merge with develop

.github/CODEOWNERS

@@ -1,8 +1,8 @@
-# Documentation files
-docs/* @saadrahim @LisaDelaney
-*.md  @saadrahim @LisaDelaney
-*.rst  @saadrahim @LisaDelaney
-# Header directory
-library/include/*  @saadrahim @LisaDelaney @kiritigowda @rrawther
 # Source code
 @kiritigowda @rrawther
+# Documentation files
+docs/* @ROCm/rocm-documentation
+*.md @ROCm/rocm-documentation
+*.rst @ROCm/rocm-documentation
+# Header directory
+library/include/* @ROCm/rocm-documentation @kiritigowda @rrawther

.jenkins/precheckin.groovy

@@ -47,7 +47,7 @@ ci: {
     def propertyList = ["compute-rocm-dkms-no-npi-hipclang":[pipelineTriggers([cron('0 1 * * 0')])]]
     propertyList = auxiliary.appendPropertyList(propertyList)
 
-    def jobNameList = ["compute-rocm-dkms-no-npi-hipclang":([ubuntu22:['gfx908'], ubuntu20:['gfx906'], centos8:['gfx908']])]
+    def jobNameList = ["compute-rocm-dkms-no-npi-hipclang":([ubuntu20:['gfx90a'], ubuntu22:['gfx1101'], sles15sp1:['gfx908'], rhel8:['gfx1030'], rhel9:['gfx908']])]
     jobNameList = auxiliary.appendJobNameList(jobNameList)
 
     propertyList.each 

CMakeLists.txt

@@ -346,6 +346,7 @@ install(FILES ${PROJECT_BINARY_DIR}/include/rpp_backend.h
 # install Test
 install(DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}/cmake DESTINATION ${CMAKE_INSTALL_DATADIR}/${PROJECT_NAME}/test COMPONENT test)
 install(DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}/utilities/test_suite/ DESTINATION ${CMAKE_INSTALL_DATADIR}/${PROJECT_NAME}/test COMPONENT test)
+install(DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}/utilities/rpp-performancetests DESTINATION ${CMAKE_INSTALL_DATADIR}/${PROJECT_NAME}/test COMPONENT test)
 
 # set license information
 set(CPACK_RESOURCE_FILE_LICENSE  "${CMAKE_CURRENT_SOURCE_DIR}/LICENSE")

docs/sphinx/requirements.in

@@ -1 +1 @@
-rocm-docs-core[api_reference]==0.33.0
+rocm-docs-core[api_reference]==0.35.0

docs/sphinx/requirements.txt

@@ -112,7 +112,7 @@ requests==2.28.2
     # via
     #   pygithub
     #   sphinx
-rocm-docs-core[api-reference]==0.33.0
+rocm-docs-core[api-reference]==0.35.0
     # via
     #   -r requirements.in
     #   rocm-docs-core

include/rppdefs.h

@@ -116,8 +116,8 @@ typedef enum
     RPP_ERROR_NOT_ENOUGH_MEMORY         = -16,
     /*! \brief Out of bound source ROI \ingroup group_rppdefs */
     RPP_ERROR_OUT_OF_BOUND_SRC_ROI      = -17,
-    /*! \brief src and dst layout mismatch \ingroup group_rppdefs */
-    RPP_ERROR_SRC_DST_LAYOUT_MISMATCH   = -18
+    /*! \brief Number of channels is invalid. (Needs to adhere to function specification.) \ingroup group_rppdefs */
+    RPP_ERROR_INVALID_CHANNELS          = -18
 } RppStatus;
 
 /*! \brief RPP rppStatus_t type enums

include/rppi_arithmetic_operations.h

@@ -320,4 +320,4 @@ RppStatus rppi_tensor_multiply_u8_gpu(RppPtr_t srcPtr1, RppPtr_t srcPtr2, RppPtr
 }
 #endif
 
-#endif
+#endif

include/rppt_tensor_audio_augmentations.h

@@ -95,7 +95,22 @@ RppStatus rppt_to_decibels_host(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_
  */
 RppStatus rppt_pre_emphasis_filter_host(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t dstPtr, RpptDescPtr dstDescPtr, Rpp32s *srcLengthTensor, Rpp32f *coeffTensor, RpptAudioBorderType borderType, rppHandle_t rppHandle);
 
