[tensor] Update GstTensorMemory coherent with ml_tensors_data_s
authorParichay Kapoor <pk.kapoor@samsung.com>
Tue, 28 Jul 2020 07:23:08 +0000 (16:23 +0900)
committerMyungJoo Ham <myungjoo.ham@samsung.com>
Fri, 31 Jul 2020 11:44:10 +0000 (20:44 +0900)
commit63d6fba3e441c7c514713c28f6309d8c013ccbee
treeb8c04d3d6fd70b0d0d92326ec4d4a1a01e4f9c51
parentf1751bbb9d74d07e72244e9930033221d42cd1eb
[tensor] Update GstTensorMemory coherent with ml_tensors_data_s

One of the efforts to minimize API latency is to avoid creating data wrappers in
API and internally in nnstreamer and avoid its copying. Rather just interpret
the API data structure as internal definition.
This requires internal and external structures to be coherent.
This commit makes GstTensorMemory coherent with ml_tensor_data_s by removing
`tensor_type type` from GstTensorMemory.

type can be remove from GstTensorMemory as GstTensorMemory is used internally
and its usage is always supported with GstTensorInfo which contains the type information.
Correspondingly update the existing elements and extensions which use GstTensorMemory
to use the GstTensorInfo type information.

Signed-off-by: Parichay Kapoor <pk.kapoor@samsung.com>
13 files changed:
api/capi/src/nnstreamer-capi-single.c
ext/nnstreamer/tensor_filter/tensor_filter_nnfw.c
ext/nnstreamer/tensor_filter/tensor_filter_openvino.cc
ext/nnstreamer/tensor_filter/tensor_filter_openvino.hh
ext/nnstreamer/tensor_filter/tensor_filter_python.cc
ext/nnstreamer/tensor_filter/tensor_filter_pytorch.cc
gst/nnstreamer/include/tensor_typedef.h
gst/nnstreamer/tensor_decoder/tensordec.c
gst/nnstreamer/tensor_filter/tensor_filter.c
tests/nnstreamer_filter_armnn/unittest_filter_armnn.cc
tests/nnstreamer_filter_extensions_common/unittest_tizen_template.cc.in
tests/nnstreamer_filter_openvino/unittest_openvino.cc
tests/tizen_nnfw_runtime/unittest_tizen_nnfw_runtime_raw.cc