Develop - Merge to TOT (#170)

* Documents - Bump rocm-docs-core from 0.13.4 to 0.14.0 in /docs/.sphinx (#162)

Bumps [rocm-docs-core](https://github.com/RadeonOpenCompute/rocm-docs-core) from 0.13.4 to 0.14.0.
- [Release notes](https://github.com/RadeonOpenCompute/rocm-docs-core/releases)
- [Changelog](https://github.com/RadeonOpenCompute/rocm-docs-core/blob/develop/CHANGELOG.md)
- [Commits](ROCm/rocm-docs-core@v0.13.4...v0.14.0)

---
updated-dependencies:
- dependency-name: rocm-docs-core
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

* Bump rocm-docs-core from 0.13.4 to 0.16.0 in /docs/.sphinx (#164)

Bumps [rocm-docs-core](https://github.com/RadeonOpenCompute/rocm-docs-core) from 0.13.4 to 0.16.0.
- [Release notes](https://github.com/RadeonOpenCompute/rocm-docs-core/releases)
- [Changelog](https://github.com/RadeonOpenCompute/rocm-docs-core/blob/develop/CHANGELOG.md)
- [Commits](ROCm/rocm-docs-core@v0.13.4...v0.16.0)

---
updated-dependencies:
- dependency-name: rocm-docs-core
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Kiriti Gowda <kiritigowda@gmail.com>

* RPP - Test suite upgrade 2 (#161)

* Add ci

* Add initial rotate HIP implementation

* Minor change

* Add test suite case  for Rotate

* Minor test suite change

* Add initial rotate HOST implementation

* Add test suite cases

* Minor fix

* Minor changes

* Use combined sincos API

* Remove rotate.hpp

Compute affine matric and internaly invoke warp affine for HOST

* Remove rotate.hpp include

* Remove ci

* HIP Test suite changes

Use pinned memory for roi pointers and image sizes
Remove the HOST mem allocation and memcopy for the 2 variables

