[Triton][Config] Add muls_add triton kernel and refactor AscendCompilationConfig (#5518)

### What this PR does / why we need it?
Add muls_add triton kernel with related fusion pass. What's more, this
PR refactors `AscendCompilationConfig` and delete `NpugraphExConfig`.

### Does this PR introduce _any_ user-facing change?
None

### How was this patch tested?
CI passed with new added test.


- vLLM version: v0.13.0
- vLLM main:
45c1ca1ca1

---------

Signed-off-by: whx-sjtu <2952154980@qq.com>
This commit is contained in:
whx
2026-03-02 17:54:25 +08:00
committed by GitHub
parent 8547520726
commit 16c879cdf7
14 changed files with 290 additions and 98 deletions

View File

@@ -279,7 +279,7 @@ class NPUPlatform(Platform):
if compilation_config.cudagraph_mode == CUDAGraphMode.NONE:
compilation_config.mode = CompilationMode.NONE
ascend_config.npugraph_ex_config.enable = False
ascend_config.ascend_compilation_config.enable_npugraph_ex = False
elif compilation_config.cudagraph_mode == CUDAGraphMode.PIECEWISE:
logger.info("PIECEWISE compilation enabled on NPU. use_inductor not supported - using only ACL Graph mode")
assert compilation_config.mode == CompilationMode.VLLM_COMPILE, (
@@ -299,7 +299,7 @@ class NPUPlatform(Platform):
# not be detected in advance assert.
compilation_config.splitting_ops.extend(["vllm::mla_forward"])
update_aclgraph_sizes(vllm_config)
ascend_config.npugraph_ex_config.enable = False
ascend_config.ascend_compilation_config.enable_npugraph_ex = False
elif (
compilation_config.cudagraph_mode == CUDAGraphMode.FULL_DECODE_ONLY
or compilation_config.cudagraph_mode == CUDAGraphMode.FULL
@@ -328,7 +328,7 @@ class NPUPlatform(Platform):
)
compilation_config.cudagraph_mode = CUDAGraphMode.NONE
compilation_config.mode = CompilationMode.NONE
ascend_config.npugraph_ex_config.enable = False
ascend_config.ascend_compilation_config.enable_npugraph_ex = False
# TODO: Remove this check when ACL Graph supports ASCEND_LAUNCH_BLOCKING=1
# Then, we will have to discuss the error handling strategy and user experience