[Submodule] Add Flash attention v2

[masahi]

Flash attention v2 was released last week and it is claimed to give good speed up compared to existing implementations. I integrated flash v2 into our Relax CUTLASS BYOC and got 2-3 iter / sec speed up on SD v1.5 UNet compared to our existing flow that uses the xformer kernel.

The original code depends on libtorch so it is hard to integrate. A version without torch dep was created in https://github.com/tlc-pack/libflash_attn and I'm adding it as a submodule.

@tqchen @vinx13 @yzh119 @sunggg

[tvm-bot]

Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews from Reviewers by @-ing them in a comment.

• No users to tag found in teams: submodule _{See #10317 for details}

_{Generated by tvm-bot}

[tqchen]

let us also update https://github.com/apache/tvm/blob/main/LICENSE#L248 and https://github.com/apache/tvm/tree/main/licenses

[junrushao]

Very cool work! This will substantially help with TVM performance on stable diffusion and LLM training workloads