diff options
| author | Giorgio Arena <giorgio.arena@arm.com> | 2018-08-23 12:00:02 +0100 |
|---|---|---|
| committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:54:54 +0000 |
| commit | 226e4b92b191491ffa57ede66eba1d5d6fcf3b76 (patch) | |
| tree | 334705a1e743e3465400208d582cf0b25bf950fa /arm_compute/core/utils/misc/ShapeCalculator.h | |
| parent | 35aea3776449557c438e264bae7af5b1fe0e5ff6 (diff) | |
| download | ComputeLibrary-226e4b92b191491ffa57ede66eba1d5d6fcf3b76.tar.gz | |
COMPMID-1470 Add auto-init of the output in NECol2im
The output of NECol2Im is already auto-initialized.
This patch is about calling ShapeCalculator instead of computing the shape inside the kernel, adding validate_and_configure_window, and standardize the way convolved dims are passed (now NEON uses Size2D, while CL passes a pair of uint values: using Size2D for both implementations)
Change-Id: I795696e1b6532f57847c3186c1b532c09f5a25da
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/145345
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Diffstat (limited to 'arm_compute/core/utils/misc/ShapeCalculator.h')
| -rw-r--r-- | arm_compute/core/utils/misc/ShapeCalculator.h | 10 |
1 files changed, 5 insertions, 5 deletions
diff --git a/arm_compute/core/utils/misc/ShapeCalculator.h b/arm_compute/core/utils/misc/ShapeCalculator.h index c40e7119b2..09f558d8b0 100644 --- a/arm_compute/core/utils/misc/ShapeCalculator.h +++ b/arm_compute/core/utils/misc/ShapeCalculator.h @@ -141,18 +141,18 @@ inline TensorShape compute_reductionB_shape(const ITensorInfo &a) return shape_vector_sum_row; } -inline TensorShape compute_col2im_shape(const ITensorInfo &input, std::pair<unsigned int, unsigned int> convolved_dims, unsigned int num_groups = 1) +inline TensorShape compute_col2im_shape(const ITensorInfo &input, const Size2D &convolved_dims, bool batch_size_on_z, unsigned int num_groups = 1) { ARM_COMPUTE_ERROR_ON(num_groups == 0); - ARM_COMPUTE_ERROR_ON(input.tensor_shape()[1] != (convolved_dims.first * convolved_dims.second)); + ARM_COMPUTE_ERROR_ON(input.tensor_shape()[1] != (convolved_dims.area())); ARM_COMPUTE_ERROR_ON((num_groups > 1) && input.tensor_shape()[2] != num_groups); TensorShape col2im_shape{ input.tensor_shape() }; - col2im_shape.set(0, convolved_dims.first); - col2im_shape.set(1, convolved_dims.second); + col2im_shape.set(0, convolved_dims.width); + col2im_shape.set(1, convolved_dims.height); col2im_shape.set(2, input.tensor_shape()[0] * num_groups); - const unsigned int batch_idx = (num_groups == 1) ? 2 : 3; + const unsigned int batch_idx = (batch_size_on_z && num_groups == 1) ? 2 : 3; col2im_shape.set(3, input.tensor_shape()[batch_idx]); return col2im_shape; |