): core::marker::Send), bound_vars: [] }]`... |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is a bit unfortunate lol. we actually print the Ty
s in TraitPredicate
/TraitRef
using Debug
now and the debug printing for aliases is not super concise
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
.
☔ The latest upstream changes (presumably #125230) made this pull request unmergeable. Please resolve the merge conflicts. |
4cde784
to
876d181
Compare
☔ The latest upstream changes (presumably #125284) made this pull request unmergeable. Please resolve the merge conflicts. |
This PR is a bit of a mess apologies. I've tried to split up into two commits.
The first commit makes the
Infcx
inWithInfcx
be stored inside of anOption
and makesNoInfcx
into an uninhabited type so that we do not have a type that is not anything like anInfcx
implement theInferCtxtLike
trait and then get constructed and passed to places that expect an infcx. Generally the "leafs" of debug printing do not share codepaths for printing with or without an infcx and so need to be able to determine which to use.While updating the leaves to match on the
Option
I changed the debug printing of inference variables to be?0_..t
for when we cant probe the variable?0
instead of "silently" downgrading to normalDebug
printing. I'm not sure how valuable this is but it was easy to do and given how many places accidentally are not going through theDebugWithInfcx
trait it seems valuable to be able to tell in logs if an inference variable is being printed through the wrong trait or just doesnt have a universe available.As an example of not going through the right trait, it seems like the
Ty: DebugWithInfcx
impl just.... did not actually pass through toTyKind: DebugWithInfcx
and so would not ever print out the universes ^^' fixed that in this commit too.The second commit actually introduces the
DebugWithInfcx
implementations forTraitRef
andTraitPredicate
. PreviouslyTraitPredicate
was formatted as:TraitPredicate(<T as Trait<U>>, polarity::Positive)
and now it is(T: Trait<U>)
. This should hopefully make reading debug logs involving lots of trait bounds easier (and it'll get a lot better with customBinder
Debug
impls)TraitRef
's formatting has not really changed. Previously it was<T as Trait<U>>
and now it is(T as Trait<U>)
.r? @compiler-errors