Remove const_in_array_repeat

[JulianKnodt]

Fixes #80371. Fixes #81315. Fixes #80767. Fixes #75682.

I thought there might be some issue with Repeats(_, 0), but if you increase the items in the array it still ICEs. I'm not sure if this is the best fix but it does fix the given issue.

[rust-highfive]

r? @lcnr

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

[lcnr]

this looks fine to me though I am kind of surprised that this error wasn't hit more frequently as borrowck uses mir_promoted. There seems to be another way to prevent borrowck from accessing promoteds or something I don't know about.

As I don't fully understand that I will leave the final review to someone else.

maybe r? @ecstatic-morse

[camelid]

r? @oli-obk (ecstatic-morse is no longer doing reviews)

[camelid]

I marked this PR as also fixing #81315, #80767, and #75682.

[tmiasko]

I also looked at this issue before noticing that it has PR opened already.

As far as I understand, when originally written this code was working only with MIR before promotion so assertion was always true. Subsequently the code was reused to check whether const_in_array_repeat_expressions feature would successfully promote the operand used in repeat expression and if so to provide that as a suggestion to fix a compilation error. In the latter case the code is running on a MIR after promotion and the assertion is no longer true.

My first impression regarding fixing it was to simply return false for promoteds, since they haven't been disqualified earlier.

[oli-obk]

cc @RalfJung we could also just delete the const_in_array_repeat_expressions feature to fix this issue 😆

[RalfJung]

@oli-obk you mean together with this problem? ;)

Fine for me. The assertion here could still be helpful to catch accidentally created nested promoteds.

[oli-obk]

The assertion here could still be helpful to catch accidentally created nested promoteds.

yea, I don't want to remove the assertion, so let's go with removing const_in_array_repeat_expressions.

[JulianKnodt]

Should I change this PR or should I close this for someone else's implementation?

[oli-obk]

Should I change this PR or should I close this for someone else's implementation?

If you have the time and motivation to do it, the job's all yours. Feel free to do it in this PR.

[rust-log-analyzer]

The job x86_64-gnu-llvm-9 failed! Check out the build log: (web) (plain)

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

[rust-log-analyzer]

The job x86_64-gnu-llvm-9 failed! Check out the build log: (web) (plain)

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

[rust-log-analyzer]

The job mingw-check failed! Check out the build log: (web) (plain)

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

configure: rust.channel         := nightly
configure: rust.debug-assertions := True
configure: llvm.assertions      := True
configure: dist.missing-tools   := True
configure: build.configure-args := ['--enable-sccache', '--disable-manage-submodu ...
configure: writing `config.toml` in current directory
configure: 
configure: run `python /checkout/x.py --help`
configure: 
---
   Compiling regex v1.4.3
   Compiling tidy v0.1.0 (/checkout/src/tools/tidy)
    Finished release [optimized] target(s) in 9.98s
tidy check
tidy error: following path contains more than 2669 entries, you should move the test to some relevant subdirectory (current: 2671): /checkout/src/test/ui/issues
Found 435 error codes
Found 0 error codes with no tests
Done!
Done!
tidy error: The Unstable Book has a 'language feature' section 'const-in-array-repeat-expressions' which doesn't correspond to an unstable language feature
some tidy checks failed

command did not execute successfully: "/checkout/obj/build/x86_64-unknown-linux-gnu/stage0-tools-bin/tidy" "/checkout" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage0/bin/cargo" "/checkout/obj/build"
expected success, got: exit code: 1

[rust-log-analyzer]

The job mingw-check failed! Check out the build log: (web) (plain)

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

configure: rust.channel         := nightly
configure: rust.debug-assertions := True
configure: llvm.assertions      := True
configure: dist.missing-tools   := True
configure: build.configure-args := ['--enable-sccache', '--disable-manage-submodu ...
configure: writing `config.toml` in current directory
configure: 
configure: run `python /checkout/x.py --help`
configure: 
---
   Compiling regex v1.4.3
   Compiling tidy v0.1.0 (/checkout/src/tools/tidy)
    Finished release [optimized] target(s) in 9.28s
tidy check
tidy error: following path contains more than 2669 entries, you should move the test to some relevant subdirectory (current: 2671): /checkout/src/test/ui/issues
Found 435 error codes
Found 0 error codes with no tests
Done!
some tidy checks failed

[oli-obk]

you need to move your new test into a test subfolder.

it's also very likey that you need to bless some tests and change or remove some tests that use the const_in_array_repeat_expressions feature gate

[bors]

📌 Commit bc653d60f640bb3ddfd905dba038ec8884026d52 has been approved by oli-obk

[RalfJung]

@bors r-
I left a comment

[RalfJung]

Thanks. :)
@bors r=oli-obk

[bors]

📌 Commit 6946534 has been approved by oli-obk