Update README.md

* Delete AMD HIP Programming Guide_v4.2.pdf

* Delete AMD_HIP_API_Guide_4.2.pdf

* Delete AMD_ROCm_DataCenter_Tool_User_Guide_v4.2.pdf

* Delete AMD_ROCm_Release_Notes_v4.2.pdf

* Delete HIP_Supported_CUDA_API_Reference_Guide_v4.2.pdf

* Delete ROCm_Data_Center_Tool_API_Guide_v4.2.pdf

* Delete ROCm_Debugger_API_Guide_v4.2.pdf

* Delete ROCm_Debugger_User_Guide_v4.2.pdf

* Delete ROCm_SMI_Manual_4.2.pdf

* Update README.md

* Update README.md

* Delete CG1.PNG

* Delete CG2.PNG

* Delete CG3.PNG

* Delete CGMain.PNG

* Delete CLI1.PNG

* Delete CLI2.PNG

* Delete SMI.PNG

* Delete keyfeatures.PNG

* Delete latestGPU.PNG

* Delete rocsolverAPI.PNG

* Create test.rst

* Add files via upload

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

* Update README.md

README.md

@@ -1,25 +1,91 @@
 
-# AMD ROCm Release Notes v3.8.0
+# AMD ROCm™ v4.3.1 Point Release Notes 
+
+This document describes the features, fixed issues, and information about downloading and installing the AMD ROCm™ software.
 
-This page describes the features, fixed issues, and information about downloading and installing the ROCm software.
 It also covers known issues in this release.
 
+## List of Supported Operating Systems
+
+The AMD ROCm platform supports the following operating systems:
+
+![Screenshot](https://github.com/RadeonOpenCompute/ROCm/blob/master/images/SuppEnv.PNG)
+
+
+## What\'s New in This Release
+
+The ROCm v4.3.1 release consists of the following enhancements:
+
+### Support for RHEL V8.4
+
+This release extends support for RHEL v8.4.
+
+### Support for SLES V15 Service Pack 3
+
+This release extends support for SLES v15 SP3.
+
+### Pass Manager Update 
+
+In the AMD ROCm 4.3.1 release, the ROCm compiler uses the legacy pass manager, by default, to provide a better performance experience with some workloads. 
+
+Previously, in ROCm v4.3, the default choice for the ROCm compiler was the new pass manager. 
+
+For more information about legacy and new pass managers, see http://llvm.org.
+
+
+## Known Issues in This Release
+
+### General Userspace and Application Freeze on MI25
+
+For some workloads on MI25, general user space and application freeze are observed, and the GPU resets intermittently. Note, the freeze may take hours to reproduce.
+
+This issue is under active investigation, and no workarounds are available currently.
+
+### hipRTC - File Not Found Error
+
+hipRTC may fail, and users may encounter the following error:
+
+
+		<built-in>:1:10: fatal error: '__clang_hip_runtime_wrapper.h' file not found
+		#include "__clang_hip_runtime_wrapper.h"
+
+
+
+#### Suggested Workarounds
+
+* Set LLVM_PATH in the environment to <path to ROCm llvm>/llvm. Note, if ROCm is installed at the default location, then LLVM_PATH must be set to /opt/rocm/llvm. 
+	
+* Add “-I <path to ROCm>/llvm/lib/clang/13.0.0/include/” to compiler options in the call to hiprtcCompileProgram (). Note, this workaround requires the following changes in the code:
+
+	
+		// set NUM_OPTIONS to one more than the number of options that was previously required
+		const char* options[NUM_OPTIONS];
+		// fill other options[] here
+		std::string sarg = "-I/opt/rocm/llvm/lib/clang/13.0.0/include/";
+		options[NUM_OPTIONS - 1] = sarg.c_str();
+		hiprtcResult compileResult{hiprtcCompileProgram(prog, NUM_OPTIONS, options)};"
+
+
+
+
+# AMD ROCm™ v4.3 Release Notes 
+
+This document describes the features, fixed issues, and information about downloading and installing the AMD ROCm™ software. It also covers known issues and deprecations in this release.
+
 - [Supported Operating Systems and Documentation Updates](#Supported-Operating-Systems-and-Documentation-Updates)
   * [Supported Operating Systems](#Supported-Operating-Systems)
+  * [ROCm Installation Updates](#ROCm-Installation-Updates)
   * [AMD ROCm Documentation Updates](#AMD-ROCm-Documentation-Updates)
+
    
 - [What\'s New in This Release](#Whats-New-in-This-Release)
-  * [Hipfort-Interface for GPU Kernel Libraries](#Hipfort-Interface-for-GPU-Kernel-Libraries)
+  * [HIP Enhancements](#HIP-Enhancements)
   * [ROCm Data Center Tool](#ROCm-Data-Center-Tool)
-  * [Error-Correcting Code Fields in ROCm Data Center Tool](#Error-Correcting-Code-Fields-in-ROCm-Data-Center-Tool)
-  * [Static Linking Libraries](#Static-Linking-Libraries)
-    
-- [Fixed Defects](#Fixed-Defects)
+  * [ROCm Math and Communication Libraries](#ROCm-Math-and-Communication-Libraries)
+  * [ROCProfiler Enhancements](#ROCProfiler-Enhancements)   
 
-- [Known Issues](#Known-Issues)
+- [Known Issues in This Release](#Known-Issues-in-This-Release)
 
-- [Deploying ROCm](#Deploying-ROCm)
- 
 - [Hardware and Software Support](#Hardware-and-Software-Support)
 
 - [Machine Learning and High Performance Computing Software Stack for AMD GPU](#Machine-Learning-and-High-Performance-Computing-Software-Stack-for-AMD-GPU)
@@ -28,34 +94,53 @@ It also covers known issues in this release.
   
