Use non-intercrate SelectionContext for normalization in coherence::with_fresh_ty_vars

[b-naber]

Fixes #85898

Previously we used a SelectionContext with the intercrate flag set in with_fresh_ty_vars, which caused projections of foreign types to fail the orphan check. We now use a new SelectionContext for the normalization.

[rust-highfive]

r? @estebank

(rust-highfive has picked a reviewer for you, use r? to override)

[b-naber]

@estebank Can you review this, please?

[estebank]

@bors r+

[bors]

📌 Commit 8fd515c has been approved by estebank

[bors]

⌛ Testing commit 8fd515c with merge 5cdeccb1f7728b38161e581bcc6d55118e063be2...

[rust-log-analyzer]

The job x86_64-gnu-aux failed! Check out the build log: (web) (plain)

Click to see the possible cause of the failure (guessed by this bot)

   Compiling quickcheck v0.9.2
   Compiling ipnetwork v0.17.0
   Compiling diesel_derives v2.0.0 (/checkout/obj/build/ct/diesel/diesel_derives)
   Compiling diesel v2.0.0 (/checkout/obj/build/ct/diesel/diesel)
error[E0275]: overflow evaluating the requirement `&(_, _, _, _, _): insertable::Insertable<_>`
  |
  = help: consider adding a `#![recursion_limit="256"]` attribute to your crate (`diesel`)
  = note: required because of the requirements on the impl of `insertable::Insertable<_>` for `(&(_, _, _, _, _), &_, &_, &_, &_)`
  = note: 127 redundant requirements hidden
  = note: required because of the requirements on the impl of `insertable::Insertable<_>` for `&(((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((_, _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _)`
For more information about this error, try `rustc --explain E0275`.
For more information about this error, try `rustc --explain E0275`.
error: could not compile `diesel` due to previous error
error: build failed
thread 'main' panicked at 'tests failed for https://github.com/diesel-rs/diesel', src/tools/cargotest/main.rs:101:9



note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
command did not execute successfully: "/checkout/obj/build/x86_64-unknown-linux-gnu/stage0-tools-bin/cargotest" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2-tools-bin/cargo" "/checkout/obj/build/ct"
expected success, got: exit status: 101


Build completed unsuccessfully in 0:18:59
make: *** [check-aux] Error 1
Makefile:44: recipe for target 'check-aux' failed

[bors]

💔 Test failed - checks-actions

[b-naber]

@estebank This fails in some macro call, which I can't minimize. The only other possible solution I'd have for now would be to just change the flag on the existing SelectionContext during the normalization, but I'm unable to build the crate irrespective of the rustc version I use on my computer. Can we maybe try another bors run with the altered flag change? Otherwise I think we should close this.

[bors]

⌛ Testing commit 8fd515c with merge 4ef29ce025ae9896a9e8e80d60cfa4f78932627f...

[rust-log-analyzer]

The job x86_64-gnu-aux failed! Check out the build log: (web) (plain)

Click to see the possible cause of the failure (guessed by this bot)

   Compiling quickcheck v0.9.2
   Compiling ipnetwork v0.17.0
   Compiling diesel_derives v2.0.0 (/checkout/obj/build/ct/diesel/diesel_derives)
   Compiling diesel v2.0.0 (/checkout/obj/build/ct/diesel/diesel)
error[E0275]: overflow evaluating the requirement `&(_, _, _, _, _): insertable::Insertable<_>`
  |
  = help: consider adding a `#![recursion_limit="256"]` attribute to your crate (`diesel`)
  = note: required because of the requirements on the impl of `insertable::Insertable<_>` for `(&(_, _, _, _, _), &_, &_, &_, &_)`
  = note: 127 redundant requirements hidden
  = note: required because of the requirements on the impl of `insertable::Insertable<_>` for `&(((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((_, _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _), _, _, _, _)`
For more information about this error, try `rustc --explain E0275`.
error: could not compile `diesel` due to previous error
warning: build failed, waiting for other jobs to finish...
error: build failed
---
expected success, got: exit status: 101


Build completed unsuccessfully in 0:26:04
make: *** [check-aux] Error 1
Makefile:44: recipe for target 'check-aux' failed

[bors]

💔 Test failed - checks-actions

[JohnCSimon]

triage:
@b-naber moving this back to author, to address the build failures or close.

@rustbot label: +S-waiting-on-author -S-waiting-on-review

[estebank]

Can we maybe try another bors run with the altered flag change?

What change were you talking about?

Looking at that macro, it seems like maybe we might be doing something wrong now with the expansion of Value, but I might be wrong. I'm as mystified as you by the failures :-/

[camelid]

@bors r-

Tests failed, but bors still has it in the merge queue.

[JohnCSimon]

triage:
@b-naber can you please post the status on this PR?

[b-naber]

@JohnCSimon Sorry, I forgot about this PR. I will try to look into this again this week. Feel free to close this, if I won't update this PR by then.

[JohnCSimon]

@b-naber
Ping from triage: I'm closing this one due to inactivity, please feel to reopen when you are ready to continue with this.

@rustbot label: +S-inactive