[rocBLAS] [hipBLAS] CHANGELOG update for 7.2 (#2668) (#2849)

## Motivation

cherry-pick update CHANGELOG.md for 7.2.

Co-authored-by: Andrew Chapman <[email protected]>
Co-authored-by: Jeffrey Novotny <[email protected]>
Co-authored-by: Andrew Chapman <[email protected]>
Co-authored-by: Val Movsik <[email protected]>

[rocBLAS] Revert "[rocBLAS] Stream Order Allocation as default (#1516)…

…" due to rocHPL performance drop (multi-gpu) (#2241)

This reverts commit cc99c1c.

## Motivation
[https://ontrack-internal.amd.com/browse/SWDEV-558744](url) ~15% to ~47%
drops in rochpl-Mxp 2, 4 and 8GPUs

## Technical Details

Fixed the rocHPL performance drop in a multi-gpu setup. 

## Test Plan

**rocBLAS test:** 
Validate rocBLAS build using the './install.sh -dc -a auto' command
Run pre_checkin tests

> ./rocblas-test --gtest_filter=*pre_checkin*

**rocHPL test:** To benchmark performance, run the following commands:

>        cd rocHPL-MxP/
>        ./install.sh
>        
>        cd build/
>        
>        ./mpirun_rochplmxp -P 2 -Q 1 -N 179200 --NB 2560 -->For 2 GPU
>        
>        ./mpirun_rochplmxp -P 2 -Q 2 -N 250880 --NB 2560 -->For 4 GPU
>       
>        ./mpirun_rochplmxp -P 2 -Q 4 -N 358400 --NB 2560 -->For 8GPU

## Test Result

- rocBLAS Build successful
- rocblas-test pre_checkin passed
- rocHPL performance improved


## Submission Checklist

- [x] Look over the contributing guidelines at
https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

---------

Co-authored-by: Jeffrey Novotny <[email protected]>
Co-authored-by: JeniferC99 <[email protected]>

[rocBLAS] clients only needs find_package hipblaslt (#1762)

* hipblaslt when enables is required for a clients only build 
* swdev-556783

[rocm-libraries] ROCm/rocm-libraries#1768 (commit 0e8c708)

[rocBLAS] [hipBLAS] version and CHANGELOG update for 7.0.2
 (#1768)

Co-authored-by: Andrew Chapman <[email protected]>

Add gfx1150 and complete gfx1151 support (#1667)

* Add gfx1150 support

* Add gfx1151 yaml files

[rocm-libraries] ROCm/rocm-libraries#624 (commit 2f44be4)

hotfix 7.0 use omp config (#565)

find OpenMP config (#517)

First search for ROCm's libomp.so via openmp-config.cmake. This is what
we would prefer instead of searching for a system libomp.so/libgomp.so
    and then manually adding in a ROCm lib path.
    This methodology should still be RHEL-10 RPATH compliant.

(cherry picked from commit 408affb61fe3a97d7eecc458f6314da8418035d5)

[rocm-libraries] ROCm/rocm-libraries#624 (commit 2f44be4)

hotfix 7.0 use omp config (#565)

find OpenMP config (#517)

First search for ROCm's libomp.so via openmp-config.cmake. This is what
we would prefer instead of searching for a system libomp.so/libgomp.so
    and then manually adding in a ROCm lib path.
    This methodology should still be RHEL-10 RPATH compliant.

(cherry picked from commit 408affb61fe3a97d7eecc458f6314da8418035d5)

Zero imag portion of diagonal in syrk/herk (#1648) (#1651)

* Zero imag portion of diagonal in syrk/herk (#1648)

* Bump version to 4.4.1 and update changelog

Zero imag portion of diagonal in syrk/herk (#1648) (#1651)

* Zero imag portion of diagonal in syrk/herk (#1648)

* Bump version to 4.4.1 and update changelog

ROCm Release for rocBLAS 4.4.0