Skip to content

Fix cuda allocator copy condition #2800

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
awni merged 1 commit into from
Nov 20, 2025
Merged

Fix cuda allocator copy condition #2800

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
8 changes: 4 additions & 4 deletions mlx/backend/cuda/allocator.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -92,7 +92,7 @@ CudaAllocator::CudaAllocator()
[this](CudaBuffer* buf) { cuda_free(buf); }) {
size_t free, total;
CHECK_CUDA_ERROR(cudaMemGetInfo(&free, &total));
memory_limit_ = total * 0.95;
memory_limit_ = total * 0.9;
max_pool_size_ = memory_limit_;

int device_count = 0;
Expand Down Expand Up @@ -176,7 +176,7 @@ CudaAllocator::malloc_async(size_t size, int device, cudaStream_t stream) {
buffer_cache_.release_cached_buffers(get_cache_memory() - max_pool_size_);
}
// Copy to managed here if the buffer is not on the right device
if (buf->device != device) {
if (buf->device >= 0 && buf->device != device) {
copy_to_managed(*buf);
}
Comment on lines +179 to 181

@awni awni Nov 20, 2025

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If the buffer is already managed (device = -1) then don't need a copy here

return Buffer{buf};
Expand Down Expand Up @@ -219,9 +219,9 @@ void CudaAllocator::cuda_free(CudaBuffer* buf) {
scalar_pool_.free(buf);
} else {
if (buf->device >= 0) {
cudaFreeAsync(buf->data, free_streams_[buf->device]);
CHECK_CUDA_ERROR(cudaFreeAsync(buf->data, free_streams_[buf->device]));
} else {
cudaFree(buf->data);
CHECK_CUDA_ERROR(cudaFree(buf->data));
}
delete buf;
}
Expand Down
Loading