restore matmul_allreduce_add_rmsnrom aclnn interface (#5119)
**What this PR does / why we need it?**
restore a2 matmul_allreduce_add_rmsnrom kernel aclnn interface
**How was this patch tested?**
- vLLM version: v0.12.0
- vLLM main:
ad32e3e19c
Signed-off-by: tongrunze <t00574058@china.huawei.com>
Co-authored-by: tongrunze <t00574058@china.huawei.com>
This commit is contained in:
@@ -26,6 +26,10 @@ enum NnopbaseHcclServerType {
|
||||
};
|
||||
extern "C" void __attribute__((weak)) NnopbaseSetHcclServerType(void *executor, NnopbaseHcclServerType sType);
|
||||
|
||||
#ifdef __cplusplus
|
||||
extern "C" {
|
||||
#endif
|
||||
|
||||
extern aclnnStatus aclnnInnerMatmulAllreduceAddRmsnormGetWorkspaceSize(
|
||||
const aclTensor *x1,
|
||||
const aclTensor *x2,
|
||||
@@ -48,10 +52,6 @@ extern aclnnStatus aclnnInnerMatmulAllreduceAddRmsnorm(
|
||||
aclOpExecutor *executor,
|
||||
aclrtStream stream);
|
||||
|
||||
#ifdef __cplusplus
|
||||
extern "C" {
|
||||
#endif
|
||||
|
||||
aclnnStatus aclnnMatmulAllreduceAddRmsnormGetWorkspaceSize(
|
||||
const aclTensor *x1,
|
||||
const aclTensor *x2,
|
||||
|
||||
Reference in New Issue
Block a user