 
 
-# Supported Operating Systems 
 
-## Support for Vega 7nm Workstation
+## ROCm Installation Updates 
 
-This release extends support to the Vega 7nm Workstation (Vega20 GL-XE) version.
-
-## List of Supported Operating Systems
+### Supported Operating Systems
 
 The AMD ROCm platform is designed to support the following operating systems:
 
-* Ubuntu 20.04 (5.4 and 5.6-oem) and 18.04.5 (Kernel 5.4)	
-* CentOS 7.8 & RHEL 7.8 (Kernel 3.10.0-1127) (Using devtoolset-7 runtime support)
-* CentOS 8.2 & RHEL 8.2 (Kernel 4.18.0 ) (devtoolset is not required)
-* SLES 15 SP1
+![Screenshot](https://github.com/RadeonOpenCompute/ROCm/blob/master/images/OSKernelupdated.PNG)
 
-## Fresh Installation of AMD ROCm v3.8 Recommended
-A fresh and clean installation of AMD ROCm v3.8 is recommended. An upgrade from previous releases to AMD ROCm v3.8 is not supported.
 
-For more information, refer to the AMD ROCm Installation Guide at:
+### Fresh Installation of AMD ROCM V4.3 Recommended
+
+Complete uninstallation of previous ROCm versions is required before installing a new version of ROCm. **An upgrade from previous releases to AMD ROCm v4.3 is not supported**. For more information, refer to the AMD ROCm Installation Guide at
+
 https://rocmdocs.amd.com/en/latest/Installation_Guide/Installation-Guide.html
 
 **Note**: AMD ROCm release v3.3 or prior releases are not fully compatible with AMD ROCm v3.5 and higher versions. You must perform a fresh ROCm installation if you want to upgrade from AMD ROCm v3.3 or older to 3.5 or higher versions and vice-versa.
 
-**Note**: *render group* is required only for Ubuntu v20.04. For all other ROCm supported operating systems, continue to use *video group*.
+**Note**: *render* group is required only for Ubuntu v20.04. For all other ROCm supported operating systems, continue to use video group. 
 
-* For ROCm v3.5 and releases thereafter,the *clinfo* path is changed to -  */opt/rocm/opencl/bin/clinfo*.
+* For ROCm v3.5 and releases thereafter, the clinfo path is changed to /opt/rocm/opencl/bin/clinfo. 
 
-* For ROCm v3.3 and older releases, the *clinfo* path remains unchanged -  */opt/rocm/opencl/bin/x86_64/clinfo*.
+* For ROCm v3.3 and older releases, the clinfo path remains /opt/rocm/opencl/bin/x86_64/clinfo. 
+ 
+## ROCm Multi-Version Installation Update
+
+With the AMD ROCm v4.3 release, the following ROCm multi-version installation changes apply:
+
+The meta packages rocm-dkms<version> are now deprecated for multi-version ROCm installs.  For example, rocm-dkms3.7.0, rocm-dkms3.8.0.
+ 
+* Multi-version installation of ROCm should be performed by installing rocm-dev<version> using each of the desired ROCm versions. For example, rocm-dev3.7.0, rocm-dev3.8.0, rocm-dev3.9.0.   
+
+* The rock-dkms loadable kernel modules should be installed using a single rock-dkms package. 
+
+* ROCm v3.9 and above will not set any ldconfig entries for ROCm libraries for multi-version installation.  Users must set LD_LIBRARY_PATH to load the ROCm library version of choice.
+
+**NOTE**: The single version installation of the ROCm stack remains the same. The rocm-dkms package can be used for single version installs and is not deprecated at this time.
+	
+
+## Support for Enviornment Modules
+	
+Environment modules are now supported. This enhancement in the ROCm v4.3 release enables users to switch between ROCm v4.2 and ROCm v4.3 easily and efficiently. 
+	
+For more information about installing environment modules, refer to
+	
+https://modules.readthedocs.io/en/latest/
+ 
 
 
 # AMD ROCm Documentation Updates
@@ -64,27 +149,71 @@ https://rocmdocs.amd.com/en/latest/Installation_Guide/Installation-Guide.html
 
 The AMD ROCm Installation Guide in this release includes:
 
-* Updated Supported Environments
-* HIP Installation Instructions
-* Tensorflow ROCm Port: Basic Installations on RHEL v8.2
+* Supported Environments
+
+* Installation Instructions 
+
+* HIP Installation Instructions 
+
+For more information, refer to the ROCm documentation website at:
+
+https://rocmdocs.amd.com/en/latest/
 
-https://rocmdocs.amd.com/en/latest/Installation_Guide/Installation-Guide.html
 
 ## AMD ROCm - HIP Documentation Updates
 
-* HIP Repository Information
+* HIP Programming Guide v4.3 
+	
+https://github.com/RadeonOpenCompute/ROCm/blob/master/AMD_HIP_Programming_Guide_v4.3.pdf
+	
+* HIP API Guide v4.3
+	
+https://github.com/RadeonOpenCompute/ROCm/blob/master/AMD_HIP_API_Guide_v4.3.pdf
+	
+* HIP-Supported CUDA API Reference Guide v4.3
+	
+https://github.com/RadeonOpenCompute/ROCm/blob/master/AMD_HIP_Supported_CUDA_API_Reference_Guide_v4.3.pdf
+	
 
-For more information, see 
+* **NEW** - AMD ROCm Compiler Reference Guide v4.3 
+	
+https://github.com/RadeonOpenCompute/ROCm/blob/master/AMD_Compiler_Reference_Guide_v4.3.pdf
+	
+* HIP FAQ
+	
+https://rocmdocs.amd.com/en/latest/Programming_Guides/HIP-FAQ.html#hip-faq
 
-https://rocmdocs.amd.com/en/latest/Programming_Guides/Programming-Guides.html#hip-repository-information
 
-## ROCm Data Center Tool User Guide
+## ROCm Data Center User and API Guide
 
-* Error-Correction Codes Field and Output Documentation
+* ROCm Data Center Tool User Guide
 
-For more information, refer to the AMD ROCm Data Center User Guide at
+  - Prometheus (Grafana) Integration with Automatic Node Detection 
+	
+https://github.com/RadeonOpenCompute/ROCm/blob/master/AMD_ROCm_DataCenter_Tool_User_Guide_v4.3.pdf
 
