AI/rocm-systems
2
0 Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

rearranged how the min and max functions are part of msccl (#1025)

Browse Source 
* rearranged how the min and max functions are part of msccl

* added more coverage on in place graph tests
This commit is contained in:
akolliasAMD authored and GitHub committed 2023-12-21 08:58:33 -07:00
1 parent 7e1cbb440d
commit f4858e14b2
4 files changed
+8 -8
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

src/include/msccl/msccl_kernel.h
+1 -1
View File
@@ -39,8 +39,8 @@ __global__ void MSCCL_KERNEL_ENTRY_NAME(devredop, type, proto, fullOps)(struct n
#define MSCCL_DECL_KERNEL_ENTRY_FUNC() \
MSCCL_DECL_KERNEL_ENTRY_FUNC_DEVREDOP(Sum, false) \
MSCCL_DECL_KERNEL_ENTRY_FUNC_DEVREDOP(Prod, false) \
MSCCL_DECL_KERNEL_ENTRY_FUNC_DEVREDOP(Min, false) \
MSCCL_DECL_KERNEL_ENTRY_FUNC_DEVREDOP(Max, false) \
MSCCL_DECL_KERNEL_ENTRY_FUNC_DEVREDOP(Min, false) \
MSCCL_DECL_KERNEL_ENTRY_FUNC_DEVREDOP(PreMulSum, false) \
MSCCL_DECL_KERNEL_ENTRY_FUNC_DEVREDOP_NOFLOAT(SumPostDiv, false)