Add a forever unstable opt-out of const qualification checks #56123

oli-obk · 2018-11-21T09:57:15Z

basically a forever unstable way to screw with const things in horribly unsafe, unsound and incoherent ways.

Note that this does not affect miri except by maybe violating assumptions that miri makes. But there's no change in how miri evaluates things.

eddyb · 2018-11-21T10:36:57Z

cc @rust-lang/compiler

RalfJung · 2018-11-21T13:17:27Z

src/test/ui/consts/miri_unleashed/assoc_const_2.stderr

+  --> $DIR/assoc_const_2.rs:27:13
+   |
+LL |     let y = <String as Bar<String>>::F; //~ ERROR erroneous constant
+   |             ^^^^^^^^^^^^^^^^^^^^^^^^^^ referenced constant has errors


"Has errors" is not very precise, why does it not show an actual miri error?

because of allow(const_err) silencing the lint that reports the error on a constant

RalfJung · 2018-11-21T13:17:52Z

src/libsyntax/feature_gate.rs

@@ -230,6 +230,11 @@ declare_features! (
    // Allows panicking during const eval (produces compile-time errors)
    (active, const_panic, "1.30.0", Some(51999), None),

+    // Don't run any static analyses on constants and `const fn` and just "let miri handle it"
+    // NOT meant to ever be stabilized, this is solely for being able to demonstrate problems with
+    // features of alternative RFCs when writing RFCs that would allow more const features.


Also sounds useful for test cases.

varkor · 2018-11-21T13:50:01Z

Would it be reasonable to add this as a -Zunleash_the_miri_inside_of_you (or similar) instead of #[feature(unleash_the_miri_inside_of_you)]? The idea of people deciding to include the feature simply because it's convenient to do so isn't a nice one.

oli-obk · 2018-11-21T13:55:39Z

I'm not particularly bound to using features. Do we have any other -Z flags that change language rules?

RalfJung · 2018-11-21T14:09:43Z

I'd count these:

    -Z                  two-phase-borrows -- use two-phase reserved/active distinction for `&mut` borrows in MIR borrowck
    -Z           two-phase-beyond-autoref -- when using two-phase-borrows, allow two phases even for non-autoref `&mut` borrows
    -Z       disable-nll-user-type-assert -- disable user provided type assertion in NLL
    -Z       nll-dont-emit-read-for-match -- in match codegen, do not include FakeRead statements (used by mir-borrowck)

Centril · 2018-11-22T00:20:18Z

@varkor The main nice thing about features is that it becomes easy to test stuff out in the playground and such so if it's not too much of a problem using features would be nice.

src/librustc_mir/transform/qualify_consts.rs

src/libsyntax/feature_gate.rs

RalfJung · 2018-11-22T09:08:14Z

TBH I also feel like a -Z flag is more suited. Maybe we can get @shepmaster to support arbitrary additional flags on playground, similar to godbolt? :D

nagisa · 2018-11-22T14:47:31Z

NB: nll had both a flag and a "feature". Either one of them would enable the feature for testing.

oli-obk · 2018-11-22T15:44:10Z

Discussed in compiler meeting. A flag is definitely better because

making this feature a flag would make it impossible to misuse within libstd
or most of the code, really

pnkfelix · 2018-11-29T13:43:18Z

This was discussed a week ago in T-compiler. Reviewing the discussion now, it seems like there was consensus that the feature should be a -Z flag. I don't think there's anything more to discuss, and the nomination tag was just accidentally left on.)

Un-nominating.

bors · 2019-01-12T14:11:27Z

☔ The latest upstream changes (presumably #57542) made this pull request unmergeable. Please resolve the merge conflicts.

TimNN · 2019-01-29T18:14:48Z

Ping from triage! What is the status of this PR?

oli-obk · 2019-02-05T16:56:14Z

cc @therealprof @jamesmunns @japaric

eddyb · 2019-02-05T17:09:38Z

@bors r+

bors · 2019-02-05T17:09:43Z

📌 Commit d0129a6 has been approved by eddyb

oli-obk · 2019-02-06T08:25:15Z

@bors r=eddyb

the bors queue did not pick this up as "approved"

bors · 2019-02-06T08:25:15Z

📌 Commit d0129a6 has been approved by eddyb

bors · 2019-02-06T08:42:53Z

⌛ Testing commit d0129a6 with merge b139669...