-https://github.com/RadeonOpenCompute/ROCm/blob/master/AMD_ROCm_DataCenter_Tool_User_Guide.pdf
+* ROCm Data Center Tool API Guide
+
+https://github.com/RadeonOpenCompute/ROCm/blob/master/AMD_RDC_API_Guide_v4.3.pdf
+	
+
+## ROCm SMI API Documentation Updates 
+
+* ROCm SMI API Guide 
+
+https://github.com/RadeonOpenCompute/ROCm/blob/master/AMD_ROCm_SMI_Guide_v4.3.pdf
+ 
+  
+## ROC Debugger User and API Guide 
+
+* ROC Debugger User Guide  
+	
+https://github.com/RadeonOpenCompute/ROCm/blob/master/AMD_ROCDebugger_User_Guide.pdf
+
+* Debugger API Guide 
+
+https://github.com/RadeonOpenCompute/ROCm/blob/master/AMD_ROCDebugger_API.pdf 
+	
 
 ## General AMD ROCm Documentation Links
 
@@ -100,129 +229,557 @@ Access the following links for more information:
 
 * For AMD ROCm binary structure, see
 
-  https://rocmdocs.amd.com/en/latest/Installation_Guide/Installation-Guide.html#build-amd-rocm
+  https://rocmdocs.amd.com/en/latest/Installation_Guide/Software-Stack-for-AMD-GPU.html
+  
 
 * For AMD ROCm Release History, see
 
-  https://rocmdocs.amd.com/en/latest/Installation_Guide/Installation-Guide.html#amd-rocm-version-history
+ https://rocmdocs.amd.com/en/latest/Current_Release_Notes/ROCm-Version-History.html
 
 
 
 # What\'s New in This Release
 
-## Hipfort-Interface for GPU Kernel Libraries
+## HIP Enhancements
 
-Hipfort is an interface library for accessing GPU Kernels. It provides support to the AMD ROCm architecture from within the Fortran programming language. Currently, the gfortran and HIP-Clang compilers support hipfort. Note, the gfortran compiler belongs to the GNU Compiler Collection (GCC). While hipfc wrapper calls hipcc for the non-fortran kernel source, gfortran is used for FORTRAN applications that call GPU kernels.
+### HIP Versioning Update
 
-The hipfort interface library is meant for Fortran developers with a focus on gfortran users. 
+The HIP version definition is updated from the ROCm v4.2 release as follows: 
+
+```
+	HIP_VERSION=HIP_VERSION_MAJOR * 10000000 + HIP_VERSION_MINOR * 100000 + 
+	HIP_VERSION_PATCH)
+```
+
+The HIP version can be queried from a HIP API call
+
+
+```	
+	hipRuntimeGetVersion(&runtimeVersion);	
+```
+	
+**Note**: The version returned will be greater than the version in previous ROCm releases.
+
+	
+### Support for Managed Memory Allocation
+
+HIP now supports and automatically manages Heterogeneous Memory Management (HMM) allocation. The HIP application performs a capability check before making the managed memory API call hipMallocManaged.
+
+**Note**: The _managed_ keyword is unsupported currently. 
+
+```
+	int managed_memory = 0;
+	HIPCHECK(hipDeviceGetAttribute(&managed_memory,
+ 	 hipDeviceAttributeManagedMemory,p_gpuDevice));
+	if (!managed_memory ) {
+  	printf ("info: managed memory access not supported on the device %d\n Skipped\n", p_gpuDevice);
+	}
+	else {
+ 	 HIPCHECK(hipSetDevice(p_gpuDevice));
+  	HIPCHECK(hipMallocManaged(&Hmm, N * sizeof(T)));
+	. . .
+	}
+```
+
+### Kernel Enqueue Serialization
+
+Developers can control kernel command serialization from the host using the following environment variable,
+AMD_SERIALIZE_KERNEL
+	
+* AMD_SERIALIZE_KERNEL = 1, Wait for completion before enqueue,
+
+* AMD_SERIALIZE_KERNEL = 2, Wait for completion after enqueue,
+
+* AMD_SERIALIZE_KERNEL = 3, Both.
+
+This environment variable setting enables HIP runtime to wait for GPU idle before/after any GPU command.
+
+
+### NUMA-aware Host Memory Allocation
+	
+The Non-Uniform Memory Architecture (NUMA) policy determines how memory is allocated and selects a CPU closest to each GPU. 
+	
+NUMA also measures the distance between the GPU and CPU devices. By default, each GPU selects a Numa CPU node that has the least NUMA distance between them; the host memory is automatically allocated closest to the memory pool of the NUMA node of the current GPU device. 
+	
+Note, using the *hipSetDevice* API with a different GPU provides access to the host allocation. However, it may have a longer NUMA distance.
+
+
+### New Atomic System Scope Atomic Operations
+	
+HIP now provides new APIs with _system as a suffix to support system scope atomic operations. For example,  atomicAnd atomic is dedicated to the GPU device, and atomicAnd_system allows developers to extend the atomic operation to system scope from the GPU device to other CPUs and GPU devices in the system.
+	
+For more information, refer to the HIP Programming Guide at,
+	
+https://github.com/RadeonOpenCompute/ROCm/blob/master/AMD_HIP_Programming_Guide_v4.3.pdf
+
+### Indirect Function Call and C++ Virtual Functions 
+	
+While the new release of the ROCm compiler supports indirect function calls and C++ virtual functions on a device, there are some known limitations and issues. 
+	
+**Limitations**
+	
+* An address to a function is device specific.  Note, a function address taken on the host can not be used on a device, and a function address taken on a device can not be used on the host.  On a system with multiple devices, an address taken on one device can not be used on a different device.
+	
+* C++ virtual functions only work on the device where the object was constructed.
+	
+* Indirect call to a device function with function scope shared memory allocation is not supported. For example, LDS.
+	
+* Indirect call to a device function defined in a source file different than the calling function/kernel is only supported when compiling the entire program with -fgpu-rdc.
+	
+**Known Issues in This Release**
+	
+* Programs containing kernels with different launch bounds may crash when making an indirect function call.  This issue is due to a compiler issue miscalculating the register budget for the callee function.
+	
+* Programs may not work correctly when making an indirect call to a function that uses more resources. For example, scratch memory, shared memory, registers made available by the caller.
+	
+* Compiling a program with objects with pure or deleted virtual functions on the device will result in a linker error.  This issue is due to the missing implementation of some C++ runtime functions on the device.
+	
+* Constructing an object with virtual functions in private or shared memory may crash the program due to a compiler issue when generating code for the constructor.  
 
