broken MIR in Item (nightly) #99866

ArthurKValladares · 2022-07-28T18:08:41Z

Code

#[derive(Debug)]
pub struct HalSetLayouts {
    vertex_layout: <B as hal::Backend>::DescriptorSetLayout,
    fragment_layout: <B as hal::Backend>::DescriptorSetLayout,
    fragment_sampler_layout: <B as hal::Backend>::DescriptorSetLayout,
}

impl HalSetLayouts {
    fn iter<DSL>(&self) -> impl ExactSizeIterator<Item = &DSL> + '_
    where
        // rust 1.37 fails to typecheck `iter` without this. This is fixed in nightly.
        B: hal::Backend<DescriptorSetLayout = DSL>,
        DSL: std::fmt::Debug + Send + Sync + 'static,
    {
        std::iter::once(&self.vertex_layout)
            .chain(std::iter::once(&self.fragment_layout))
            .chain(std::iter::once(&self.fragment_sampler_layout))
            .exact(3)
    }
}

Error output

error: internal compiler error: no errors encountered even though `delay_span_bug` issued

error: internal compiler error: broken MIR in Item(WithOptConstParam { did: DefId(0:3021 ~ renderer[c01e]::shader::{impl#11}::iter), const_param_did: None }) (end of phase transition to Optimized) at bb0[0]:
                                Field projection `(*_1).field[0]` specified type `DSL`, but actual type is <gfx_backend_metal::Backend as gfx_hal::Backend>::DescriptorSetLayout
   --> brainstorm/src/renderer/src/shader.rs:642:25
    |
642 |         std::iter::once(&self.vertex_layout)
    |                         ^^^^^^^^^^^^^^^^^^^
    |
    = note: delayed at compiler/rustc_const_eval/src/transform/validate.rs:129:36

error: internal compiler error: broken MIR in Item(WithOptConstParam { did: DefId(0:3021 ~ renderer[c01e]::shader::{impl#11}::iter), const_param_did: None }) (end of phase transition to Optimized) at bb1[0]:
                                Field projection `(*_1).field[1]` specified type `DSL`, but actual type is <gfx_backend_metal::Backend as gfx_hal::Backend>::DescriptorSetLayout
   --> brainstorm/src/renderer/src/shader.rs:643:36
    |
643 |             .chain(std::iter::once(&self.fragment_layout))
    |                                    ^^^^^^^^^^^^^^^^^^^^^
    |
    = note: delayed at compiler/rustc_const_eval/src/transform/validate.rs:129:36

error: internal compiler error: broken MIR in Item(WithOptConstParam { did: DefId(0:3021 ~ renderer[c01e]::shader::{impl#11}::iter), const_param_did: None }) (end of phase transition to Optimized) at bb3[0]:
                                Field projection `(*_1).field[2]` specified type `DSL`, but actual type is <gfx_backend_metal::Backend as gfx_hal::Backend>::DescriptorSetLayout
   --> brainstorm/src/renderer/src/shader.rs:644:36
    |
644 |             .chain(std::iter::once(&self.fragment_sampler_layout))
    |                                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    |
    = note: delayed at compiler/rustc_const_eval/src/transform/validate.rs:129:36

Backtrace

thread 'rustc' panicked at 'Box<dyn Any>', compiler/rustc_errors/src/lib.rs:1426:13
stack backtrace:
   0: std::panicking::begin_panic::<rustc_errors::ExplicitBug>
   1: std::panic::panic_any::<rustc_errors::ExplicitBug>
   2: <rustc_errors::HandlerInner as core::ops::drop::Drop>::drop
   3: core::ptr::drop_in_place::<rustc_session::parse::ParseSess>
   4: <alloc::rc::Rc<rustc_session::session::Session> as core::ops::drop::Drop>::drop
   5: core::ptr::drop_in_place::<rustc_interface::interface::Compiler>
   6: rustc_span::with_source_map::<core::result::Result<(), rustc_errors::ErrorGuaranteed>, rustc_interface::interface::create_compiler_and_run<core::result::Result<(), rustc_errors::ErrorGuaranteed>, rustc_driver::run_compiler::{closure#1}>::{closure#1}>
   7: rustc_interface::interface::create_compiler_and_run::<core::result::Result<(), rustc_errors::ErrorGuaranteed>, rustc_driver::run_compiler::{closure#1}>
   8: <scoped_tls::ScopedKey<rustc_span::SessionGlobals>>::set::<rustc_interface::interface::run_compiler<core::result::Result<(), rustc_errors::ErrorGuaranteed>, rustc_driver::run_compiler::{closure#1}>::{closure#0}, core::result::Result<(), rustc_errors::ErrorGuaranteed>>
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.

