Panic due to assertion `chunk_id_removed.is_some()` #7992

grtlr · 2024-11-04T15:15:39Z

Describe the bug

I'm currently working on the graph primitives and I'm encountering a panic when using log_static multiple times. Although, of course, there should only ever be a single call to log_static for a given entity, having multiple calls should not cause Rerun to crash.

To Reproduce
Steps to reproduce the behavior:

I'm on branch #7500, and I'm using the following code:

use napkin::force::SimulationBuilder;
use rerun::GraphNodes;

const NUM_NODES: usize = 10;

fn main() -> anyhow::Result<()> {
    let rec = rerun::RecordingStreamBuilder::new("napkin_disjoint").spawn()?;

    let nodes = (0..NUM_NODES)
        .map(|i| (format!("node{}", i), [0.0, 0.0]))
        .collect::<Vec<_>>();

    let builder = SimulationBuilder::default();
    let mut simulation = builder
        .build(nodes.clone())
        .add_force_collide("collide".into(), Default::default());

    simulation.step_with(
        10.into(),
        || {
            let nodes = nodes.iter().map(|(key, _)| key.clone()).collect::<Vec<_>>();
            println!("Step!");
            rec.log_static("/nodes", &GraphNodes::new(nodes)).unwrap();
        },
        || {
            let nodes = nodes.iter().map(|(key, _)| key.clone()).collect::<Vec<_>>();
            println!("Finish!");
            rec.log_static("/nodes", &GraphNodes::new(nodes)).unwrap();
        },
    );

    Ok(())
}

If it helps, I can also share the complete example, including the unreleased napkin crate.

Removing the _static makes the panic go away. Unortunately, I was not able to create a self-contained example yet, but I'm happy to try some more if it helps.

Expected behavior

Rerun should not crash but rather emit an error / warning.

