Enabling Manywheel builds here: pytorch#138732

During the build I observe the failure with cuda jobs:

```
-- Compiler does not support SVE extension. Will not build perfkernels.
-- Found CUDA: /usr/local/cuda (found version "11.8")
-- The CUDA compiler identification is unknown
CMake Error at cmake/public/cuda.cmake:47 (enable_language):
  No CMAKE_CUDA_COMPILER could be found.

  Tell CMake where to find the compiler by setting either the environment
  variable "CUDACXX" or the CMake cache entry CMAKE_CUDA_COMPILER to the full
  path to the compiler, or to the compiler name if it is in the PATH.
Call Stack (most recent call first):
  cmake/Dependencies.cmake:44 (include)
  CMakeLists.txt:851 (include)
```

While correct sequence suppose to be:
```
-- Found CUDA: /usr/local/cuda (found version "11.8")
-- The CUDA compiler identification is NVIDIA 11.8.89
-- Detecting CUDA compiler ABI info
-- Detecting CUDA compiler ABI info - done
-- Check for working CUDA compiler: /usr/local/cuda/bin/nvcc - skipped
-- Detecting CUDA compile features
-- Detecting CUDA compile features - done
-- Found CUDAToolkit: /usr/local/cuda/include (found version "11.8.89")
```

Issue found to be missing PATH setting in 2_28 Docker file.  This section exist in CentOS Docker file here:
https://github.com/pytorch/pytorch/blob/main/.ci/docker/manywheel/Dockerfile#L174-L175

(Please Note these Docker images are not used yet. The pytorch#138732 should enable using these images)
Pull Request resolved: pytorch#139631
Approved by: https://github.com/malfet, https://github.com/huydhn