sagudev

Here is similar crash from another test with `debug=true`: ``` Running WPT tests with /home/user/Desktop/servo/target/production/servo Running 1 tests in web-platform-tests This GDB supports auto-downloading debuginfo from the following URLs: Enable...

I suspect something is wrong with hasher for url when hashing this urls for specific random states.

https://github.com/rust-lang/rust/blob/79e9716c980570bfd1f666e3b16ac583f0168962/library/std/src/collections/hash/map.rs#L3143

on first test: ``` 0x0000555558868c61 in hashbrown::map::HashMap::insert (self=0x7fffd967dc80, k=..., v=) at /cargo/registry/src/index.crates.io-6f17d22bba15001f/hashbrown-0.14.0/src/map.rs:1747 #0 0x0000555558868c61 in hashbrown::map::HashMap::insert (self=0x7fffd967dc80, k=..., v=()) at /cargo/registry/src/index.crates.io-6f17d22bba15001f/hashbrown-0.14.0/src/map.rs:1747 #1 0x000055555854148b in hashbrown::set::HashSet::insert (self=, value=...) at /cargo/registry/src/index.crates.io-6f17d22bba15001f/hashbrown-0.14.0/src/set.rs:1113 #2...

On `_webgpu/webgpu/cts.https.html?q=webgpu:shader,execution,expression,unary,i32_conversion:f16:*`: ``` Process 203675 stopped and restarted: thread 1 received signal: SIGCHLD Process 203675 stopped and restarted: thread 1 received signal: SIGCHLD Process 203675 stopped and restarted: thread 1...

on `_webgpu/webgpu/cts.https.html?q=webgpu:shader,execution,expression,call,builtin,dot:f32_vec4:*`: ``` Process 227947 stopped and restarted: thread 1 received signal: SIGCHLD Process 227947 stopped and restarted: thread 1 received signal: SIGCHLD Process 227947 stopped and restarted: thread 1...

Same report with ASan: [asan_crash.txt](https://github.com/servo/servo/files/14403702/asan_crash.txt)

I check all three examples with [asan and using system allocator](https://github.com/servo/servo/pull/31194/commits/2f8de65c5fa52b89636dbf3f50c8e665560dae2e) and all fail in hashbrown. Here are ASAN logs: [`_webgpu/webgpu/cts.https.html?q=webgpu:shader,execution,expression,unary,u32_conversion:u32:*`](https://github.com/servo/servo/files/14431811/u32_conversion_u32.txt) [`_webgpu/webgpu/cts.https.html?q=webgpu:shader,execution,expression,unary,i32_conversion:f16:*`](https://github.com/servo/servo/files/14431813/i32_conversion_f16.txt) [`_webgpu/webgpu/cts.https.html?q=webgpu:shader,execution,expression,call,builtin,dot:f32_vec4:*`](https://github.com/servo/servo/files/14431814/dot_f32_vec4_crash.txt)

> I wonder if upgrading to a newer rust version would fix this? I just tested on nightly-2024-03-01 and crashes are still happening (testing commit: https://github.com/servo/servo/commit/ed20907af347a554536965ff3df68db0bc96d0b2).

Switching to IndexSet/Map and/or changing hasher does not help. All crashes are caused by doing alloc (resize to be exact).