Backtrace

thread 'main' panicked at 'assertion failed: chunk_id_removed.is_some()'
re_chunk_store/src/writes.rs:145
stack backtrace:
   7: core::panicking::panic_fmt
             at core/src/panicking.rs:72:14
   8: core::panicking::panic
             at core/src/panicking.rs:146:5
   9: re_chunk_store::writes::<impl re_chunk_store::store::ChunkStore>::insert_chunk
             at re_chunk_store/src/writes.rs:145:25
  10: re_entity_db::entity_db::EntityDb::add_chunk
             at re_entity_db/src/entity_db.rs:367:28
  11: re_entity_db::entity_db::EntityDb::add
             at re_entity_db/src/entity_db.rs:353:17
  12: re_viewer::app::App::receive_messages
             at re_viewer/src/app.rs:1160:19
  13: <re_viewer::app::App as eframe::epi::App>::update
             at re_viewer/src/app.rs:1718:9
  14: eframe::native::epi_integration::EpiIntegration::update::{{closure}}
             at eframe-0.29.1/src/native/epi_integration.rs:283:17
      egui::context::Context::run
             at egui-0.29.1/src/context.rs:815:13
  15: eframe::native::epi_integration::EpiIntegration::update
             at eframe-0.29.1/src/native/epi_integration.rs:276:27
  16: eframe::native::wgpu_integration::WgpuWinitRunning::run_ui_and_paint
             at eframe-0.29.1/src/native/wgpu_integration.rs:603:27
      <eframe::native::wgpu_integration::WgpuWinitApp as eframe::native::winit_integration::WinitApp>::run_ui_and_paint
             at eframe-0.29.1/src/native/wgpu_integration.rs:372:13
  17: <eframe::native::run::WinitAppWrapper<T> as winit::application::ApplicationHandler<eframe::native::winit_integration::UserEvent>>::window_event::{{closure}}
             at eframe-0.29.1/src/native/run.rs:283:21
      eframe::native::event_loop_context::with_event_loop_context
             at eframe-0.29.1/src/native/event_loop_context.rs:53:5
      <eframe::native::run::WinitAppWrapper<T> as winit::application::ApplicationHandler<eframe::native::winit_integration::UserEvent>>::window_event
             at eframe-0.29.1/src/native/run.rs:280:9
  18: winit::event_loop::dispatch_event_for_app
             at winit-0.30.5/src/event_loop.rs:642:52
      winit::event_loop::EventLoop<T>::run_app::{{closure}}
             at winit-0.30.5/src/event_loop.rs:265:49
  19: winit::platform_impl::macos::event_loop::map_user_event::{{closure}}
             at winit-0.30.5/src/platform_impl/macos/event_loop.rs:174:22
  20: <alloc::boxed::Box<F,A> as core::ops::function::FnMut<Args>>::call_mut
             at alloc/src/boxed.rs:2029:9
      winit::platform_impl::macos::event_handler::EventHandler::handle_event
             at winit-0.30.5/src/platform_impl/macos/event_handler.rs:125:17
  21: winit::platform_impl::macos::app_state::ApplicationDelegate::handle_event
             at winit-0.30.5/src/platform_impl/macos/app_state.rs:303:9
  22: winit::platform_impl::macos::app_state::ApplicationDelegate::cleared
             at winit-0.30.5/src/platform_impl/macos/app_state.rs:365:13
  23: winit::platform_impl::macos::observer::control_flow_end_handler::{{closure}}
             at winit-0.30.5/src/platform_impl/macos/observer.rs:84:21
      winit::platform_impl::macos::observer::control_flow_handler::{{closure}}
             at winit-0.30.5/src/platform_impl/macos/observer.rs:46:9
      std::panicking::try::do_call
             at std/src/panicking.rs:559:40
      std::panicking::try
             at std/src/panicking.rs:523:19
  24: std::panic::catch_unwind
             at std/src/panic.rs:149:14
      winit::platform_impl::macos::event_loop::stop_app_on_panic
             at winit-0.30.5/src/platform_impl/macos/event_loop.rs:435:11
      winit::platform_impl::macos::observer::control_flow_handler
             at winit-0.30.5/src/platform_impl/macos/observer.rs:44:5
      winit::platform_impl::macos::observer::control_flow_end_handler
             at winit-0.30.5/src/platform_impl/macos/observer.rs:79:9
  25: <unknown>
  26: <unknown>
  27: <unknown>
  28: <unknown>
  29: <unknown>
  30: <unknown>
  31: <unknown>
  32: <unknown>
  33: <unknown>
  34: <unknown>
  35: winit::platform_impl::macos::event_loop::EventLoop<T>::run_on_demand::{{closure}}::{{closure}}
             at winit-0.30.5/src/platform_impl/macos/event_loop.rs:306:26
      objc2::rc::autorelease::autoreleasepool
             at objc2-0.5.2/src/rc/autorelease.rs:438:15
      winit::platform_impl::macos::event_loop::EventLoop<T>::run_on_demand::{{closure}}
             at winit-0.30.5/src/platform_impl/macos/event_loop.rs:292:13
      winit::platform_impl::macos::event_handler::EventHandler::set
             at winit-0.30.5/src/platform_impl/macos/event_handler.rs:98:9
  36: winit::platform_impl::macos::app_state::ApplicationDelegate::set_event_handler
             at winit-0.30.5/src/platform_impl/macos/app_state.rs:172:9
      winit::platform_impl::macos::event_loop::EventLoop<T>::run_on_demand
             at winit-0.30.5/src/platform_impl/macos/event_loop.rs:291:9
  37: <winit::event_loop::EventLoop<T> as winit::platform::run_on_demand::EventLoopExtRunOnDemand>::run_on_demand
             at winit-0.30.5/src/platform/run_on_demand.rs:89:9
      winit::platform::run_on_demand::EventLoopExtRunOnDemand::run_app_on_demand
             at winit-0.30.5/src/platform/run_on_demand.rs:75:9
      eframe::native::run::run_and_return
             at eframe-0.29.1/src/native/run.rs:300:16
      eframe::native::run::run_wgpu::{{closure}}
             at eframe-0.29.1/src/native/run.rs:357:13
      eframe::native::run::with_event_loop::{{closure}}
             at eframe-0.29.1/src/native/run.rs:52:12
      std::thread::local::LocalKey<T>::try_with
             at std/src/thread/local.rs:286:12
      std::thread::local::LocalKey<T>::with
             at std/src/thread/local.rs:262:9
      eframe::native::run::with_event_loop
             at eframe-0.29.1/src/native/run.rs:42:16
      eframe::native::run::run_wgpu
             at eframe-0.29.1/src/native/run.rs:355:16
  38: eframe::run_native
             at eframe-0.29.1/src/lib.rs:265:13

Troubleshooting Rerun: https://www.rerun.io/docs/getting-started/troubleshooting 
Report bugs: https://github.com/rerun-io/rerun/issues

Desktop (please complete the following information):

OS: macOS Ventura 13.7 (Intel)

Rerun version
#7500.

Additional context

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

grtlr · 2024-11-04T17:22:16Z

The following .rrd file can be used to reproduce the issue, but most likely will require #7500 to reproduce as well.

test.rrd.zip

The crash does not happen on the following (web) versions:

The crash does happen on:

current main (59773c9)
v0.19.0 (5efb166)

teh-cmc · 2024-11-05T08:41:52Z

I'll have a look but I'm removing the 0.20 milestone blocker in the meantime: these are just my usual too-paranoid-until-proven-wrong debug assertions, they have no impact on release builds whatsoever.