+/*! \brief Down Mixing augmentation on HOST backend
+* \details Down Mixing augmentation for audio data
+* \param[in] srcPtr source tensor in HOST memory
+* \param[in] srcDescPtr source tensor descriptor (Restrictions - numDims = 3, offsetInBytes >= 0, dataType = F32)
+* \param[out] dstPtr destination tensor in HOST memory
+* \param[in] dstDescPtr destination tensor descriptor (Restrictions - numDims = 3, offsetInBytes >= 0, dataType = F32)
+* \param[in] srcDimsTensor source audio buffer length and number of channels (1D tensor in HOST memory, of size batchSize * 2)
+* \param[in] normalizeWeights bool flag to specify if normalization of weights is needed
+* \param[in] rppHandle RPP HOST handle created with <tt>\ref rppCreateWithBatchSize()</tt>
+* \return A <tt> \ref RppStatus</tt> enumeration.
+* \retval RPP_SUCCESS Successful completion.
+* \retval RPP_ERROR* Unsuccessful completion.
+*/
+RppStatus rppt_down_mixing_host(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t dstPtr, RpptDescPtr dstDescPtr, Rpp32s *srcDimsTensor, bool normalizeWeights, rppHandle_t rppHandle);
+
 #ifdef __cplusplus
 }
 #endif
-#endif // RPPT_TENSOR_AUDIO_AUGMENTATIONS_H
+#endif // RPPT_TENSOR_AUDIO_AUGMENTATIONS_H

include/rppt_tensor_color_augmentations.h

@@ -417,6 +417,48 @@ RppStatus rppt_lut_host(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t dstPtr
 RppStatus rppt_lut_gpu(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t dstPtr, RpptDescPtr dstDescPtr, RppPtr_t lutPtr, RpptROIPtr roiTensorPtrSrc, RpptRoiType roiType, rppHandle_t rppHandle);
 #endif // GPU_SUPPORT
 