This came from a workaround for what seems to be an old compiler bug. removing the workaround gets ride of the ICE:

#[derive(Debug)]
pub struct HalSetLayouts {
    vertex_layout: <B as hal::Backend>::DescriptorSetLayout,
    fragment_layout: <B as hal::Backend>::DescriptorSetLayout,
    fragment_sampler_layout: <B as hal::Backend>::DescriptorSetLayout,
}

impl HalSetLayouts {
    fn iter(
        &self,
    ) -> impl ExactSizeIterator<Item = &<B as hal::Backend>::DescriptorSetLayout> + '_ {
        std::iter::once(&self.vertex_layout)
            .chain(std::iter::once(&self.fragment_layout))
            .chain(std::iter::once(&self.fragment_sampler_layout))
            .exact(3)
    }
}

The text was updated successfully, but these errors were encountered:

5225225 · 2022-07-28T19:15:20Z

Can you post what your dependencies are? Or include the use statements? Ideally with the versions too.

Or something that can just be stuck in a main.rs that compiles? I'm not sure where the B is coming from in the code above (in the struct layout)

ArthurKValladares · 2022-07-28T19:48:49Z

Of course, my bad for not providing this upfront. I've been able to repo it with this setup:

deps/features:

[features]
default = ["metal"]
dx12 = ["gfx-backend-dx12"]
empty = ["gfx-backend-empty"]
gl = ["gfx-backend-gl"]
metal = ["gfx-backend-metal"]
vulkan = ["gfx-backend-vulkan"]

[dependencies]
gfx-backend-dx12 = { version = "0.6.3", optional = true }
gfx-backend-empty = { version = "0.6.0", optional = true }
gfx-backend-gl = { version = "0.5.1", optional = true }
gfx-backend-metal = { version = "0.6.2", optional = true }
gfx-backend-vulkan = { version = "0.6.1", optional = true }
gfx-hal = "0.6.0"

lib.rs

#[cfg(feature = "metal")]
use gfx_backend_metal as back;
pub use gfx_hal as hal;

#[cfg(feature = "vulkan")]
use gfx_backend_vulkan as back;

#[cfg(not(any(feature = "metal", feature = "vulkan")))]
use gfx_backend_empty as back;

pub type B = back::Backend;

pub struct Exacterator<I: Iterator<Item = T>, T> {
    iter: I,
    size: usize,
}

impl<I: Iterator<Item = T>, T> Exacterator<I, T> {
    pub fn new(iter: I, size: usize) -> Self {
        Self { iter, size }
    }
}

impl<I: Iterator<Item = T>, T> Iterator for Exacterator<I, T> {
    type Item = I::Item;

    fn next(&mut self) -> Option<Self::Item> {
        self.iter.next()
    }

    fn size_hint(&self) -> (usize, Option<usize>) {
        (self.size, Some(self.size))
    }
}

impl<I: Iterator<Item = T>, T> ExactSizeIterator for Exacterator<I, T> {}

pub trait Exact<T>: Iterator<Item = T> + Sized {
    fn exact(self, size: usize) -> Exacterator<Self, T>;
}

impl<I: Iterator<Item = T>, T> Exact<T> for I {
    fn exact(self, size: usize) -> Exacterator<Self, T> {
        Exacterator::new(self, size)
    }
}

#[derive(Debug)]
pub struct HalSetLayouts {
    vertex_layout: <B as hal::Backend>::DescriptorSetLayout,
    fragment_layout: <B as hal::Backend>::DescriptorSetLayout,
    fragment_sampler_layout: <B as hal::Backend>::DescriptorSetLayout,
}

impl HalSetLayouts {
    pub fn iter<DSL>(&self) -> impl ExactSizeIterator<Item = &DSL> + '_
    where
        // rust 1.37 fails to typecheck `iter` without this. This is fixed in nightly.
        B: hal::Backend<DescriptorSetLayout = DSL>,
        DSL: std::fmt::Debug + Send + Sync + 'static,
    {
        std::iter::once(&self.vertex_layout)
            .chain(std::iter::once(&self.fragment_layout))
            .chain(std::iter::once(&self.fragment_sampler_layout))
            .exact(3)
    }
}

5225225 · 2022-07-28T19:53:40Z

Ah yes, thank you very much. I can reproduce it (with the empty feature, so all the other crates can presumably be removed). I'll look into this.

