[Diffusion] Refactor fuse qkv with QKVParallelLinear linear

[BBuf]

Motivation

Follow #14505 , Refactor fuse qkv with QKVParallelLinear linear to unify code style.

sglang generate --model-path black-forest-labs/FLUX.1-dev  \
    --prompt "A logo With Bold Large text: SGL Diffusion" --tp-size 1 --height 1024 --width 1024

sglang generate --model-path black-forest-labs/FLUX.2-dev  \
    --prompt "A logo With Bold Large text: SGL Diffusion" --tp-size 1 --height 1024 --width 1024

sglang generate --model-path Qwen/Qwen-Image --prompt "A logo With Bold Large text: SGL Diffusion" 

Profile

Qwen-Image

main:

pr:

QKVParallelLinear reduce some cpu overhead, sppedup from 248us->176us.

Modifications

Accuracy Tests

Benchmarking and Profiling

Checklist

• Format your code according to the Format code with pre-commit.
• Add unit tests according to the Run and add unit tests.
• Update documentation according to Write documentations.
• Provide accuracy and speed benchmark results according to Test the accuracy and Benchmark the speed.
• Follow the SGLang code style guidance.
• Work with maintainers to merge your PR. See the PR Merge Process

[gemini-code-assist]

Warning

You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again!

[mickqian]

/tag-and-rerun-ci