Skip to content

Deduplicate instances caching allocator, so that we only have one instance. #16720

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

Closed
wants to merge 3 commits into from
Closed

Deduplicate instances caching allocator, so that we only have one instance. #16720

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
e1ebf81
V1: Initial commit
ezyang Feb 4, 2019
81b05e1
V2: Merge with parent diff changes
ezyang Feb 5, 2019
970d5ac
V3: Merge with parent diff changes
ezyang Feb 7, 2019
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
11 changes: 2 additions & 9 deletions c10/cuda/CUDACachingAllocator.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -660,21 +660,14 @@ std::shared_ptr<void> getIpcDevPtr(std::string handle) {
return sp;
}

Legacy::Legacy()
: _impl(new THCCachingAllocator()) {}

Legacy::~Legacy() {
delete _impl;
}

cudaError_t
Legacy::Alloc(void** refPtr, size_t nbytes, cudaStream_t stream) {
_impl->malloc(refPtr, nbytes, stream);
caching_allocator.malloc(refPtr, nbytes, stream);
return cudaSuccess;
}

cudaError_t Legacy::Free(void* ptr) {
_impl->free(ptr);
caching_allocator.free(ptr);
return cudaSuccess;
}

Expand Down
6 changes: 0 additions & 6 deletions c10/cuda/CUDACachingAllocator.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -47,14 +47,8 @@ struct THCCachingAllocator;

class C10_CUDA_API Legacy {
public:
Legacy();
~Legacy();

cudaError_t Alloc(void** refPtr, size_t nbytes, cudaStream_t stream);
cudaError_t Free(void* ptr);

private:
THCCachingAllocator* _impl;
};

} // namespace CUDACachingAllocator
Expand Down