-For information on HIPFort installation and examples, see
-https://github.com/ROCmSoftwarePlatform/hipfort
-  
 
 ## ROCm Data Center Tool 
 
-The ROCm™ Data Center Tool™ simplifies the administration and addresses key infrastructure challenges in AMD GPUs in cluster and datacenter environments. The important features of this tool are:
+### Prometheus (Grafana) Integration with Automatic Node Detection 
 
-* GPU telemetry  
+The ROCm Data Center (RDC) tool enables you to use Consul to discover the rdc_prometheus service automatically. Consul is “a service mesh solution providing a full-featured control plane with service discovery, configuration, and segmentation functionality.” For more information, refer to their website at https://www.consul.io/docs/intro.
+	
+The ROCm Data Center Tool uses Consul for health checks of RDC’s integration with the Prometheus plug-in (rdc_prometheus), and these checks provide information on its efficiency.  
+	
+Previously, when a new compute node was added, users had to change prometheus_targets.json to use Consul manually. Now, with the Consul agent integration, a new compute node can be discovered automatically.
+	
+https://github.com/RadeonOpenCompute/ROCm/blob/master/AMD_ROCm_DataCenter_Tool_User_Guide_v4.3.pdf
 
-* GPU statistics for jobs 
+### Coarse Grain Utilization
+	
+This feature provides a counter that displays the coarse grain GPU usage information, as shown below.
+	
+Sample output
 
-* Integration with third-party tools 
+```
+		$ rocm_smi.py --showuse
+		============================== % time GPU is busy =============================
+               GPU[0] : GPU use (%): 0
+               GPU[0] : GFX Activity: 3401
+```
 
-* Open source 
+	
+### Add 64-bit Energy Accumulator In-band
+	
+This feature provides an average value of energy consumed over time in a free-flowing RAPL counter, a 64-bit Energy Accumulator.
+	
+Sample output
+	
+```
+	$ rocm_smi.py --showenergycounter
+	=============================== Consumed Energy ================================
+	GPU[0] : Energy counter: 2424868
+	GPU[0] : Accumulated Energy (uJ): 0.0	
 
-The ROCm Data Center Tool can be used in the standalone mode if all components are installed. The same set of features is also available in a library format that can be used by existing management tools. 
+```	
+	
+### Support for Continuous Clocks Values
+	
+ROCm SMI will support continuous clock values instead of the previous discrete levels. Moving forward the updated sysfs file will consist of only MIN and MAX values and the user can set the clock value in the given range. 
+	
+Sample output:
 
-![ScreenShot](https://github.com/Rmalavally/ROCm/blob/master/RDCComponentsrevised.png)
+```
+	$ rocm_smi.py --setsrange 551 1270 
+	Do you accept these terms? [y/N] y                                                                                    
+	============================= Set Valid sclk Range=======
+	GPU[0]          : Successfully set sclk from 551(MHz) to 1270(MHz)                                                     
+	GPU[1]          : Successfully set sclk from 551(MHz) to 1270(MHz)                                                     
+	=========================================================================
+                       
+	$ rocm_smi.py --showsclkrange                                                                                                                                                                    
+	============================ Show Valid sclk Range======                     
+
+	GPU[0]          : Valid sclk range: 551Mhz - 1270Mhz                                                                  
+	GPU[1]          : Valid sclk range: 551Mhz - 1270Mhz             
+```
+	
+### Memory Utilization Counters
+
+This feature provides a counter display memory utilization information as shown below.
+
+Sample output
+	
+```
+       $ rocm_smi.py --showmemuse
+	========================== Current Memory Use ==============================
+
+	GPU[0] : GPU memory use (%): 0
+	GPU[0] : Memory Activity: 0
+```	
+
+### Performance Determinism
+
+ROCm SMI supports performance determinism as a unique mode of operation. Performance variations are minimal as this enhancement allows users to control the entry and exit to set a soft maximum (ceiling) for the GFX clock.
+	
+Sample output
+
+```
+	$ rocm_smi.py --setperfdeterminism 650
+	cat pp_od_clk_voltage
+	GFXCLK:                
+	0: 500Mhz
+	1: 650Mhz *
+	2: 1200Mhz
+	$ rocm_smi.py --resetperfdeterminism 	
+```	
+	
+**Note**: The idle clock will not take up higher clock values if no workload is running. After enabling determinism, users can run a GFX workload to set performance determinism to the desired clock value in the valid range.
+
+	* GFX clock could either be less than or equal to the max value set in this mode. GFX clock will be at the max clock set in this mode only when required by the running 	workload.
+	
+	* VDDGFX will be higher by an offset (75mv or so based on PPTable) in the determinism mode.
+	
+### HBM Temperature Metric Per Stack
+
+This feature will enable ROCm SMI to report all HBM temperature values as shown below.
+
+Sample output
+
+```	
+   	$ rocm_smi.py –showtemp
+	================================= Temperature =================================
+	GPU[0] : Temperature (Sensor edge) (C): 29.0
+	GPU[0] : Temperature (Sensor junction) (C): 36.0
+	GPU[0] : Temperature (Sensor memory) (C): 45.0
+	GPU[0] : Temperature (Sensor HBM 0) (C): 43.0
+	GPU[0] : Temperature (Sensor HBM 1) (C): 42.0
+	GPU[0] : Temperature (Sensor HBM 2) (C): 44.0
+	GPU[0] : Temperature (Sensor HBM 3) (C): 45.0
+```	
+
+	
+## ROCm Math and Communication Libraries 
+
+### rocBLAS
+
+**Optimizations**
+
+* Improved performance of non-batched and batched rocblas_Xgemv for gfx908 when m <= 15000 and n <= 15000
+	
+* Improved performance of non-batched and batched rocblas_sgemv and rocblas_dgemv for gfx906 when m <= 6000 and n <= 6000
+	
+* Improved the overall performance of non-batched and batched rocblas_cgemv for gfx906
+	
+* Improved the overall performance of rocblas_Xtrsv
+
+For more information, refer to 
+
+https://rocblas.readthedocs.io/en/master/
 
 
-Refer to the ROCm Data Center Tool™ User Guide for more details on the different modes of operation.  
+### rocRAND
 
-NOTE: The ROCm Data Center User Guide is intended to provide an overview of ROCm Data Center Tool features and how system administrators and Data Center (or HPC) users can administer and configure AMD GPUs. The guide also provides an overview of its components and open source developer handbook. 
+**Enhancements**
+	
+* gfx90a support added
+	
+* gfx1030 support added
 
