GitLab

feat(v1 attention): 为 ROCm FlashAttention 接入 unified kv layout，并打通 mm_prefix、qq_bias 与 use_alibi_sqrt 透传
在 ROCm FlashAttention 后端增加 unified KV layout 选择逻辑
接入 unified varlen kernel 调用路径
在 FlashAttention metadata 中补充 mm_prefix_range 与 qq_bias 透传

set fp8_e4m3 only supported on nmz and support q&kvcache fp8
set VLLM_PCIE_USE_CUSTOM_ALLREDUCE=1

Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Michael Goin <mgoin64@gmail.com>
(cherry picked from commit 0a3c71e7)

add prepare_so_files to prepare so

not supported FlashMLASchedMeta

Signed-off-by: NickLucche <nlucches@redhat.com>
(cherry picked from commit 1f3a2c29)

Signed-off-by: ElizaWszola <ewszola@redhat.com>
Signed-off-by: mgoin <mgoin64@gmail.com>
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Signed-off-by: Luka Govedič <luka.govedic@gmail.com>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Signed-off-by: Luka Govedič <lgovedic@redhat.com>
Co-authored-by: mgoin <mgoin64@gmail.com>
Co-authored-by: Varun Sundar Rabindranath <varunsundar08@gmail.com>
Co-authored-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: Luka Govedič <luka.govedic@gmail.com>
Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Luka Govedič <lgovedic@redhat.com>

Signed-off-by: Eldar Kurtic <8884008+eldarkurtic@users.noreply.github.com>
Signed-off-by: eldarkurtic <8884008+eldarkurtic@users.noreply.github.com>

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>

区分pcie和hglink custom allreduce的使用
vllm：export VLLM_CUSTOM_CACHE=1
dtk：export HIP_KERNEL_EVENT_SYSTENFENCE=1

set VLLM_USE_FUSED_RMS_ROPE=1
add SUPPORT_MOE_MARLIN_W16A16 to use moe marlin on bw
support fa kvcache fp8 (todo: add VLLM_USE_QUERY_QUANT to not use q quant)
update moe_align_block_size

fix _forward_encoder_attention
remove medusa
set VLLM_PCIE_USE_CUSTOM_ALLREDUCE=1

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>

Signed-off-by: Zhuohao Yang <zy242@cornell.edu>
Co-authored-by: Zhuohao Yang <zy242@cornell.edu>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>

Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Stanislaw Wozniak <stw@zurich.ibm.com>

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Signed-off-by: Matthew Bonanni <mbonanni001@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>

add VLLM_USE_OPT_RESHAPE_AND_CACHE、VLLM_USE_FUSE_SILU_AND_MUL and VLLM_USE_TOPK_RENORM for qwen3-30b

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>

[Attention][Async] Eliminate `seq_lens_cpu` in FlashAttention metadata building with DCP > 1 (#29449)
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>

Signed-off-by: NickLucche <nlucches@redhat.com>

Signed-off-by: Or Ozeri <oro@il.ibm.com>

Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
Signed-off-by: FENP <yuanyongjie.yyj@antgroup.com>
Signed-off-by: LookAround <lixushi@huawei.com>
Signed-off-by: Jingchun Gao <gaojingchun1@huawei.com>
Signed-off-by: zhenwenqi2024 <zhenwenqi_2022@qq.com>
Co-authored-by: FENP <yuanyongjie.yyj@antgroup.com>
Co-authored-by: LookAround <lixushi@huawei.com>
Co-authored-by: Jingchun Gao <gaojingchun1@huawei.com>
Co-authored-by: zhenwenqi2024 <zhenwenqi_2022@qq.com>
Co-authored-by: Jingchun Gao <63247409+gjc0824@users.noreply.github.com>

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>

Signed-off-by: Huamin Li <3ericli@gmail.com>

[Perf] Refactor cudagraph_support to enable full CUDA graphs for spec decoding with FlashInfer (#28479)
Signed-off-by: Benjamin Chislett <bchislett@nvidia.com>

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Signed-off-by: Matthew Bonanni <mbonanni001@gmail.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>

Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>

Signed-off-by: zhangsicheng5 <zhangsicheng5@huawei.com>
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
Signed-off-by: Qiu <qiuchunshuo@huawei.com>
Co-authored-by: QiuChunshuo <qiuchunshuo@huawei.com>

Signed-off-by: 22quinn <33176974+22quinn@users.noreply.github.com>

Signed-off-by: Kunshang Ji <kunshang.ji@intel.com>

Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com>

Signed-off-by: Yeshwanth Surya <yeshsurya@gmail.com>
Signed-off-by: Yeshwanth N <yeshsurya@gmail.com>
Signed-off-by: yeshsurya <yeshsurya@gmail.com>