5225225 · 2022-07-28T20:07:56Z

For reference, minimised it down to this:

pub trait Backend {
    type DescriptorSetLayout;
}

pub struct Back;

impl Backend for Back {
    type DescriptorSetLayout = u32;
}

pub struct HalSetLayouts {
    vertex_layout: <Back as Backend>::DescriptorSetLayout,
}

impl HalSetLayouts {
    pub fn iter<DSL>(self) -> DSL
    where
        Back: Backend<DescriptorSetLayout = DSL>,
    {
        self.vertex_layout
    }
}

I also bisected this, and it turns out to be the same rollup as in #99852... So I suspect both these errors were introduced in the same commit. Interesting!

Edit: Adjusted the code (edit2: twice!), check comment history for the original.

Noratrieb · 2022-07-28T20:34:53Z

Here is the MIR for the iter function when I manually disabled validation (I just noticed, validation doesn't have to be disabled, -Zunpretty=mir works anyways)

fn <impl at lol.rs:17:1: 17:19>::iter(_1: HalSetLayouts) -> DSL {
    debug self => _1;                    // in scope 0 at lol.rs:18:22: 18:26
    let mut _0: DSL;                     // return place in scope 0 at lol.rs:18:31: 18:34

    bb0: {
        _0 = move (_1.0: DSL);           // scope 0 at lol.rs:22:9: 22:27
        return;                          // scope 0 at lol.rs:23:6: 23:6
    }
}

Noratrieb · 2022-07-28T20:44:50Z

After looking into this a bit, I got the following:

The body is treating the generic type DST and the concrete associated type <Back as Backend>::DescriptorSetLayout to be of the same type. The MIR validator probably thinks that this is very sketchy and bails out (I haven't looked at it closely).

This is actually ok though, since in the where bounds DST is constrained to be equal to <Back as Backend>::DescriptorSetLayout.

To me, this looks like a bug in the MIR validator.

Noratrieb · 2022-07-28T20:50:52Z

This was probably regressed by #96856. cc @drmeepster

Noratrieb · 2022-07-28T20:54:39Z

Yes, manually reverting the PR fixed the issue for me locally.

Noratrieb · 2022-07-28T21:49:13Z

While it regressed in that PR, the PR only called validation that wasn't called before. The validation has been broken for longer.

Noratrieb · 2022-07-29T21:00:46Z

@rustbot claim

rust-lang/rust#99866 rust-lang/rust#99914 rust-lang/rust#99945

compiler-errors · 2022-08-03T18:31:24Z

Probably just needs a normalize call somewhere 🤔

c410-f3r · 2022-08-25T12:40:16Z

Does not ICE any more after #100571. Test added in #100998.

c410-f3r · 2022-08-26T15:59:32Z

Oooppsss... Sorry, this test still ICE.

On the bright side, I was able to reduce the compiler arguments to --crate-type lib --emit=link.

… r=oli-obk Try normalizing types without RevealAll in ParamEnv in MIR validation Before, the MIR validator used RevealAll in its ParamEnv for type checking. This could cause false negatives in some cases due to RevealAll ParamEnvs not always use all predicates as expected here. Since some MIR passes like inlining use RevealAll as well, keep using it in the MIR validator too, but when it fails usign RevealAll, also try the check without it, to stop false negatives. Fixes rust-lang#99866 cc `@compiler-errors` who nicely helped me on zulip

… r=oli-obk Try normalizing types without RevealAll in ParamEnv in MIR validation Before, the MIR validator used RevealAll in its ParamEnv for type checking. This could cause false negatives in some cases due to RevealAll ParamEnvs not always use all predicates as expected here. Since some MIR passes like inlining use RevealAll as well, keep using it in the MIR validator too, but when it fails usign RevealAll, also try the check without it, to stop false negatives. Fixes rust-lang#99866 cc ``@compiler-errors`` who nicely helped me on zulip