-For installation information on different distributions, refer to the ROCm Data Center User Guide at
+* gfx803 supported re-enabled
 
-https://github.com/RadeonOpenCompute/ROCm/blob/master/AMD_ROCm_DataCenter_Tool_User_Guide.pdf
+**Fixed**
+	
+* Memory leaks in Poisson tests has been fixed.
+	
+* Memory leaks when generator has been created but setting seed/offset/dimensions display an exception has been fixed.
+
+For more information, refer to
+
+https://rocrand.readthedocs.io/en/latest/
 
 
-### Error Correcting Code Fields in ROCm Data Center Tool
+### rocSOLVER	
 
-The ROCm Data Center (RDC) tool is enhanced to provide counters to track correctable and uncorrectable errors. While a single bit per word error can be corrected, double bit per word errors cannot be corrected. 
+**Enhancements**
+	
+Linear solvers for general non-square systems:
+	
+* GELS now supports underdetermined and transposed cases
+	
+* Inverse of triangular matrices
+	
+* TRTRI (with batched and strided_batched versions)
+	
+* Out-of-place general matrix inversion
+	
+* GETRI_OUTOFPLACE (with batched and strided_batched versions)
+	
+* Argument names for the benchmark client now match argument names from the public API
+	
+**Fixed Issues**
+	
+* Known issues with Thin-SVD. The problem was identified in the test specification, not in the thin-SVD implementation or the rocBLAS gemm_batched routines.
 
-The RDC tool now helps monitor and protect undetected memory data corruption. If the system is using ECC- enabled memory, the ROCm Data Center tool can report the error counters to monitor the status of the memory. 
+* Benchmark client longer crashes as a result of leading dimension or stride arguments not being provided on the command line.
 
-![ScreenShot](https://github.com/Rmalavally/ROCm/blob/master/forweb.PNG)
+**Optimizations**
+	
+* Improved general performance of matrix inversion (GETRI)
 
-## Static Linking Libraries 
+For more information, refer to
 
-The underlying libraries of AMD ROCm are dynamic and are called shared objects (.so) in Linux.
-The AMD ROCm v3.8 release includes the capability to build static ROCm libraries and link to the applications statically. CMake target files enable linking an application statically to ROCm libraries and each component exports the required dependencies for linking. The static libraries are called Archives (.a) in Linux.
+https://rocsolver.readthedocs.io/en/latest/
 
-This release also comprises of the requisite changes required for all the components to work in a static environment. The components have been successfully tested for basic functionalities like *rocminfo /rocm_bandwidth_test* and archives.
 
-In the AMD ROCm v3.8 release, the following libraries support static linking: 
+### rocSPARSE	
+	
+**Enhancements**
+	
+* (batched) tridiagonal solver with and without pivoting
+	
+* dense matrix sparse vector multiplication (gemvi)
+	
+* support for gfx90a
+	
+* sampled dense-dense matrix multiplication (sddmm)
+	
+**Improvements**
+	
+* client matrix download mechanism
+	
+* boost dependency in clients removed
 
-![ScreenShot](https://github.com/Rmalavally/ROCm/blob/master/staticlinkinglib.PNG)
 
-# Fixed Defects
-The following defects are fixed in this release:
+For more information, refer to
 
-* GPU Kernel C++ Names Not Demangled
-* MIGraphX Fails for fp16 Datatype
-* Issue with Peer-to-Peer Transfers
-* ‘rocprof’ option ‘--parallel-kernels’ Not Supported in this Release
+https://rocsparse.readthedocs.io/en/latest/usermanual.html#rocsparse-gebsrmv
 
-# Known Issues 
 
-## Undefined Reference Issue in Statically Linked Libraries
+### hipBLAS
 
-Libraries and applications statically linked using flags -rtlib=compiler-rt, such as rocBLAS, have an implicit dependency on gcc_s not captured in their CMAKE configuration.  
+**Enhancements**
+	
+* Added *hipblasStatusToString*
+	
+**Fixed**
+	
+* Added catch() blocks around API calls to prevent the leak of C++ exceptions
+	
 
-Client applications may require linking with an additional library -lgcc_s to resolve the undefined reference to symbol '_Unwind_Resume@@GCC_3.0'.
+### rocFFT
 
-## MIGraphX Pooling Operation Fails for Some Models
+**Changes**
+	
+* Re-split device code into single-precision, double-precision, and miscellaneous kernels.
+	
+**Fixed Issues**
+	
+* double-precision planar->planar transpose.
+	
+* 3D transforms with unusual strides, for SBCC-optimized sizes.
+	
+* Improved buffer placement logic.
 
-MIGraphX does not work for some models with pooling operations and the following error appears:
+For more information, refer to
 
-*‘test_gpu_ops_test FAILED’*
+https://rocfft.readthedocs.io/en/rocm-4.3.0/
+	
 
-This issue is currently under investigation and there is no known workaround currently. 
+### hipFFT	
 
-## MIVisionX Installation Error on CentOS/RHEL8.2 and SLES 15
+**Fixed Issues**
+	
+* CMAKE updates
+	
+* Added callback API in hipfftXt.h header.
 
-Installing ROCm on MIVisionX results in the following error on CentOS/RHEL8.2 and SLES 15:
 
-*"Problem: nothing provides opencv needed"*
+### rocALUTION
+	
+**Enhancements**
+	
+* Support for gfx90a target
+	
+* Support for gfx1030 target
+	
+**Improvements**
+	
+* Install script
+	
+For more information, refer to
+	
+### rocTHRUST	
 
