imgui UI out of bounds in example "imgui-winit"

[martinteoharov]

Whenever I switch between my monitors and run my program I get an instant crash before the first render. I narrowed it down to the UI drawing to a slightly bigger frame. I tried to recreate it in the example mentioned in the title and I also got the same error. A restart of my PC fixes it sometimes.

[2023-02-24T15:14:51Z ERROR wgpu::backend::direct] Handling wgpu errors as fatal by default
thread 'main' panicked at 'wgpu error: Validation Error

Caused by:
    In a RenderPass
      note: encoder = `pixels_command_encoder`
    In a set_scissor_rect command
    Scissor Rect { x: 0, y: 0, w: 911, h: 650 } is not contained in the render target Extent3d { width: 910, height: 650, depth_or_array_layers: 1 }

I tried tweaking the code, but I couldn't find a solution. Any advice?

[parasyte]

It looks like an off by one based on the error message, but this is a new failure mode to me.

Do your displays have different logical pixel sizes? Maybe that has something to do with it. I honestly have not done a lot of testing with multiple displays and handling changes to the scaling factor. I can put together a setup and try to reproduce this.

Because that example uses the scaling factor within its resize handler, I suspect these two events need to be handled simultaneously. That's my current theory.

[martinteoharov]

thanks @parasyte . My setup is basically a laptop that i connect to 2 different monitors (1920x1080 and 1920x1200) throughout the day (not at the same time).

ill have a look tomorrow and will try to come up with a solution, thanks for the advice and for the great project, ive found it very useful :))

[tommythorn]

I just cloned the repo and ran through some examples and hit this. My message was a bit different though:

    Finished `dev` profile [unoptimized + debuginfo] target(s) in 14.25s
     Running `/Users/tommy/projects/Rust/pixels/target/debug/imgui-winit`
[2024-11-12T22:16:56Z ERROR imgui_winit] pixels.resize_surface() failed: Texture width is invalid: 4294967295
[2024-11-12T22:16:56Z ERROR wgpu::backend::direct] Handling wgpu errors as fatal by default
thread 'main' panicked at /Users/tommy/.cargo.x86_64-darwin/registry/src/index.crates.io-6f17d22bba15001f/wgpu-0.17.2/src/backend/direct.rs:3056:5:
wgpu error: Validation Error

Caused by:
    In a RenderPass
      note: encoder = `pixels_command_encoder`
    In a set_scissor_rect command
    Scissor Rect { x: 0, y: 0, w: 4294967295, h: 4294967295 } is not contained in the render target Extent3d { width: 1280, height: 960, depth_or_array_layers: 1 }


note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
-[_MTLCommandEncoder dealloc]:134: failed assertion `Command encoder released without endEncoding'
Abort trap: 6

This is on an iMac 5K (x86) running the latest OS and run: 15.1 and Rust rustc 1.82.0 (f6e511eec 2024-10-15).

[tommythorn]

The (truncated) backtrace:

stack backtrace:
   0: rust_begin_unwind
             at /rustc/f6e511eec7342f59a25f7c0534f1dbea00d01b14/library/std/src/panicking.rs:662:5
   1: core::panicking::panic_fmt
             at /rustc/f6e511eec7342f59a25f7c0534f1dbea00d01b14/library/core/src/panicking.rs:74:14
   2: wgpu::backend::direct::default_error_handler
             at /Users/tommy/.cargo.x86_64-darwin/registry/src/index.crates.io-6f17d22bba15001f/wgpu-0.17.2/src/backend/direct.rs:3056:5
   3: core::ops::function::Fn::call
             at /rustc/f6e511eec7342f59a25f7c0534f1dbea00d01b14/library/core/src/ops/function.rs:79:5
   4: <alloc::boxed::Box<F,A> as core::ops::function::Fn<Args>>::call
             at /rustc/f6e511eec7342f59a25f7c0534f1dbea00d01b14/library/alloc/src/boxed.rs:2245:9
   5: wgpu::backend::direct::ErrorSinkRaw::handle_error
             at /Users/tommy/.cargo.x86_64-darwin/registry/src/index.crates.io-6f17d22bba15001f/wgpu-0.17.2/src/backend/direct.rs:3042:17
   6: wgpu::backend::direct::Context::handle_error
             at /Users/tommy/.cargo.x86_64-darwin/registry/src/index.crates.io-6f17d22bba15001f/wgpu-0.17.2/src/backend/direct.rs:324:9
   7: <wgpu::backend::direct::Context as wgpu::context::Context>::command_encoder_end_render_pass
             at /Users/tommy/.cargo.x86_64-darwin/registry/src/index.crates.io-6f17d22bba15001f/wgpu-0.17.2/src/backend/direct.rs:1946:13
   8: <T as wgpu::context::DynContext>::command_encoder_end_render_pass
             at /Users/tommy/.cargo.x86_64-darwin/registry/src/index.crates.io-6f17d22bba15001f/wgpu-0.17.2/src/context.rs:2796:9
   9: <wgpu::RenderPass as core::ops::drop::Drop>::drop
             at /Users/tommy/.cargo.x86_64-darwin/registry/src/index.crates.io-6f17d22bba15001f/wgpu-0.17.2/src/lib.rs:3898:13
  10: core::ptr::drop_in_place<wgpu::RenderPass>
             at /rustc/f6e511eec7342f59a25f7c0534f1dbea00d01b14/library/core/src/ptr/mod.rs:574:1
  11: imgui_winit::gui::Gui::render
             at ./src/gui.rs:128:5
  12: imgui_winit::main::{{closure}}::{{closure}}
             at ./src/main.rs:74:17

[tommythorn]

Clarification: this is a plain vanilla iMac setup and no second monitor. I don't think there's anything unusual about my setup.

[parasyte]

Scissor Rect { x: 0, y: 0, w: 4294967295, h: 4294967295 }

This is indicative of the window being hidden. Which should not cause a draw. (I would consider it a bug to redraw while the window is hidden. But I don't remember ever making a conscious decision to avoid redraws on hidden windows. It was identified in #174, though.)

Also, see this discussion: #400.