Cuda Component Refactor

[Treece-Burgess]

Pull Request Description

This pull request addresses adding a device qualifier to the Cuda component. This will greatly decrease the number of native events that are output when running ./papi_native_avail. For example, if running on Guyot which has a total of eight A100 NVIDIA GPUs. The total number of native events that are output is 2,105,856 with output currently being formatted as:

===============================================================================
| cuda:::dram__bytes.avg:device=0                                              |     
|            # of bytes accessed in DRAM. Units=(bytes) Numpass=1              |     
--------------------------------------------------------------------------------

With adding a device qualifier to the Cuda component with this pull request, we would see the total number of native events on Guyot decreased to 263,232 with the updated output formatted as:

===============================================================================
| cuda:::dram__bytes.avg                                                       |
|            # of bytes accessed in DRAM. Units=(bytes) Numpass=1              |
|     :device=0                                                                |
|            Mandatory device qualifier [0,1,2,3,4,5,6,7]                      |
--------------------------------------------------------------------------------

Other example outputs will be shown below.

Hopper1 has a single NVIDIA GH200 GPU and output for that system appears as follows:

===============================================================================
| cuda:::FE_A.TriageCompute.gr__cycles_active.avg                              |
|            # of cycles where GR was active. Units=(sys_clks) Numpass=1       |
|     :device=0                                                                |
|            Mandatory device qualifier [0]                                    |
--------------------------------------------------------------------------------

Leconte has eight NVIDIA Tesla V100 GPUs and output for that system appears as follows:

===============================================================================
| cuda:::dram__bytes.avg                                                       |
|            # of bytes accessed in DRAM. Units=(bytes) Numpass=1              |
|     :device=0                                                                |
|            Mandatory device qualifier [0,1,2,3,4,5,6,7]                      |
--------------------------------------------------------------------------------

Author Checklist

• Description
  Why this PR exists. Reference all relevant information, including background, issues, test failures, etc
• Commits
  Commits are self contained and only do one thing
  Commits have a header of the form: module: short description
  Commits have a body (whenever relevant) containing a detailed description of the addressed problem and its solution
• Tests
  The PR needs to pass all the tests