-As a workaround, install opencv before installing MIVisionX.
+**Enhancements**
+	
+* Updated to match upstream Thrust 1.11
+	
+* gfx90a support added
+	
+* gfx803 support re-enabled
+
+hipCUB	
+
+Enhancements
+
+* DiscardOutputIterator to backend header
+	
+
+## ROCProfiler Enhancements
+
+### Tracing Multiple MPI Ranks 
+
+When tracing multiple MPI ranks in ROCm v4.3, users must use the form:
+
+```
+	mpirun ... <mpi args> ... rocprof ... <rocprof args> ... application ... <application args>
+	
+```
+
+**NOTE**: This feature differs from ROCm v4.2 (and lower), which used "rocprof ... mpirun ... application".
+	
+This change was made to enable ROCProfiler to handle process forking better and launching via mpirun (and related) executables. 
+
+From a user perspective, this new execution mode requires:
+	
+1. Generation of trace data per MPI (or process) rank.
+	
+2. Use of a new ["merge_traces.sh" utility script](https://github.com/ROCm-Developer-Tools/rocprofiler/blob/rocm-4.3.x/bin/merge_traces.sh) to combine traces from multiple processes into a unified trace for profiling.
+	
+For example, to accomplish step #1, ROCm provides a simple bash wrapper that demonstrates how to generate a unique output directory per process:
+	
+```
+	$ cat wrapper.sh
+	#! /usr/bin/env bash
+	if [[ -n ${OMPI_COMM_WORLD_RANK+z} ]]; then
+ 	  # mpich
+  	  export MPI_RANK=${OMPI_COMM_WORLD_RANK}
+	elif [[ -n ${MV2_COMM_WORLD_RANK+z} ]]; then
+  	  # ompi
+  	  export MPI_RANK=${MV2_COMM_WORLD_RANK}
+	fi
+	args="$*"
+	pid="$$"
+	outdir="rank_${pid}_${MPI_RANK}"
+	outfile="results_${pid}_${MPI_RANK}.csv"
+	eval "rocprof -d ${outdir} -o ${outdir}/${outfile} $*"
+```
+
+This script:
+	
+* Determines the global MPI rank (implemented here for OpenMPI and MPICH only)
+	
+* Determines the process id of the MPI rank
+	
+* Generates a unique output directory using the two
+
+To invoke this wrapper, use the following command:
+
+```
+	mpirun <mpi args> ./wrapper.sh --hip-trace <application> <args>
+```
+
+This generates an output directory for each used MPI rank. For example, 	
+
+```
+	$ ls -ld rank_* | awk {'print $5" "$9'}
+	4096 rank_513555_0
+	4096 rank_513556_1
+```	
+	
+Finally, these traces may be combined using the [merge traces script](https://github.com/ROCm-Developer-Tools/rocprofiler/blob/rocm-4.3.x/bin/merge_traces.sh). For example, 
+
+```
+	$  ./merge_traces.sh -h
+	Script for aggregating results from multiple rocprofiler out directries.
+	Full path: /opt/rocm/bin/merge_traces.sh
+	Usage:
+  	merge_traces.sh -o <outputdir> [<inputdir>...]
+```	
+
+Use the following input arguments to the merge_traces.sh script to control which traces are merged and where the resulting merged trace is saved.
+	
+* -o <*outputdir*> - output directory where the results are aggregated.
+	
+* <*inputdir*>... - space-separated list of rocprofiler directories. If not specified, CWD is used.
+
+For example, if an output directory named "unified" was supplied to the `merge_traces.sh` script, the file 'unified/results.json' will be generated, and the contains trace data from both MPI ranks. 
+
+Known issue for ROCProfiler
+	
+Collecting several counter collection passes (multiple "pmc:" lines in an counter input file) is not supported in a single run.  
+The workaround is to break the multiline counter input file into multiple single-line counter input files and execute runs.
+
+	
+# Known Issues in This Release
+
+The following are the known issues in this release.
+
+## Upgrade to AMD ROCm v4.3 Not Supported
+
+An upgrade from previous releases to AMD ROCm v4.2 is not supported. Complete uninstallation of previous ROCm versions is required before installing a new version of ROCm.
+	
+## _LAUNCH BOUNDS_Ignored During Kernel Launch
+
+The HIP runtime returns the hipErrorLaunchFailure error code when an application tries to launch kernel with a block size larger than the launch bounds mentioned during compile time. If no launch bounds were specified during the compile time, the default value of 1024 is assumed.  Refer to the HIP trace for more information about the failing kernel. A sample error in the trace is shown below:
+ 
+Snippet of the HIP trace
+	
+```
+	:3:devprogram.cpp           :2504: 2227377746776 us: Using Code Object V4.
+	:3:hip_module.cpp           :361 : 2227377768546 us: 7670 : [7f7c6eddd180] ihipModuleLaunchKernel ( 0x0x16fe080, 2048, 1, 1, 		1024, 1, 1, 0, stream:<null>, 0x7ffded8ad260, char array:<null>, event:0, event:0, 0, 0 )
+	:1:hip_module.cpp           :254 : 2227377768572 us: Launch params (1024, 1, 1) are larger than launch bounds (64) for 		kernel _Z8MyKerneliPd
+	:3:hip_platform.cpp         :667 : 2227377768577 us: 7670 : [7f7c6eddd180] ihipLaunchKernel: Returned hipErrorLaunchFailure 		:
+	:3:hip_module.cpp           :493 : 2227377768581 us: 7670 : [7f7c6eddd180] hipLaunchKernel: Returned hipErrorLaunchFailure :
+```
+
+There is no known workaround at this time. 
+	
+## PYCACHE Folder Exists After ROCM SMI Library Uninstallation
+	
+Users may observe that the /opt/rocm-x/bin/__pycache__ folder continues to exist even after the rocm_smi_lib uninstallation.
+Workaround: Delete the /opt/rocm-x/bin/__pycache__ folder manually before uninstalling rocm_smi_lib.
 
 
 # Deploying ROCm
-AMD hosts both Debian and RPM repositories for the ROCm v3.8.x packages. 
+
+AMD hosts both Debian and RPM repositories for the ROCm packages. 
 
 For more information on ROCM installation on all platforms, see
 
 https://rocmdocs.amd.com/en/latest/Installation_Guide/Installation-Guide.html
 
+
+# Machine Learning and High Performance Computing Software Stack for AMD GPU
+
+For an updated version of the software stack for AMD GPU, see
+
+https://rocmdocs.amd.com/en/latest/Installation_Guide/Installation-Guide.html#software-stack-for-amd-gpu
+
+
+
 # Hardware and Software Support
 ROCm is focused on using AMD GPUs to accelerate computational tasks such as machine learning, engineering workloads, and scientific computing.
 In order to focus our development efforts on these domains of interest, ROCm supports a targeted set of hardware configurations which are detailed further in this section.
 
