Add Support for Geekbench 6 Benchmark

[leyli16]

This PR addresses the requirements outlined in [Issue #5238]. The following changes have been implemented:

Summary of Changes:

1. Benchmarking Framework Setup

• Created geekbench_benchmark.py files in both linux_benchmarks and windows_benchmarks directories.
• Defined BENCHMARK_NAME and BENCHMARK_CONFIG constants, allowing PKB to recognize and locate the benchmark.
• Added GetConfig, Prepare, Run, and Cleanup functions, each containing initial structure with pass/return statements, making the benchmark operational within PKB.

2. Geekbench Installation Support

• Developed geekbench.py package within linux_packages and windows_packages for installing Geekbench.
• Implemented an Install function that takes a vm as input, executing necessary shell commands to install Geekbench.
• Prioritized installation support for Ubuntu 24.04 and Windows. Added unit tests using mock to ensure the vm.RemoteCommand calls are correctly issued for installation.

3. Running and Parsing Geekbench6

• Manually ran Geekbench6 locally, producing output used for parsing and testing.

• Stored Geekbench output in data/geekbench as a raw data file for testing the parsing function.

• Developed ParseResults function in geekbench.py to process Geekbench output. The function produces a list of PKB Samples, with each sample representing a unique metric with:
  = metric_name, metric_value, metric_unit, and metric_metadata.

• Added unit tests to verify accurate parsing and structured output for each metric from both CPU and GPU tests.
  

---

Update from 11/29:

1. Moved Geekbench test files to appropriate directories for better organization.
2. Updated Prepare() to install Geekbench (vm.install('geekbench')).
3. Added a TODO in Run() to trigger Geekbench execution.
4. Logged informative messages for parsing failures in ParseResults while continuing execution.
5. Refactored ParseResults() to use sample.Sample for consistent output.

Example Output:
        [
            Sample(
                metric="Single-Core Score",
                value=1795,
                unit="points",
                metadata={"category": "Single-Core"},
                timestamp=1699815932.123
            ),
            Sample(
                metric="Single-Core File Compression",
                value=269.3,
                unit="MB/sec",
                metadata={
                    "category": "Single-Core",
                    "test": "File Compression",
                    "score": 1875
                },
                timestamp=1699815932.123
            )
        ]

6. Renamed references to geekbench_benchmark for naming consistency.

[jellyfishcake]

Very nice! Some small nits and questions.

[jellyfishcake]

Hi leyli16, it has came to my attention that we already have a geekbench6 linux implementation, you cannot see it because it is not externalized here on github.
I am really sorry about the confusion. I can externalize our existing implementation so that we can build upon it to add the windows implementation.
Happy to work with you to contribute towards a benchmark that does not have a conflict. Feel free to reach out.

[leyli16]

Hi leyli16, it has came to my attention that we already have a geekbench6 linux implementation, you cannot see it because it is not externalized here on github.
I am really sorry about the confusion. I can externalize our existing implementation so that we can build upon it to add the windows implementation.
Happy to work with you to contribute towards a benchmark that does not have a conflict. Feel free to reach out.

Thank you for the clarification. I’ve already addressed the Windows implementation, and I’d be happy to collaborate further to ensure there’s no conflict. Let me know how I can assist!