Skip to content

Fix 12.6 build with cudaMemcpyBatchAsync #5799

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
xwang233 merged 3 commits into from
Jan 14, 2026
Merged

Fix 12.6 build with cudaMemcpyBatchAsync #5799

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f2e9b53
Throw instead of calling cudaMemcpyBatchAsyn for Cuda<12.8
samnordmann Jan 12, 2026
76ce6d4
lint
samnordmann Jan 12, 2026
c04a3b8
account for API change
samnordmann Jan 13, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 12 additions & 0 deletions csrc/multidevice/cuda_p2p.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -268,6 +268,11 @@ void postBroadcastWithCudaBackend(
launchMulticastKernel(
multicast_handle->bufferMulticastPtr(), src_ptr, count, stream);
} else if (protocol == MulticastProtocol::BatchMemcpy) {
#if CUDA_VERSION < 12080
NVF_THROW(
"cudaMemcpyBatchAsync backend is not supported for CUDA version < "
"12.8");
#else
std::vector<void*> dsts(world_size);
std::vector<const void*> srcs(world_size, src_ptr);
std::vector<size_t> counts(world_size, count);
Expand Down Expand Up @@ -311,6 +316,7 @@ void postBroadcastWithCudaBackend(
numAttrs,
&failIdx,
(cudaStream_t)stream));
#endif
#endif
} else {
NVFUSER_CUDA_RT_SAFE_CALL(cudaMemcpyAsync(
Expand Down Expand Up @@ -419,6 +425,11 @@ void postAllgatherWithCudaBackend(
count,
stream);
} else if (protocol == MulticastProtocol::BatchMemcpy) {
#if CUDA_VERSION < 12080
NVF_THROW(
"cudaMemcpyBatchAsync backend is not supported for CUDA version < "
"12.8");
#else
std::vector<void*> dsts(world_size);
std::vector<const void*> srcs(world_size, src_ptr);
std::vector<size_t> counts(world_size, count);
Expand Down Expand Up @@ -462,6 +473,7 @@ void postAllgatherWithCudaBackend(
numAttrs,
&failIdx,
(cudaStream_t)stream));
#endif
#endif
} else {
NVFUSER_CUDA_RT_SAFE_CALL(cudaMemcpyAsync(
Expand Down