… r=oli-obk Try normalizing types without RevealAll in ParamEnv in MIR validation Before, the MIR validator used RevealAll in its ParamEnv for type checking. This could cause false negatives in some cases due to RevealAll ParamEnvs not always use all predicates as expected here. Since some MIR passes like inlining use RevealAll as well, keep using it in the MIR validator too, but when it fails usign RevealAll, also try the check without it, to stop false negatives. Fixes rust-lang#99866 cc ```@compiler-errors``` who nicely helped me on zulip

… r=oli-obk Try normalizing types without RevealAll in ParamEnv in MIR validation Before, the MIR validator used RevealAll in its ParamEnv for type checking. This could cause false negatives in some cases due to RevealAll ParamEnvs not always use all predicates as expected here. Since some MIR passes like inlining use RevealAll as well, keep using it in the MIR validator too, but when it fails usign RevealAll, also try the check without it, to stop false negatives. Fixes rust-lang#99866 cc ````@compiler-errors```` who nicely helped me on zulip

… r=oli-obk Try normalizing types without RevealAll in ParamEnv in MIR validation Before, the MIR validator used RevealAll in its ParamEnv for type checking. This could cause false negatives in some cases due to RevealAll ParamEnvs not always use all predicates as expected here. Since some MIR passes like inlining use RevealAll as well, keep using it in the MIR validator too, but when it fails usign RevealAll, also try the check without it, to stop false negatives. Fixes rust-lang#99866 cc `````@compiler-errors````` who nicely helped me on zulip

… r=oli-obk Try normalizing types without RevealAll in ParamEnv in MIR validation Before, the MIR validator used RevealAll in its ParamEnv for type checking. This could cause false negatives in some cases due to RevealAll ParamEnvs not always use all predicates as expected here. Since some MIR passes like inlining use RevealAll as well, keep using it in the MIR validator too, but when it fails usign RevealAll, also try the check without it, to stop false negatives. Fixes rust-lang#99866 cc `@compiler-errors` who nicely helped me on zulip

… r=oli-obk Try normalizing types without RevealAll in ParamEnv in MIR validation Before, the MIR validator used RevealAll in its ParamEnv for type checking. This could cause false negatives in some cases due to RevealAll ParamEnvs not always use all predicates as expected here. Since some MIR passes like inlining use RevealAll as well, keep using it in the MIR validator too, but when it fails usign RevealAll, also try the check without it, to stop false negatives. Fixes rust-lang#99866 cc ``@compiler-errors`` who nicely helped me on zulip

… r=oli-obk Try normalizing types without RevealAll in ParamEnv in MIR validation Before, the MIR validator used RevealAll in its ParamEnv for type checking. This could cause false negatives in some cases due to RevealAll ParamEnvs not always use all predicates as expected here. Since some MIR passes like inlining use RevealAll as well, keep using it in the MIR validator too, but when it fails usign RevealAll, also try the check without it, to stop false negatives. Fixes rust-lang#99866 cc ```@compiler-errors``` who nicely helped me on zulip

… r=oli-obk Try normalizing types without RevealAll in ParamEnv in MIR validation Before, the MIR validator used RevealAll in its ParamEnv for type checking. This could cause false negatives in some cases due to RevealAll ParamEnvs not always use all predicates as expected here. Since some MIR passes like inlining use RevealAll as well, keep using it in the MIR validator too, but when it fails usign RevealAll, also try the check without it, to stop false negatives. Fixes rust-lang#99866 cc ````@compiler-errors```` who nicely helped me on zulip

… r=oli-obk Try normalizing types without RevealAll in ParamEnv in MIR validation Before, the MIR validator used RevealAll in its ParamEnv for type checking. This could cause false negatives in some cases due to RevealAll ParamEnvs not always use all predicates as expected here. Since some MIR passes like inlining use RevealAll as well, keep using it in the MIR validator too, but when it fails usign RevealAll, also try the check without it, to stop false negatives. Fixes rust-lang#99866 cc `````@compiler-errors````` who nicely helped me on zulip