+/*! \brief Color Temperature augmentation on HOST backend for a NCHW/NHWC layout tensor
+ * \details The color temperature augmentation does a image temperature adjustment operation, taking a pixel adjustment value as argument for each image in a batch of RGB(3 channel) with an NHWC/NCHW tensor layout.<br>
+ * - srcPtr depth ranges - Rpp8u (0 to 255), Rpp16f (0 to 1), Rpp32f (0 to 1), Rpp8s (-128 to 127).
+ * - dstPtr depth ranges - Will be same depth as srcPtr.
+ * \image html img150x150.jpg Sample Input
+ * \image html color_augmentations_color_temperature_img150x150.jpg Sample Output
+ * \param [in] srcPtr source tensor in HOST memory
+ * \param [in] srcDescPtr source tensor descriptor (Restrictions - numDims = 4, offsetInBytes >= 0, dataType = U8/F16/F32/I8, layout = NCHW/NHWC, c = 1/3)
+ * \param [out] dstPtr destination tensor in HOST memory
+ * \param [in] dstDescPtr destination tensor descriptor (Restrictions - numDims = 4, offsetInBytes >= 0, dataType = U8/F16/F32/I8, layout = NCHW/NHWC, c = same as that of srcDescPtr)
+ * \param [in] adjustmentValueTensor adjustment values for color temperature calculation (1D tensor of size sizeof(Rpp8s) * batchSize with -100 <= adjustmentValueTensor[i] >= 100 for each image in batch)
+ * \param [in] roiTensorSrc ROI data in HOST memory, for each image in source tensor (2D tensor of size batchSize * 4, in either format - XYWH(xy.x, xy.y, roiWidth, roiHeight) or LTRB(lt.x, lt.y, rb.x, rb.y))
+ * \param [in] roiType ROI type used (RpptRoiType::XYWH or RpptRoiType::LTRB)
+ * \param [in] rppHandle RPP HOST handle created with <tt>\ref rppCreateWithBatchSize()</tt>
+ * \return A <tt> \ref RppStatus</tt> enumeration.
+ * \retval RPP_SUCCESS Successful completion.
+ * \retval RPP_ERROR* Unsuccessful completion.
+ */
+RppStatus rppt_color_temperature_host(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t dstPtr, RpptDescPtr dstDescPtr, Rpp8s *adjustmentValueTensor, RpptROIPtr roiTensorPtrSrc, RpptRoiType roiType, rppHandle_t rppHandle);
+
+#ifdef GPU_SUPPORT
+/*! \brief Color Temperature augmentation on HIP backend for a NCHW/NHWC layout tensor
+ * \details The color temperature augmentation does a image temperature adjustment operation, taking a pixel adjustment value as argument for each image in a batch of RGB(3 channel) with an NHWC/NCHW tensor layout.<br>
+ * - srcPtr depth ranges - Rpp8u (0 to 255), Rpp16f (0 to 1), Rpp32f (0 to 1), Rpp8s (-128 to 127).
+ * - dstPtr depth ranges - Will be same depth as srcPtr.
+ * \image html img150x150.jpg Sample Input
+ * \image html color_augmentations_color_temperature_img150x150.jpg Sample Output
+ * \param [in] srcPtr source tensor in HIP memory
+ * \param [in] srcDescPtr source tensor descriptor (Restrictions - numDims = 4, offsetInBytes >= 0, dataType = U8/F16/F32/I8, layout = NCHW/NHWC, c = 1/3)
+ * \param [out] dstPtr destination tensor in HIP memory
+ * \param [in] dstDescPtr destination tensor descriptor (Restrictions - numDims = 4, offsetInBytes >= 0, dataType = U8/F16/F32/I8, layout = NCHW/NHWC, c = same as that of srcDescPtr)
+ * \param [in] adjustmentValueTensor adjustment values for color temperature calculation (1D tensor of size sizeof(Rpp8s) * batchSize with -100 <= adjustmentValueTensor[i] >= 100 for each image in batch)
+ * \param [in] roiTensorSrc ROI data in HIP memory, for each image in source tensor (2D tensor of size batchSize * 4, in either format - XYWH(xy.x, xy.y, roiWidth, roiHeight) or LTRB(lt.x, lt.y, rb.x, rb.y))
+ * \param [in] roiType ROI type used (RpptRoiType::XYWH or RpptRoiType::LTRB)
+ * \param [in] rppHandle RPP HIP handle created with <tt>\ref rppCreateWithStreamAndBatchSize()</tt>
+ * \return A <tt> \ref RppStatus</tt> enumeration.
+ * \retval RPP_SUCCESS Successful completion.
+ * \retval RPP_ERROR* Unsuccessful completion.
+ */
+RppStatus rppt_color_temperature_gpu(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t dstPtr, RpptDescPtr dstDescPtr, Rpp32s *adjustmentValueTensor, RpptROIPtr roiTensorPtrSrc, RpptRoiType roiType, rppHandle_t rppHandle);
+#endif // GPU_SUPPORT
+
 /*! @}
  */
 

include/rppt_tensor_statistical_operations.h

@@ -24,6 +24,7 @@ SOFTWARE.
 
 #ifndef RPPT_TENSOR_STATISTICAL_OPERATIONS_H
 #define RPPT_TENSOR_STATISTICAL_OPERATIONS_H
+
 #include "rpp.h"
 #include "rppdefs.h"
 #ifdef __cplusplus
@@ -77,6 +78,78 @@ RppStatus rppt_tensor_sum_host(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t
 RppStatus rppt_tensor_sum_gpu(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t tensorSumArr, Rpp32u tensorSumArrLength, RpptROIPtr roiTensorPtrSrc, RpptRoiType roiType, rppHandle_t rppHandle);
 #endif // GPU_SUPPORT
 
