aboutsummaryrefslogtreecommitdiff
path: root/src/runtime/gpu/cl/utils/ClAuxTensorHandler.h
diff options
context:
space:
mode:
authorGeorgios Pinitas <georgios.pinitas@arm.com>2021-04-22 21:13:21 +0100
committerGeorgios Pinitas <georgios.pinitas@arm.com>2021-05-18 14:48:39 +0000
commit856f66e6c61b77d03f754cd0fa8439891f0e4aca (patch)
treef9379cd0853ac407109e54c3d53b385ceee066c2 /src/runtime/gpu/cl/utils/ClAuxTensorHandler.h
parent37f4b2ef1ea225a90ccb563fcb2c08f8fb0fb5d5 (diff)
downloadComputeLibrary-856f66e6c61b77d03f754cd0fa8439891f0e4aca.tar.gz
Port CLGEMM to memory injecting interface
Moves the following kernels: - CLGEMMMatrixMultiplyKernel - CLGEMMMatrixMultiplyNativeKernel - CLGEMMMatrixMultipluReshapedKernel - CLGEMMMatrixMultiplyReshapedOnlyRHSKernel Moves the following functions - CLGEMM Introduces facilities to easy handling of auxiliary temporary buffers under then new run interface. Such are: - CLAuxTensorHandler: That allows wrapping of workspace buffers memory to CLBuffer objects - Ability to inject TensorInfo to allocator without transferring ownership. This reduce the copy overhead if needed. Resolves: COMPMID-4188 Signed-off-by: Georgios Pinitas <georgios.pinitas@arm.com> Change-Id: I7055435d831b05b749b26302082e4ac45f26dfb0 Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5498 Tested-by: Arm Jenkins <bsgcomp@arm.com> Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com> Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/runtime/gpu/cl/utils/ClAuxTensorHandler.h')
-rw-r--r--src/runtime/gpu/cl/utils/ClAuxTensorHandler.h86
1 files changed, 86 insertions, 0 deletions
diff --git a/src/runtime/gpu/cl/utils/ClAuxTensorHandler.h b/src/runtime/gpu/cl/utils/ClAuxTensorHandler.h
new file mode 100644
index 0000000000..ad893acea5
--- /dev/null
+++ b/src/runtime/gpu/cl/utils/ClAuxTensorHandler.h
@@ -0,0 +1,86 @@
+/*
+ * Copyright (c) 2021 Arm Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef ARM_COMPUTE_CL_UTILS_CL_AUX_TENSOR_HANDLER_H
+#define ARM_COMPUTE_CL_UTILS_CL_AUX_TENSOR_HANDLER_H
+
+#include "arm_compute/core/ITensorPack.h"
+#include "arm_compute/core/TensorInfo.h"
+#include "arm_compute/runtime/CL/CLTensor.h"
+
+#include "support/Cast.h"
+
+namespace arm_compute
+{
+namespace opencl
+{
+/* Tensor handler to wrap and handle tensor allocations on workspace buffers */
+class CLAuxTensorHandler
+{
+public:
+ CLAuxTensorHandler(int slot_id, TensorInfo &info, ITensorPack &pack, bool pack_inject = false)
+ : _tensor()
+ {
+ _tensor.allocator()->soft_init(info);
+
+ ICLTensor *packed_tensor = utils::cast::polymorphic_downcast<ICLTensor *>(pack.get_tensor(slot_id));
+ if((packed_tensor == nullptr) || (info.total_size() > packed_tensor->info()->total_size()))
+ {
+ _tensor.allocator()->allocate();
+ if(pack_inject)
+ {
+ pack.add_tensor(slot_id, &_tensor);
+ }
+ }
+ else
+ {
+ _tensor.allocator()->import_memory(packed_tensor->cl_buffer());
+ }
+ }
+
+ CLAuxTensorHandler(TensorInfo &info, ICLTensor &tensor)
+ : _tensor()
+ {
+ _tensor.allocator()->soft_init(info);
+ if(info.total_size() <= tensor.info()->total_size())
+ {
+ _tensor.allocator()->import_memory(tensor.cl_buffer());
+ }
+ }
+
+ ICLTensor *get()
+ {
+ return &_tensor;
+ }
+
+ ICLTensor *operator()()
+ {
+ return &_tensor;
+ }
+
+private:
+ CLTensor _tensor{};
+};
+} // namespace opencl
+} // namespace arm_compute
+#endif /* ARM_COMPUTE_CL_UTILS_CL_AUX_TENSOR_HANDLER_H */ \ No newline at end of file