Getting a seg fault with no error or line number

[JohnAllen]

Getting a seg fault with no error printed. I'm not using any other profilers. Tried a bunch of things (turning off flame graph and not using parallel iterations and stuff) but to no avail. Using 0.11.1. Tried --release and without it. Sorry I can't help more.

The only thing i do get is a number before segmentation fault which changes, e.g. 2953 segmentation fault and 3108 segmentation fault. Not sure what that is as I'm still new to Rust.

[0xPoe]

Could you please provide more information about your environment? Or could provide a minimal project which can reproduce it.

[ccqpein]

@hi-rustin . I think I got the same problem when I was benchmarking my code. I make a minimal piece of code to reproduce the issue here. Copy this bench file and run it can cause the issue.

• cargo bench --bench bench-fault works well
• give --profile-time will cause issues like: cargo bench --bench bench-fault -- --profile-time=1

I think it is because of format! macro.

• I don't know why it gives me signal: 10, SIGBUS: access to undefined memory
• I am not sure if the format! is the only reason my code have this issue.

My env:

• macOS 13.3.1
• M2 CPU
• rustc 1.71.0-nightly (4a59ba4d5 2023-05-12)