COMPMID-713: Address failures in OCLGrind for CLDirectConvolution
-Changes way of clamping in the kernel side.
-Fills padding with quantized values
Change-Id: I94d17c341fd637fbb24390722162b551b62d16cb
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/111114
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
diff --git a/src/core/CL/cl_kernels/direct_convolution_1x1_3x3_5x5_quantized.cl b/src/core/CL/cl_kernels/direct_convolution_1x1_3x3_5x5_quantized.cl
index cbe8266..d0cf032 100644
--- a/src/core/CL/cl_kernels/direct_convolution_1x1_3x3_5x5_quantized.cl
+++ b/src/core/CL/cl_kernels/direct_convolution_1x1_3x3_5x5_quantized.cl
@@ -243,9 +243,7 @@
pixels0 = ASYMM_MULT_BY_QUANT_MULTIPLIER_LESS_THAN_ONE(pixels0, output_multiplier, output_shift, 8);
pixels0 = pixels0 + output_offset;
- pixels0 = max(pixels0, 0);
- pixels0 = min(pixels0, 255);
- vstore8(convert_uchar8(pixels0), 0, (__global uchar *)dst.ptr);
+ vstore8(convert_uchar8_sat(pixels0), 0, (__global uchar *)dst.ptr);
}
#endif // defined(DATA_TYPE) && defined(STRIDE_X) && defined(WEIGHTS_DEPTH)
diff --git a/src/runtime/CL/functions/CLDirectConvolutionLayer.cpp b/src/runtime/CL/functions/CLDirectConvolutionLayer.cpp
index 8ef7068..7596925 100644
--- a/src/runtime/CL/functions/CLDirectConvolutionLayer.cpp
+++ b/src/runtime/CL/functions/CLDirectConvolutionLayer.cpp
@@ -23,6 +23,7 @@
*/
#include "arm_compute/runtime/CL/functions/CLDirectConvolutionLayer.h"
+#include "arm_compute/core/CL/ICLTensor.h"
#include "arm_compute/core/CL/kernels/CLDirectConvolutionLayerKernel.h"
#include "arm_compute/core/PixelValue.h"
#include "arm_compute/core/Utils.h"
@@ -45,7 +46,12 @@
_direct_conv_kernel.configure(input, weights, biases, output, conv_info);
// Configure border handler
- _input_border_handler.configure(input, _direct_conv_kernel.border_size(), BorderMode::CONSTANT, PixelValue(0));
+ PixelValue &&zero_value(0.f);
+ if(is_data_type_quantized_asymmetric(input->info()->data_type()))
+ {
+ zero_value = PixelValue(static_cast<uint8_t>(input->info()->quantization_info().offset));
+ }
+ _input_border_handler.configure(input, _direct_conv_kernel.border_size(), BorderMode::CONSTANT, zero_value);
}
Error CLDirectConvolutionLayer::validate(const ITensorInfo *input, const ITensorInfo *weights, const ITensorInfo *biases, const ITensorInfo *output, const PadStrideInfo &conv_info)
diff --git a/tests/validation/CL/DirectConvolutionLayer.cpp b/tests/validation/CL/DirectConvolutionLayer.cpp
index 08d53d5..e6a196a 100644
--- a/tests/validation/CL/DirectConvolutionLayer.cpp
+++ b/tests/validation/CL/DirectConvolutionLayer.cpp
@@ -196,7 +196,7 @@
TEST_SUITE(Quantized)
TEST_SUITE(QASYMM8)
FIXTURE_DATA_TEST_CASE(Run, CLDirectConvolutionLayerQuantizedFixture<uint8_t>, framework::DatasetMode::ALL, combine(combine(data, framework::dataset::make("DataType", DataType::QASYMM8)),
- framework::dataset::make("QuantizationInfo", { QuantizationInfo(2.f / 255, 127) })))
+ framework::dataset::make("QuantizationInfo", { QuantizationInfo(2.f / 255, 10) })))
{
// Validate output
validate(CLAccessor(_target), _reference, tolerance_qasymm8);
diff --git a/tests/validation/CPP/ConvolutionLayer.cpp b/tests/validation/CPP/ConvolutionLayer.cpp
index a767912..3884a93 100644
--- a/tests/validation/CPP/ConvolutionLayer.cpp
+++ b/tests/validation/CPP/ConvolutionLayer.cpp
@@ -210,8 +210,7 @@
acc = asymm_rounding_divide_by_pow2(asymm_int_mult(acc, output_multiplier), output_shift);
acc += output_offset;
- acc = std::max<int32_t>(acc, 0);
- acc = std::min<int32_t>(acc, 255);
+ acc = clamp<int32_t>(acc, 0, 255);
// Store the result
*out_ptr = acc;
diff --git a/tests/validation/fixtures/DirectConvolutionLayerFixture.h b/tests/validation/fixtures/DirectConvolutionLayerFixture.h
index 1ec4d31..b78f13a 100644
--- a/tests/validation/fixtures/DirectConvolutionLayerFixture.h
+++ b/tests/validation/fixtures/DirectConvolutionLayerFixture.h
@@ -73,7 +73,7 @@
{
case DataType::QASYMM8:
{
- std::uniform_int_distribution<uint8_t> distribution(0, 10);
+ std::uniform_int_distribution<uint8_t> distribution(0, 50);
library->fill(tensor, distribution, i);
break;
}
@@ -86,7 +86,7 @@
}
case DataType::S32:
{
- std::uniform_int_distribution<int32_t> distribution(-1000, 1000);
+ std::uniform_int_distribution<int32_t> distribution(-5, 5);
library->fill(tensor, distribution, i);
break;
}