The crash does not happen on the following (web) versions:
* https://rerun.io/viewer/version/0.19.0
* https://rerun.io/viewer/version/nightly

That's because these are release builds.

grtlr · 2024-11-05T14:27:33Z

Sounds good—thank you! I have encountered the assertion again in another context, again when logging a second time to an entity that has been statically logged before.

The use case of (statically) logging two different archetypes to the same entity came up during #7500, where we need to log nodes and edges.

teh-cmc · 2024-11-05T15:09:53Z

Statically logging to the same entity multiple times is perfectly legitimate btw, and even something we encourage in some scenarios (e.g. soft real-time telemetry).

Feel free to comment those debug_assertions in your branch to make your life simpler in the meantime.

teh-cmc · 2024-11-06T14:51:38Z

Repro:

use std::f32::consts::TAU;

use rerun::demo_util::color_spiral;

const NUM_POINTS: usize = 100;

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let rec = rerun::RecordingStreamBuilder::new("rerun_example_dna_abacus").connect_tcp()?;

    rec.set_time_seconds("stable_time", 0f64);

    for _ in 0..100 {
        let (points1, _colors1) = color_spiral(NUM_POINTS, 2.0, 0.02, 0.0, 0.1);
        let (points2, colors2) = color_spiral(NUM_POINTS, 2.0, 0.02, TAU * 0.5, 0.1);
        rec.log_static("/nodes", &rerun::Points3D::new(points1))?;
        rec.log_static(
            "/nodes",
            &rerun::Points3D::new(points2).with_colors(colors2),
        )?;
    }

    for _ in 0..100 {
        let (points1, colors1) = color_spiral(NUM_POINTS, 2.0, 0.02, 0.0, 0.1);
        let (points2, _colors2) = color_spiral(NUM_POINTS, 2.0, 0.02, TAU * 0.5, 0.1);
        rec.log_static(
            "/nodes",
            &rerun::Points3D::new(points1).with_colors(colors1),
        )?;
        rec.log_static("/nodes", &rerun::Points3D::new(points2))?;
    }

    Ok(())
}

Start a viewer in debug:

cargo r -p rerun-cli --no-default-features --features native_viewer,nasm

Run the code above in release (might need to do it a handful of times, it's timing dependent)

@grtlr

…ks (#8020) The assertion was correct 🥳🎈. There was a real memleak in there when dealing with partially dangling static chunks (simplifying a bit: those are static chunks being overwritten for some components but not all, which is a very niche use case, thus why we've never hit it until now). * Fixes #7992 cc @grtlr

grtlr added 🪳 bug Something isn't working 👀 needs triage This issue needs to be triaged by the Rerun team labels Nov 4, 2024

grtlr changed the title ~~Panic due to chunk_id_removed.is_some()~~ Panic due to assertion chunk_id_removed.is_some() Nov 4, 2024

jleibs added this to the 0.20 - Maps, H.264, Undo milestone Nov 4, 2024

abey79 added ⛃ re_datastore affects the datastore itself and removed 👀 needs triage This issue needs to be triaged by the Rerun team labels Nov 4, 2024

emilk added the 💣 crash crash, deadlock/freeze, do-no-start label Nov 5, 2024

teh-cmc removed this from the 0.20 - Maps, H.264, Undo milestone Nov 5, 2024

grtlr mentioned this issue Nov 5, 2024

MVP for the graph view (node-link diagrams) #7897

Closed

49 tasks

teh-cmc self-assigned this Nov 6, 2024

teh-cmc mentioned this issue Nov 6, 2024

ChunkStore: fix row-id computation when removing dangling static chunks #8020

Merged

6 tasks

teh-cmc closed this as completed in #8020 Nov 7, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Panic due to assertion `chunk_id_removed.is_some()` #7992

Panic due to assertion `chunk_id_removed.is_some()` #7992

grtlr commented Nov 4, 2024

grtlr commented Nov 4, 2024 •

edited

Loading

teh-cmc commented Nov 5, 2024 •

edited

Loading

grtlr commented Nov 5, 2024

teh-cmc commented Nov 5, 2024

teh-cmc commented Nov 6, 2024

Panic due to assertion chunk_id_removed.is_some() #7992

Panic due to assertion chunk_id_removed.is_some() #7992

Comments

grtlr commented Nov 4, 2024

grtlr commented Nov 4, 2024 • edited Loading

teh-cmc commented Nov 5, 2024 • edited Loading

grtlr commented Nov 5, 2024

teh-cmc commented Nov 5, 2024

teh-cmc commented Nov 6, 2024

Panic due to assertion `chunk_id_removed.is_some()` #7992

Panic due to assertion `chunk_id_removed.is_some()` #7992

grtlr commented Nov 4, 2024 •

edited

Loading

teh-cmc commented Nov 5, 2024 •

edited

Loading