… r=oli-obk Try normalizing types without RevealAll in ParamEnv in MIR validation Before, the MIR validator used RevealAll in its ParamEnv for type checking. This could cause false negatives in some cases due to RevealAll ParamEnvs not always use all predicates as expected here. Since some MIR passes like inlining use RevealAll as well, keep using it in the MIR validator too, but when it fails usign RevealAll, also try the check without it, to stop false negatives. Fixes rust-lang#99866 cc ``````@compiler-errors`````` who nicely helped me on zulip

… r=oli-obk Try normalizing types without RevealAll in ParamEnv in MIR validation Before, the MIR validator used RevealAll in its ParamEnv for type checking. This could cause false negatives in some cases due to RevealAll ParamEnvs not always use all predicates as expected here. Since some MIR passes like inlining use RevealAll as well, keep using it in the MIR validator too, but when it fails usign RevealAll, also try the check without it, to stop false negatives. Fixes rust-lang#99866 cc ```````@compiler-errors``````` who nicely helped me on zulip

ArthurKValladares added C-bug Category: This is a bug. I-ICE Issue: The compiler panicked, giving an Internal Compilation Error (ICE) ❄️ T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. labels Jul 28, 2022

5225225 mentioned this issue Jul 28, 2022

ICE when returning value bound by if let ... && let ... #99852

Closed

rustbot assigned Noratrieb Jul 29, 2022

matthiaskrgr added a commit to matthiaskrgr/glacier that referenced this issue Jul 31, 2022

add 3 ices

55cef5d

rust-lang/rust#99866 rust-lang/rust#99914 rust-lang/rust#99945

matthiaskrgr mentioned this issue Jul 31, 2022

add 3 ices rust-lang/glacier#1368

Merged

rust-lang-glacier-bot added the glacier ICE tracked in rust-lang/glacier. label Jul 31, 2022

Noratrieb mentioned this issue Aug 3, 2022

Try normalizing types without RevealAll in ParamEnv in MIR validation #100121

Merged

ehuss mentioned this issue Aug 14, 2022

ICE when compiling nalgebra 0.25.4 in Release mode #100550

Closed

Jarcho mentioned this issue Aug 20, 2022

Replace expr_visitor with for_each_expr rust-lang/rust-clippy#8762

Merged

c410-f3r mentioned this issue Aug 25, 2022

Add tests for 99866 and 99852 #100998

Closed

bors closed this as completed in 938897e Sep 2, 2022

github-actions bot mentioned this issue Sep 3, 2022

ices/99866.rs: fixed with no errors rust-lang/glacier#1410

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

broken MIR in Item (nightly) #99866

broken MIR in Item (nightly) #99866

ArthurKValladares commented Jul 28, 2022 •

edited by rustbot

Loading

5225225 commented Jul 28, 2022 •

edited

Loading

ArthurKValladares commented Jul 28, 2022

5225225 commented Jul 28, 2022

5225225 commented Jul 28, 2022 •

edited

Loading

Noratrieb commented Jul 28, 2022 •

edited

Loading

Noratrieb commented Jul 28, 2022

Noratrieb commented Jul 28, 2022

Noratrieb commented Jul 28, 2022

Noratrieb commented Jul 28, 2022

Noratrieb commented Jul 29, 2022

compiler-errors commented Aug 3, 2022

c410-f3r commented Aug 25, 2022

c410-f3r commented Aug 26, 2022

broken MIR in Item (nightly) #99866

broken MIR in Item (nightly) #99866

Comments

ArthurKValladares commented Jul 28, 2022 • edited by rustbot Loading

Code

Meta

Error output

5225225 commented Jul 28, 2022 • edited Loading

ArthurKValladares commented Jul 28, 2022

5225225 commented Jul 28, 2022

5225225 commented Jul 28, 2022 • edited Loading

Noratrieb commented Jul 28, 2022 • edited Loading

Noratrieb commented Jul 28, 2022

Noratrieb commented Jul 28, 2022

Noratrieb commented Jul 28, 2022

Noratrieb commented Jul 28, 2022

Noratrieb commented Jul 29, 2022

compiler-errors commented Aug 3, 2022

c410-f3r commented Aug 25, 2022

c410-f3r commented Aug 26, 2022

ArthurKValladares commented Jul 28, 2022 •

edited by rustbot

Loading

5225225 commented Jul 28, 2022 •

edited

Loading

5225225 commented Jul 28, 2022 •

edited

Loading

Noratrieb commented Jul 28, 2022 •

edited

Loading