* Minor fix HOST Test suite

Release memory for image sizes buffers

* Change name of ROI variable

* Add changes to Non linear blend test cases in HIP test suite

* Minor fix - Rotate affine matrix

* initial working commit for merged test suite

* added helper function for converting PLN3 to PKD3 in test suite

* fixed the issue with CMN outputs

* minor changes

* Add new HIP merged test suite

* Test suite merge for HIP backend

* Add function to check the correctness of unit-tests output

* add a function to check the correctness of HIP unit-test output

* removed the dependency of hardcoded path in test suite

* fixed the issue with noise outputs from HOST test suite

minor cleanup in shell script files for unittests and performance tests

* fix the output difference between Avx and c outputs

* Add function to convert PKD3 input to PLN3

* fix errors in common test suite

* Add three augmentation refrence pixel values for unit tests

* minor changes in PKD3 to PLN3 conversion function

removed unnecessary variables in helpers in test suite

* Add rotate test cases in new test suite

* Minor changes

* fixed issue with HIP backend test suite

changed few variable names to camelCase

minor code cleanup

* minor change in compare_output function

* added a Map for finding augmentation name from case number

* minor changes

* removed the usage of device i8, f32,f16 buffers

* Add goldens outputs for some augmentations

* avoided the usage of tempirary buffer for HIP test suite

* Made changes in HOST teat suite similar to HIP test suite

removed unnecessary buffer usage in HOST test suite

* minor changes

* made changes to pass list of cases for test suite

* fixed the issue with folder dump for case list

* added support to generate kernel times for HIP test suite

* minor change

* fix seg faultin compare output function

* minor change in compare_output function

* restructure compare output function and code cleanup

* add common function for opencv dump

* Add common code for HOST and HIP in a common file

* Add common functions for similar code in HOST and HIP backends

* changes in write_output function

* made changes in timer variable names

converted measured time in ms

* Add readme file for new test suite

* combined set_strides() and set_nchw_values() functions into single function

* reverted changes happended with rotate merge

* readded spatter vector code

renamed reference output file names

* made changes to common ROI updation in test suite

added function to read images from imread

* updated readme

* Fixed the issue with bulleting in readme

* fix error in runTest file

* fix errors in runTest.py in Hip backend

* minor changes

* fix file odering issue

* change file ordering logic

* minor changes in parser and validator functions

* minor changes

* Bug fix for HIP Kernel time calculations

* added validation cases for PLN1

* fixed incorrect reference output for CMN PLN1

* Change ONE_OVER_255 precision

* change refFile path for interpolation functions

* Add new option to dump output array into csv file

* change golden outputs for color_twist

* Add Nearbyintf in Uchar store functions

* Add new Option QA_mode to compare the output with golden outputs

* Add new common function to set layout for descriptor pointers
Change logic in compare output function

* Remove function calls for augmentations with missing golden outputs

* change color_cast logic for pkd3-pln3 and pkd3-pkd3 conversions
change golden outputs of pkd3-pkd3 conversion of color_cast augmentation

* bug fix for CMN HIP

* minor changes

* Fix formatting in readme

* minor changes to parameter names in python

* Add TurboJpeg decoder to read Input data
Resolve exist_ok issue in python2
Change QA function logic

* minor code cleanup

* added modes of operation for test suite readme

* modifed case numbers in augmentation map

* removed nearbyintf() changes in rpp_hip_common.hpp

* changed max case (case end) in new test suite based on current functions supported

minor code cleanup

* added changes to print the summary of qa tests in a text file

* minor changes in README.md and runTests.py

* Minor changes w.r.t turbojpeg installation in test suite CMakeLists.txt

* added instructions for installing turbojpeg

* Codacy fixes

* Codacy fix

* Codacy fix

* Remove C-style pointer casting and codacy fixes

* updated instructions for installing turbojpeg

* removed reinterpret cast for u8 conversions

* Codacy fixes

* codacy fixes

* Remove C style type casting

* Add new option num_iterations

* made changes to display the total valid cases being run in QA mode

* minor formatting changes

* minor changes

* codacy fixes

* change TurboJpeg read from BGR to RGB
change reference outputs

* minor changes

* Turbojpeg find package mods

* funcCount to str

* Add new option preserve output
fix tee warnings
changes in readme

* fix tee warnings
fix PLN1 image dump issue
cleared mkdir warnings

* minor changes

* minor changes

* Add recursive file search
Add new option batch Size
ROI strides updation

* Flag for handling dual Input cases

* minor change

* Changes in batchSize implementation

* Minor Change

* Add Validation check for DST_FOLDER path

* Add Validation checks for all options in testAllScript.sh

* Add sanity check for dual Input cases
Set Max Dimension and Max Image Dump
Replaced Fast DCT tag with Accurate DCT

* Regenerate golden outputs using accurate dct Flag
Add golden outputs for some new augmentations

* Fix Flip golden outputs mismatch
Fix PLN3 variants mismatch in QA mode

* Add MAX_BATCH_SIZE check
removed Augmentations function calls for failing Qa modes
code cleanup

* Add crop and gamma correction augmentations
code cleanup

* Add comments to functions in rpp_test_suite_common.h

* minor change

* code cleanup

* minor code changes

* Change roi and Image sizes for crop augmentation

* Change numIterations option to numRuns
Addressed PR comments

* Add turboJpeg header to update maxHeight and maxWidth values

* Change the performance Timings logic

* minor change to support u8_f16 and u8_f32 cases

* Regenerate LUT golden outputs with ACCURATE_DCT tag

* Minor code changes

* Update testAllScript.sh in HIP - quotes to prevent globbing

* Update testAllScript.sh in HOST - quotes to prevent globbing

---------

Co-authored-by: fiona-gladwin <fionagladwin@multicorewareinc.com>
Co-authored-by: sampath1117 <sampath.rachumallu@multicorewareinc.com>
Co-authored-by: HazarathKumarM <hazarathkumar@multicorewareinc.com>
Co-authored-by: Snehaa Giridharan <snehaa@multicorewareinc.com>
Co-authored-by: sampath1117 <89194005+sampath1117@users.noreply.github.com>

