Paul Fultz II

I think you will need to suppress `cppcoreguidelines-use-default-member-init`, and `bugprone-switch-missing-default-case,-warnings-as-errors`. For the `cppcoreguidelines-noexcept-swap` warning you can just add `noexcept` to the swap function.

> This still needed or do we want to close this and aim to merge in 6.2 once it's ready? We should try to get this merged in. We just...

> Can we target 6.1.3 then instead of 6.1? I opened #3335 to upgrade to this version and fix the tidy issues.

> Without this we attempt to get mlir to try to create int8 x uint8 which fail at compile. Can you explain this in more detail?

But why are we getting mixed types like this in the first place? I am wondering if we should have a pass to fix this up instead of doing this...

This should be a different operator. We can add a `compile_blas` similar to `compile_miopen` that can do the tuning for convolution through the [`compile`](https://github.com/ROCm/AMDMIGraphX/blob/e1f75de84e231de6673e31fcd9f4b5ffb7946c68/src/include/migraphx/operation.hpp#L519) function in operation interface. We just...

> ROCm used for testing (in the docker container) must be of the same version as the base installation; otherwise failures are possible The rocm versions should be backwards-compatible(except releases...

> To avoid unpleasant consequences, I suggest marking exported functions as noexcept. `noexcept` wont prevent us from throwing an exception. The application will just abort instead. Using `noexcept` is just...

> @firewave > > I do not really understand how the CI error in undefined behaviour sanitizer build has anything to do with my changes. > > ``` > Run...

You should capture the onnxruntime results and just create a ref test.