+**Note:** The AMD ROCm™ open software platform is a compute stack for headless system deployments. GUI-based software applications are currently not supported.
+
 #### Supported GPUs
 Because the ROCm Platform has a focus on particular computational domains, we offer official support for a selection of AMD GPUs that are designed to offer good performance and price in these domains.
 
+**Note:** The integrated GPUs of Ryzen are not officially supported targets for ROCm.
+
 ROCm officially supports AMD GPUs that use following chips:
 
-  * GFX8 GPUs
-    * "Fiji" chips, such as on the AMD Radeon R9 Fury X and Radeon Instinct MI8
-    * "Polaris 10" chips, such as on the AMD Radeon RX 580 and Radeon Instinct MI6
-  * GFX9 GPUs
-    * "Vega 10" chips, such as on the AMD Radeon RX Vega 64 and Radeon Instinct MI25
-    * "Vega 7nm" chips, such as on the Radeon Instinct MI50, Radeon Instinct MI60 or AMD Radeon VII
+* GFX9 GPUs
+
+  - "Vega 10" chips, such as on the AMD Radeon RX Vega 64 and Radeon Instinct MI25
+  
+  - "Vega 7nm" chips, such as on the Radeon Instinct MI50, Radeon Instinct MI60 or AMD Radeon VII, Radeon Pro VII
+
+* CDNA GPUs
+
+  - MI100 chips such as on the AMD Instinct™ MI100
+
 
 ROCm is a collection of software ranging from drivers and runtimes to libraries and developer tools.
 Some of this software may work with more GPUs than the "officially supported" list above, though AMD does not make any official claims of support for these devices on the ROCm software platform.
+
 The following list of GPUs are enabled in the ROCm software, though full support is not guaranteed:
 
   * GFX8 GPUs
@@ -238,7 +795,7 @@ As described [below](#limited-support), "Carrizo", "Bristol Ridge", and "Raven R
 However, they are not enabled in the HIP runtime, and may not work due to motherboard or OEM hardware limitations.
 As such, they are not yet officially supported targets for ROCm.
 
-For a more detailed list of hardware support, please see [the following documentation](https://rocm.github.io/hardware.html).
+For a more detailed list of hardware support, please see [the following documentation](https://en.wikipedia.org/wiki/List_of_AMD_graphics_processing_units).
 
 #### Supported CPUs
 As described above, GFX8 GPUs require PCIe 3.0 with PCIe atomics in order to run ROCm.
@@ -281,7 +838,7 @@ from the list provided above for compatibility purposes.
 
 ##### Limited support
 
-* ROCm 2.9.x should support PCIe 2.0 enabled CPUs such as the AMD Opteron, Phenom, Phenom II, Athlon, Athlon X2, Athlon II and older Intel Xeon and Intel Core Architecture and Pentium CPUs. However, we have done very limited testing on these configurations, since our test farm has been catering to CPUs listed above. This is where we need community support. _If you find problems on such setups, please report these issues_.
+* ROCm 4.x should support PCIe 2.0 enabled CPUs such as the AMD Opteron, Phenom, Phenom II, Athlon, Athlon X2, Athlon II and older Intel Xeon and Intel Core Architecture and Pentium CPUs. However, we have done very limited testing on these configurations, since our test farm has been catering to CPUs listed above. This is where we need community support. _If you find problems on such setups, please report these issues_.
 * Thunderbolt 1, 2, and 3 enabled breakout boxes should now be able to work with ROCm. Thunderbolt 1 and 2 are PCIe 2.0 based, and thus are only supported with GPUs that do not require PCIe 3.1.0 atomics (e.g. Vega 10). However, we have done no testing on this configuration and would need community support due to limited access to this type of equipment.
 * AMD "Carrizo" and "Bristol Ridge" APUs are enabled to run OpenCL, but do not yet support HIP or our libraries built on top of these compilers and runtimes.
   * As of ROCm 2.1, "Carrizo" and "Bristol Ridge" require the use of upstream kernel drivers.
@@ -294,11 +851,13 @@ from the list provided above for compatibility purposes.
 
 ##### Not supported
 
-* "Tonga", "Iceland", "Vega M", and "Vega 12" GPUs are not supported in ROCm 2.9.x
+* "Tonga", "Iceland", "Vega M", and "Vega 12" GPUs are not supported.
 * We do not support GFX8-class GPUs (Fiji, Polaris, etc.) on CPUs that do not have PCIe 3.0 with PCIe atomics.
   * As such, we do not support AMD Carrizo and Kaveri APUs as hosts for such GPUs.
   * Thunderbolt 1 and 2 enabled GPUs are not supported by GFX8 GPUs on ROCm. Thunderbolt 1 & 2 are based on PCIe 2.0.
 
+In the default ROCm configuration, GFX8 and GFX9 GPUs require PCI Express 3.0 with PCIe atomics. The ROCm platform leverages these advanced capabilities to allow features such as user-level submission of work from the host to the GPU. This includes PCIe atomic Fetch and Add, Compare and Swap, Unconditional Swap, and AtomicOp Completion.
+
 #### ROCm support in upstream Linux kernels
 
 As of ROCm 1.9.0, the ROCm user-level software is compatible with the AMD drivers in certain upstream Linux kernels.
@@ -325,9 +884,17 @@ For users that have the option of using either AMD's or the upstreamed driver, t
 |      |   | Does not include most up-to-date firmware |
 
 
+# Disclaimer
 
-## Machine Learning and High Performance Computing Software Stack for AMD GPU
+AMD®, the AMD Arrow logo, AMD Instinct™, Radeon™, ROCm® and combinations thereof are trademarks of Advanced Micro Devices, Inc.
 
-For an updated version of the software stack for AMD GPU, see
+Linux® is the registered trademark of Linus Torvalds in the U.S. and other countries.
+
+PCIe® is a registered trademark of PCI-SIG Corporation. Other product names used in this publication are for identification purposes only and may be trademarks of their respective companies.  
+
+Google®  is a registered trademark of Google LLC.
+
+Ubuntu and the Ubuntu logo are registered trademarks of Canonical Ltd.
+
+Other product names used in this publication are for identification purposes only and may be trademarks of their respective companies.
 