+/*! \brief Tensor min operation on HOST backend for a NCHW/NHWC layout tensor
+ * \details The tensor min is a reduction operation that finds the channel-wise (R min / G min / B min) and overall min for each image in a batch of RGB(3 channel) / greyscale(1 channel) images with an NHWC/NCHW tensor layout.<br>
+ * - srcPtr depth ranges - Rpp8u (0 to 255), Rpp16f (0 to 1), Rpp32f (0 to 1), Rpp8s (-128 to 127).
+ * - dstPtr depth ranges - Will be same depth as srcPtr.
+ * \param [in] srcPtr source tensor in HOST memory
+ * \param [in] srcDescPtr source tensor descriptor (Restrictions - numDims = 4, offsetInBytes >= 0, dataType = U8/F16/F32/I8, layout = NCHW/NHWC, c = 1/3)
+ * \param [out] minArr destination array in HOST memory
+ * \param [in] minArrLength length of provided destination array (Restrictions - if srcDescPtr->c == 1 then tensorSumArrLength >= srcDescPtr->n, and if srcDescPtr->c == 3 then tensorSumArrLength >= srcDescPtr->n * 4)
+ * \param [in] roiTensorSrc ROI data in HOST memory, for each image in source tensor (2D tensor of size batchSize * 4, in either format - XYWH(xy.x, xy.y, roiWidth, roiHeight) or LTRB(lt.x, lt.y, rb.x, rb.y)) | (Restrictions - roiTensorSrc[i].xywhROI.roiWidth <= 3840 and roiTensorSrc[i].xywhROI.roiHeight <= 2160)
+ * \param [in] roiType ROI type used (RpptRoiType::XYWH or RpptRoiType::LTRB)
+ * \param [in] rppHandle RPP HOST handle created with <tt>\ref rppCreateWithBatchSize()</tt>
+ * \return A <tt> \ref RppStatus</tt> enumeration.
+ * \retval RPP_SUCCESS Successful completion.
+ * \retval RPP_ERROR* Unsuccessful completion.
+ */
+RppStatus rppt_tensor_min_host(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t minArr, Rpp32u minArrLength, RpptROIPtr roiTensorPtrSrc, RpptRoiType roiType, rppHandle_t rppHandle);
+
+#ifdef GPU_SUPPORT
+/*! \brief Tensor min operation on HIP backend for a NCHW/NHWC layout tensor
+ * \details The tensor min is a reduction operation that finds the channel-wise (R min / G min / B min) and overall min for each image in a batch of RGB(3 channel) / greyscale(1 channel) images with an NHWC/NCHW tensor layout.<br>
+ * - srcPtr depth ranges - Rpp8u (0 to 255), Rpp16f (0 to 1), Rpp32f (0 to 1), Rpp8s (-128 to 127).
+ * - dstPtr depth ranges - Will be same depth as srcPtr.
+ * \param [in] srcPtr source tensor in HIP memory
+ * \param [in] srcDescPtr source tensor descriptor (Restrictions - numDims = 4, offsetInBytes >= 0, dataType = U8/F16/F32/I8, layout = NCHW/NHWC, c = 1/3)
+ * \param [out] minArr destination array in HIP memory
+ * \param [in] minArrLength length of provided destination array (Restrictions - if srcDescPtr->c == 1 then tensorSumArrLength >= srcDescPtr->n, and if srcDescPtr->c == 3 then tensorSumArrLength >= srcDescPtr->n * 4)
+ * \param [in] roiTensorSrc ROI data in HIP memory, for each image in source tensor (2D tensor of size batchSize * 4, in either format - XYWH(xy.x, xy.y, roiWidth, roiHeight) or LTRB(lt.x, lt.y, rb.x, rb.y)) | (Restrictions - roiTensorSrc[i].xywhROI.roiWidth <= 3840 and roiTensorSrc[i].xywhROI.roiHeight <= 2160)
+ * \param [in] roiType ROI type used (RpptRoiType::XYWH or RpptRoiType::LTRB)
+ * \param [in] rppHandle RPP HIP handle created with <tt>\ref rppCreateWithStreamAndBatchSize()</tt>
+ * \return A <tt> \ref RppStatus</tt> enumeration.
+ * \retval RPP_SUCCESS Successful completion.
+ * \retval RPP_ERROR* Unsuccessful completion.
+ */
+RppStatus rppt_tensor_min_gpu(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t imageMinArr, Rpp32u imageMinArrLength, RpptROIPtr roiTensorPtrSrc, RpptRoiType roiType, rppHandle_t rppHandle);
+#endif // GPU_SUPPORT
