[TRTLLM-9431][perf] Enable multistream for Linear Attention in Qwen3-…#9696

nv-guomingz merged 1 commit intoNVIDIA:mainNVIDIA/TensorRT-LLM:mainfrom

nv-guomingz:user/guomingz/enable_attn_multi_streamnv-guomingz/TensorRT-LLM:user/guomingz/enable_attn_multi_streamCopy head branch name to clipboard