Unable to build ubuntu 24.04

[countess427]

Today ran into the following error after installing some NVIDIA lib updates. Previously had been running fine.
Currently on NVIDIA-SMI Driver 550.127.05 CUDA Version: 12.4

~/work/brush$ RUST_BACKTRACE=full cargo run --release
    Finished `release` profile [optimized] target(s) in 0.41s
     Running `target/release/brush_bin`
[2024-12-04T22:49:57Z ERROR wgpu_hal::vulkan::adapter] get_physical_device_surface_formats: ERROR_SURFACE_LOST_KHR
[2024-12-04T22:49:57Z ERROR wgpu_hal::vulkan::adapter] get_physical_device_surface_formats: ERROR_SURFACE_LOST_KHR
[2024-12-04T22:49:57Z ERROR eframe::native::run] Exiting because of error: WGPU error: There was no valid format for the surface at all.
thread 'main' panicked at crates/brush-desktop/src/main.rs:61:14:
called `Result::unwrap()` on an `Err` value: Wgpu(NoSurfaceFormatsAvailable)
stack backtrace:
   0:     0x568056773bfa - <std::sys::backtrace::BacktraceLock::print::DisplayBacktrace as core::fmt::Display>::fmt::h5b6bd5631a6d1f6b
   1:     0x568055e05123 - core::fmt::write::h7550c97b06c86515
   2:     0x56805676f553 - std::io::Write::write_fmt::h7b09c64fe0be9c84
   3:     0x568056773a52 - std::sys::backtrace::BacktraceLock::print::h2395ccd2c84ba3aa
   4:     0x568056774b96 - std::panicking::default_hook::{{closure}}::he19d4c7230e07961
   5:     0x5680567749e0 - std::panicking::default_hook::hf614597d3c67bbdb
   6:     0x568056775137 - std::panicking::rust_panic_with_hook::h8942133a8b252070
   7:     0x568056774fda - std::panicking::begin_panic_handler::{{closure}}::hb5f5963570096b29
   8:     0x5680567740a9 - std::sys::backtrace::__rust_end_short_backtrace::h6208cedc1922feda
   9:     0x568056774c7c - rust_begin_unwind
  10:     0x568054d62960 - core::panicking::panic_fmt::h0c3082644d1bf418
  11:     0x568054d62d66 - core::result::unwrap_failed::hd20b4aa073bda1e2
  12:     0x568054e66baa - tokio::runtime::park::CachedParkThread::block_on::h500135ca182c3910
  13:     0x568054e6a586 - tokio::runtime::context::runtime::enter_runtime::he9f4e666bc1966bd
  14:     0x568054e636ce - tokio::runtime::runtime::Runtime::block_on::h1091f522ef53d6d1
  15:     0x568054e65347 - brush_bin::main::h781452a17ff623b1
  16:     0x568054e6ac73 - std::sys::backtrace::__rust_begin_short_backtrace::h8de96d3b6d8acf0a
  17:     0x568054e69a79 - std::rt::lang_start::{{closure}}::h656ec3be0f89a618
  18:     0x568056767787 - std::rt::lang_start_internal::h1c66660c99c8424c
  19:     0x568054e65455 - main
  20:     0x7f4160429d90 - __libc_start_call_main
                               at ./csu/../sysdeps/nptl/libc_start_call_main.h:58:16
  21:     0x7f4160429e40 - __libc_start_main_impl
                               at ./csu/../csu/libc-start.c:392:3
  22:     0x568054dc6325 - _start
  23:                0x0 - <unknown>

[countess427]

(Resolved with system restart)

[ArthurBrussee]

Oh phew! Strange though. There is a suspiciously similair bug in wgpu that was fixed recently gfx-rs/wgpu#6510 so hopefully this is more solid when the wgpu 23 update is in (just waiting for a burn update)

[countess427]

It was odd since all my other GPU processes (colmap, glomap, gslplat) were still working normally, so something peculiar with wgpu makes sense