* Jenkins - Update common.groovy

Support Dropped for U18

* OCL Backend - build fix (#171)

---------

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Abishek <52214183+r-abishek@users.noreply.github.com>
Co-authored-by: fiona-gladwin <fionagladwin@multicorewareinc.com>
Co-authored-by: sampath1117 <sampath.rachumallu@multicorewareinc.com>
Co-authored-by: HazarathKumarM <hazarathkumar@multicorewareinc.com>
Co-authored-by: Snehaa Giridharan <snehaa@multicorewareinc.com>
Co-authored-by: sampath1117 <89194005+sampath1117@users.noreply.github.com>

.jenkins/common.groovy

@@ -15,7 +15,7 @@ def runCompileCommand(platform, project, jobName, boolean debug=false, boolean s
             enableSCL = 'source scl_source enable llvm-toolset-7'
         }
     }
-    else if (platform.jenkinsLabel.contains('ubuntu18')) {
+    else if (platform.jenkinsLabel.contains('ubuntu20')) {
         backend = 'OCL'
     }
     else {

docs/.sphinx/requirements.in

@@ -1 +1,2 @@
 rocm-docs-core==0.18.2
+

src/modules/cl/handleocl.cpp

@@ -29,6 +29,7 @@
 #include <unistd.h>
 #endif
 
+#include <thread>
 #include "config.h"
 #include "rpp/device_name.hpp"
 #include "rpp/errors.hpp"
@@ -260,6 +261,7 @@ struct HandleImpl
     bool enable_profiling = false;
     float profiling_result = 0.0;
     size_t nBatchSize = 1;
+    Rpp32u numThreads = 0;
     InitHandle* initHandle = nullptr;
 
     ContextPtr create_context()
@@ -442,9 +444,13 @@ Handle::Handle(rppAcceleratorQueue_t stream) : impl(new HandleImpl())
     impl->PreInitializeBuffer();
 }
 
-Handle::Handle(size_t batchSize) : impl(new HandleImpl())
+Handle::Handle(size_t batchSize, Rpp32u numThreads) : impl(new HandleImpl())
 {
     impl->nBatchSize = batchSize;
+    numThreads = std::min(numThreads, std::thread::hardware_concurrency());
+    if(numThreads == 0)
+        numThreads = batchSize;
+    impl->numThreads = numThreads;
     this->SetAllocator(nullptr, nullptr, nullptr);
     impl->PreInitializeBufferCPU();
 }
@@ -517,6 +523,9 @@ Handle::Handle() : impl(new HandleImpl())
         RPP_THROW("Creating Command Queue. (clCreateCommandQueue)");
     }
     this->SetAllocator(nullptr, nullptr, nullptr);
+    impl->numThreads = std::min(impl->numThreads, std::thread::hardware_concurrency());
+    if(impl->numThreads == 0)
+        impl->numThreads = impl->nBatchSize;
     // RPP_LOG_I(*this);
 }
 
@@ -607,6 +616,11 @@ size_t Handle::GetBatchSize() const
     return this->impl->nBatchSize;
 }
 
+Rpp32u Handle::GetNumThreads() const
+{
+    return this->impl->numThreads;
+}
+
 void Handle::SetBatchSize(size_t bSize) const
 {
     this->impl->nBatchSize = bSize;

src/modules/hip/handlehip.cpp

@@ -29,7 +29,7 @@
 #include <unistd.h>
 #endif
 
-#include<thread>
+#include <thread>
 #include "config.h"
 #include "rpp/device_name.hpp"
 #include "rpp/errors.hpp"