-https://rocmdocs.amd.com/en/latest/Installation_Guide/Installation-Guide.html#software-stack-for-amd-gpu

default.xml

@@ -1,27 +1,26 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <manifest>
     <remote name="roc-github"
-            fetch="http://github.com/RadeonOpenCompute/" />
+fetch="http://github.com/RadeonOpenCompute/" />
     <remote name="rocm-devtools"
-            fetch="https://github.com/ROCm-Developer-Tools/" />
+fetch="https://github.com/ROCm-Developer-Tools/" />
     <remote name="rocm-swplat"
-            fetch="https://github.com/ROCmSoftwarePlatform/" />
+fetch="https://github.com/ROCmSoftwarePlatform/" />
     <remote name="gpuopen-libs"
-            fetch="https://github.com/GPUOpen-ProfessionalCompute-Libraries/" />
+fetch="https://github.com/GPUOpen-ProfessionalCompute-Libraries/" />
     <remote name="gpuopen-tools"
-            fetch="https://github.com/GPUOpen-Tools/" />
+fetch="https://github.com/GPUOpen-Tools/" />
     <remote name="KhronosGroup"
-            fetch="https://github.com/KhronosGroup/" />
-    <default revision="refs/tags/rocm-3.8.0"
-             remote="roc-github"
-             sync-c="true"
-             sync-j="4" />
-    <!--list of projects for ROCM-->
+fetch="https://github.com/KhronosGroup/" />
+    <default revision="refs/tags/rocm-4.3.1"
+     remote="roc-github"
+     sync-c="true"
+     sync-j="4" />
+<!--list of projects for ROCM-->
     <project name="ROCK-Kernel-Driver" />
     <project name="ROCT-Thunk-Interface" />
     <project name="ROCR-Runtime" />
-    <project name="ROC-smi" />
-    <project name="rocm_smi_lib" remote="roc-github" />
+    <project name="rocm_smi_lib" />
     <project name="rocm-cmake" />
     <project name="rocminfo" />
     <project name="rocprofiler" remote="rocm-devtools" />
@@ -29,26 +28,30 @@
     <project name="ROCm-OpenCL-Runtime" />
     <project path="ROCm-OpenCL-Runtime/api/opencl/khronos/icd" name="OpenCL-ICD-Loader" remote="KhronosGroup" revision="6c03f8b58fafd9dd693eaac826749a5cfad515f8" />
     <project name="clang-ocl" />
-    <!--HIP Projects-->
+<!--HIP Projects-->
     <project name="HIP" remote="rocm-devtools" />
     <project name="HIP-Examples" remote="rocm-devtools" />
     <project name="ROCclr" remote="rocm-devtools" />
     <project name="HIPIFY" remote="rocm-devtools" />
-    <!-- The following projects are all associated with the AMDGPU LLVM compiler -->
-    <project name="llvm-project" path="llvm_amd-stg-open" />
+<!-- The following projects are all associated with the AMDGPU LLVM compiler -->
+    <project name="llvm-project" />
     <project name="ROCm-Device-Libs" />
     <project name="atmi" />
     <project name="ROCm-CompilerSupport" />
     <project name="rocr_debug_agent" remote="rocm-devtools" />
     <project name="rocm_bandwidth_test" />
+    <project name="half" remote="rocm-swplat" revision="37742ce15b76b44e4b271c1e66d13d2fa7bd003e" />
     <project name="RCP" remote="gpuopen-tools" revision="3a49405a1500067c49d181844ec90aea606055bb" />
-    <!-- gdb projects -->
+<!-- gdb projects -->
     <project name="ROCgdb" remote="rocm-devtools" />
     <project name="ROCdbgapi" remote="rocm-devtools" />
-    <!-- ROCm Libraries -->
+<!-- ROCm Libraries -->
+    <project name="rdc" remote="roc-github" />
     <project name="rocBLAS" remote="rocm-swplat" />
+    <project name="Tensile" remote="rocm-swplat" />
     <project name="hipBLAS" remote="rocm-swplat" />
     <project name="rocFFT" remote="rocm-swplat" />
+    <project name="hipFFT" remote="rocm-swplat" />
     <project name="rocRAND" remote="rocm-swplat" />
     <project name="rocSPARSE" remote="rocm-swplat" />
     <project name="rocSOLVER" remote="rocm-swplat" />
@@ -62,18 +65,10 @@
     <project name="hipCUB" remote="rocm-swplat" />
     <project name="rocPRIM" remote="rocm-swplat" />
     <project name="hipfort" remote="rocm-swplat" />
+    <project name="AMDMIGraphX" remote="rocm-swplat" />
     <project name="ROCmValidationSuite" remote="rocm-devtools" />
-    <!-- Projects for AOMP -->
-    <project name="ROCT-Thunk-Interface" path="aomp/roct-thunk-interface" remote="roc-github" />
-    <project name="ROCR-Runtime" path="aomp/rocr-runtime" remote="roc-github" />
-    <project name="ROCm-Device-Libs" path="aomp/rocm-device-libs" remote="roc-github" />
-    <project name="ROCm-CompilerSupport" path="aomp/rocm-compilersupport" remote="roc-github" />
-    <project name="rocminfo" path="aomp/rocminfo" remote="roc-github" />
-    <project name="HIP" path="aomp/hip-on-vdi" remote="rocm-devtools" />
-    <project name="aomp" path="aomp/aomp" remote="rocm-devtools" />
-    <project name="aomp-extras" path="aomp/aomp-extras" remote="rocm-devtools" />
-    <project name="flang" path="aomp/flang" remote="rocm-devtools" />
-    <project name="amd-llvm-project" path="aomp/amd-llvm-project" remote="rocm-devtools" />
-    <project name="ROCclr" path="aomp/vdi" remote="rocm-devtools" />
-    <project name="ROCm-OpenCL-Runtime" path="aomp/opencl-on-vdi" remote="roc-github" />
+<!-- Projects for OpenMP-Extras -->
+    <project name="aomp" path="openmp-extras/aomp" remote="rocm-devtools" />
+    <project name="aomp-extras" path="openmp-extras/aomp-extras" remote="rocm-devtools" />
+    <project name="flang" path="openmp-extras/flang" remote="rocm-devtools" />
 </manifest>

images/test.rst

@@ -0,0 +1 @@
+