+
+/*! \brief Tensor max operation on HOST backend for a NCHW/NHWC layout tensor
+ * \details The tensor max is a reduction operation that finds the channel-wise (R max / G max / B max) and overall max for each image in a batch of RGB(3 channel) / greyscale(1 channel) images with an NHWC/NCHW tensor layout.<br>
+ * - srcPtr depth ranges - Rpp8u (0 to 255), Rpp16f (0 to 1), Rpp32f (0 to 1), Rpp8s (-128 to 127).
+ * - dstPtr depth ranges - Will be same depth as srcPtr.
+ * \param [in] srcPtr source tensor in HOST memory
+ * \param [in] srcDescPtr source tensor descriptor (Restrictions - numDims = 4, offsetInBytes >= 0, dataType = U8/F16/F32/I8, layout = NCHW/NHWC, c = 1/3)
+ * \param [out] maxArr destination array in HOST memory
+ * \param [in] maxArrLength length of provided destination array (Restrictions - if srcDescPtr->c == 1 then tensorSumArrLength >= srcDescPtr->n, and if srcDescPtr->c == 3 then tensorSumArrLength >= srcDescPtr->n * 4)
+ * \param [in] roiTensorSrc ROI data in HOST memory, for each image in source tensor (2D tensor of size batchSize * 4, in either format - XYWH(xy.x, xy.y, roiWidth, roiHeight) or LTRB(lt.x, lt.y, rb.x, rb.y)) | (Restrictions - roiTensorSrc[i].xywhROI.roiWidth <= 3840 and roiTensorSrc[i].xywhROI.roiHeight <= 2160)
+ * \param [in] roiType ROI type used (RpptRoiType::XYWH or RpptRoiType::LTRB)
+ * \param [in] rppHandle RPP HOST handle created with <tt>\ref rppCreateWithBatchSize()</tt>
+ * \return A <tt> \ref RppStatus</tt> enumeration.
+ * \retval RPP_SUCCESS Successful completion.
+ * \retval RPP_ERROR* Unsuccessful completion.
+ */
+RppStatus rppt_tensor_max_host(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t maxArr, Rpp32u maxArrLength, RpptROIPtr roiTensorPtrSrc, RpptRoiType roiType, rppHandle_t rppHandle);
+
+#ifdef GPU_SUPPORT
+/*! \brief Tensor max operation on HIP backend for a NCHW/NHWC layout tensor
+ * \details The tensor max is a reduction operation that finds the channel-wise (R max / G max / B max) and overall max for each image in a batch of RGB(3 channel) / greyscale(1 channel) images with an NHWC/NCHW tensor layout.<br>
+ * - srcPtr depth ranges - Rpp8u (0 to 255), Rpp16f (0 to 1), Rpp32f (0 to 1), Rpp8s (-128 to 127).
+ * - dstPtr depth ranges - Will be same depth as srcPtr.
+ * \param [in] srcPtr source tensor in HIP memory
+ * \param [in] srcDescPtr source tensor descriptor (Restrictions - numDims = 4, offsetInBytes >= 0, dataType = U8/F16/F32/I8, layout = NCHW/NHWC, c = 1/3)
+ * \param [out] maxArr destination array in HIP memory
+ * \param [in] maxArrLength length of provided destination array (Restrictions - if srcDescPtr->c == 1 then tensorSumArrLength >= srcDescPtr->n, and if srcDescPtr->c == 3 then tensorSumArrLength >= srcDescPtr->n * 4)
+ * \param [in] roiTensorSrc ROI data in HIP memory, for each image in source tensor (2D tensor of size batchSize * 4, in either format - XYWH(xy.x, xy.y, roiWidth, roiHeight) or LTRB(lt.x, lt.y, rb.x, rb.y)) | (Restrictions - roiTensorSrc[i].xywhROI.roiWidth <= 3840 and roiTensorSrc[i].xywhROI.roiHeight <= 2160)
+ * \param [in] roiType ROI type used (RpptRoiType::XYWH or RpptRoiType::LTRB)
+ * \param [in] rppHandle RPP HIP handle created with <tt>\ref rppCreateWithBatchSize()</tt>
+ * \return A <tt> \ref RppStatus</tt> enumeration.
+ * \retval RPP_SUCCESS Successful completion.
+ * \retval RPP_ERROR* Unsuccessful completion.
+ */
+RppStatus rppt_tensor_max_gpu(RppPtr_t srcPtr, RpptDescPtr srcDescPtr, RppPtr_t imageMaxArr, Rpp32u imageMaxArrLength, RpptROIPtr roiTensorPtrSrc, RpptRoiType roiType, rppHandle_t rppHandle);
+#endif // GPU_SUPPORT
+
 /*! @}
  */