Skip to content

[Offload] Remove non-blocking allocation type #159851

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
merged 1 commit into from
Sep 20, 2025
Merged

[Offload] Remove non-blocking allocation type #159851

Show file tree
Hide file tree
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
2 changes: 0 additions & 2 deletions offload/include/omptarget.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -101,8 +101,6 @@ enum TargetAllocTy : int32_t {
TARGET_ALLOC_HOST,
TARGET_ALLOC_SHARED,
TARGET_ALLOC_DEFAULT,
/// The allocation will not block on other streams.
TARGET_ALLOC_DEVICE_NON_BLOCKING,
};

inline KernelArgsTy CTorDTorKernelArgs = {
Expand Down
2 changes: 0 additions & 2 deletions offload/plugins-nextgen/amdgpu/src/rtl.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2350,7 +2350,6 @@ struct AMDGPUDeviceTy : public GenericDeviceTy, AMDGenericDeviceTy {
switch (Kind) {
case TARGET_ALLOC_DEFAULT:
case TARGET_ALLOC_DEVICE:
case TARGET_ALLOC_DEVICE_NON_BLOCKING:
MemoryPool = CoarseGrainedMemoryPools[0];
break;
case TARGET_ALLOC_HOST:
Expand Down Expand Up @@ -3847,7 +3846,6 @@ void *AMDGPUDeviceTy::allocate(size_t Size, void *, TargetAllocTy Kind) {
switch (Kind) {
case TARGET_ALLOC_DEFAULT:
case TARGET_ALLOC_DEVICE:
case TARGET_ALLOC_DEVICE_NON_BLOCKING:
MemoryPool = CoarseGrainedMemoryPools[0];
break;
case TARGET_ALLOC_HOST:
Expand Down
1 change: 0 additions & 1 deletion offload/plugins-nextgen/common/include/ErrorReporting.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -61,7 +61,6 @@ class ErrorReporter {
/// Return a nice name for an TargetAllocTy.
static StringRef getAllocTyName(TargetAllocTy Kind) {
switch (Kind) {
case TARGET_ALLOC_DEVICE_NON_BLOCKING:
case TARGET_ALLOC_DEFAULT:
case TARGET_ALLOC_DEVICE:
return "device memory";
Expand Down
2 changes: 0 additions & 2 deletions offload/plugins-nextgen/common/src/PluginInterface.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1295,7 +1295,6 @@ Expected<void *> GenericDeviceTy::dataAlloc(int64_t Size, void *HostPtr,

switch (Kind) {
case TARGET_ALLOC_DEFAULT:
case TARGET_ALLOC_DEVICE_NON_BLOCKING:
case TARGET_ALLOC_DEVICE:
if (MemoryManager) {
Alloc = MemoryManager->allocate(Size, HostPtr);
Expand Down Expand Up @@ -1386,7 +1385,6 @@ Error GenericDeviceTy::dataDelete(void *TgtPtr, TargetAllocTy Kind) {
int Res;
switch (Kind) {
case TARGET_ALLOC_DEFAULT:
case TARGET_ALLOC_DEVICE_NON_BLOCKING:
case TARGET_ALLOC_DEVICE:
if (MemoryManager) {
Res = MemoryManager->free(TgtPtr);
Expand Down
6 changes: 3 additions & 3 deletions offload/plugins-nextgen/common/src/RPC.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -28,15 +28,15 @@ rpc::Status handleOffloadOpcodes(plugin::GenericDeviceTy &Device,
switch (Port.get_opcode()) {
case LIBC_MALLOC: {
Port.recv_and_send([&](rpc::Buffer *Buffer, uint32_t) {
Buffer->data[0] = reinterpret_cast<uintptr_t>(Device.allocate(
Buffer->data[0], nullptr, TARGET_ALLOC_DEVICE_NON_BLOCKING));
Buffer->data[0] = reinterpret_cast<uintptr_t>(
Device.allocate(Buffer->data[0], nullptr, TARGET_ALLOC_DEVICE));
});
break;
}
case LIBC_FREE: {
Port.recv([&](rpc::Buffer *Buffer, uint32_t) {
Device.free(reinterpret_cast<void *>(Buffer->data[0]),
TARGET_ALLOC_DEVICE_NON_BLOCKING);
TARGET_ALLOC_DEVICE);
});
break;
}
Expand Down
19 changes: 0 additions & 19 deletions offload/plugins-nextgen/cuda/src/rtl.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -587,16 +587,6 @@ struct CUDADeviceTy : public GenericDeviceTy {
Res = cuMemAllocManaged(&DevicePtr, Size, CU_MEM_ATTACH_GLOBAL);
MemAlloc = (void *)DevicePtr;
break;
case TARGET_ALLOC_DEVICE_NON_BLOCKING: {
CUstream Stream;
if ((Res = cuStreamCreate(&Stream, CU_STREAM_NON_BLOCKING)))
break;
if ((Res = cuMemAllocAsync(&DevicePtr, Size, Stream)))
break;
cuStreamSynchronize(Stream);
Res = cuStreamDestroy(Stream);
MemAlloc = (void *)DevicePtr;
}
}

if (auto Err =
Expand Down Expand Up @@ -627,15 +617,6 @@ struct CUDADeviceTy : public GenericDeviceTy {
case TARGET_ALLOC_HOST:
Res = cuMemFreeHost(TgtPtr);
break;
case TARGET_ALLOC_DEVICE_NON_BLOCKING: {
CUstream Stream;
if ((Res = cuStreamCreate(&Stream, CU_STREAM_NON_BLOCKING)))
break;
cuMemFreeAsync(reinterpret_cast<CUdeviceptr>(TgtPtr), Stream);
cuStreamSynchronize(Stream);
if ((Res = cuStreamDestroy(Stream)))
break;
}
}

if (auto Err = Plugin::check(Res, "error in cuMemFree[Host]: %s")) {
Expand Down
1 change: 0 additions & 1 deletion offload/plugins-nextgen/host/src/rtl.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -250,7 +250,6 @@ struct GenELF64DeviceTy : public GenericDeviceTy {
case TARGET_ALLOC_DEVICE:
case TARGET_ALLOC_HOST:
case TARGET_ALLOC_SHARED:
case TARGET_ALLOC_DEVICE_NON_BLOCKING:
MemAlloc = std::malloc(Size);
break;
}
Expand Down
Loading