@eddyb

Add a forever unstable opt-out of const qualification checks r? @eddyb cc @RalfJung @Centril basically a forever unstable way to screw with const things in horribly unsafe, unsound and incoherent ways. Note that this does *not* affect miri except by maybe violating assumptions that miri makes. But there's no change in how miri evaluates things.

bors · 2019-02-06T11:38:30Z

☀️ Test successful - checks-travis, status-appveyor
Approved by: eddyb
Pushing b139669 to master...

nagisa · 2019-02-06T13:42:26Z

rust-highfive assigned eddyb Nov 21, 2018

rust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Nov 21, 2018

This comment has been minimized.

Sign in to view

eddyb added I-nominated T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. labels Nov 21, 2018

RalfJung reviewed Nov 21, 2018

View reviewed changes

Centril reviewed Nov 22, 2018

View reviewed changes

src/librustc_mir/transform/qualify_consts.rs Outdated Show resolved Hide resolved

src/libsyntax/feature_gate.rs Outdated Show resolved Hide resolved

RalfJung mentioned this pull request Nov 24, 2018

Dynamic checks for all const soundness concerns? rust-lang/const-eval#17

Open

6 tasks

pnkfelix removed the I-nominated label Nov 29, 2018

oli-obk force-pushed the import_miri_from_future branch from 0377cc4 to 051fcd9 Compare November 29, 2018 16:25

This comment has been minimized.

Sign in to view

oli-obk force-pushed the import_miri_from_future branch from 051fcd9 to 2e102ca Compare December 6, 2018 09:02

This comment has been minimized.

Sign in to view

oli-obk force-pushed the import_miri_from_future branch from bfea651 to 2747960 Compare December 10, 2018 15:46

This comment has been minimized.

Sign in to view

oli-obk mentioned this pull request Dec 11, 2018

Add a way to pass custom rustc flags rust-lang/rust-playground#446

Open

oli-obk force-pushed the import_miri_from_future branch from 2747960 to 8a949a3 Compare December 11, 2018 10:10

Add a forever unstable opt-out of const qualification checks

d0129a6

oli-obk force-pushed the import_miri_from_future branch from 8a949a3 to d0129a6 Compare January 31, 2019 13:03

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Feb 5, 2019

bors merged commit d0129a6 into rust-lang:master Feb 6, 2019

oli-obk deleted the import_miri_from_future branch February 7, 2019 08:45

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add a forever unstable opt-out of const qualification checks #56123

Add a forever unstable opt-out of const qualification checks #56123

oli-obk commented Nov 21, 2018

This comment has been minimized.

eddyb commented Nov 21, 2018

RalfJung Nov 21, 2018

oli-obk Nov 21, 2018

RalfJung Nov 21, 2018

varkor commented Nov 21, 2018

oli-obk commented Nov 21, 2018

RalfJung commented Nov 21, 2018

Centril commented Nov 22, 2018

RalfJung commented Nov 22, 2018

nagisa commented Nov 22, 2018

oli-obk commented Nov 22, 2018

pnkfelix commented Nov 29, 2018

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

bors commented Jan 12, 2019

TimNN commented Jan 29, 2019

oli-obk commented Feb 5, 2019

eddyb commented Feb 5, 2019

bors commented Feb 5, 2019

oli-obk commented Feb 6, 2019

bors commented Feb 6, 2019

bors commented Feb 6, 2019

bors commented Feb 6, 2019

nagisa commented Feb 6, 2019

Add a forever unstable opt-out of const qualification checks #56123

Add a forever unstable opt-out of const qualification checks #56123

Conversation

oli-obk commented Nov 21, 2018

This comment has been minimized.

eddyb commented Nov 21, 2018

RalfJung Nov 21, 2018

Choose a reason for hiding this comment

oli-obk Nov 21, 2018

Choose a reason for hiding this comment

RalfJung Nov 21, 2018

Choose a reason for hiding this comment

varkor commented Nov 21, 2018

oli-obk commented Nov 21, 2018

RalfJung commented Nov 21, 2018

Centril commented Nov 22, 2018

RalfJung commented Nov 22, 2018

nagisa commented Nov 22, 2018

oli-obk commented Nov 22, 2018

pnkfelix commented Nov 29, 2018

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

bors commented Jan 12, 2019

TimNN commented Jan 29, 2019

oli-obk commented Feb 5, 2019

eddyb commented Feb 5, 2019

bors commented Feb 5, 2019

oli-obk commented Feb 6, 2019

bors commented Feb 6, 2019

bors commented Feb 6, 2019

bors commented Feb 6, 2019

nagisa commented Feb 6, 2019