Develop stream 2024-07-01

[Naraenda]

This PR brings various updates, intended for ROCm 6.3.

It contains the following merge commits:

• 7ea783fc perf: improve performance on rocm 6.2.
• 1a40ddc2 fix: performance regression introduced with hip graph support.
• 6a7e0ff3 docs: add documentation for hip graph support
• a94ea245 feat: hip graph support for poisson distribution
• 2b55fd26 feat: host implementation for mt19937
• addbf170 perf: use alias method instead of binary search in cdf
• bfba2925 docs: reference GPU_TARGETS instead of AMDGPU_TARGET
• 798acc6e fix: fix host generators not always supporting large sizes
• 8f27c488 refactor: remove deprecated internal headers
• ed8541d6 test: improve accuracy of poisson histogram test
• 894abb4c docs: fix styling

This PR does not contain fixes for the recent performance regressions, we'll either add them here or create a new PR for those.

[Naraenda]

@samjwu The clang format CI step seems to not handle external PRs properly, I think this might also be happening in other repositories.

[Naraenda]

@stanleytsang-amd Let me know if I should squash the commits into their respective merge commits.

[stanleytsang-amd]

@samjwu The clang format CI step seems to not handle external PRs properly, I think this might also be happening in other repositories. @samjwu Can you take a look at the clang-format CI?

Run git fetch origin develop From https://github.com/ROCm/rocRAND

• branch develop -> FETCH_HEAD
• [new branch] develop -> origin/develop fatal: couldn't find remote ref develop_stream_20240701 Error: Process completed with exit code 128.

[Naraenda]

Rebased to get the clang-format ci fixes.

[Naraenda]

Rebased and added threefry fixes by @ex-rzr