benchdnn: memory: gpu: enable support for RNG memory fill

[yehudaorel]

Description

Systems with data compression enabled by the driver by default wont generate meaningful performance data with benchdnn ... --mode=F

• Reuse ocl_philox.h routine to directly fill gpu mem with random vals via a dedicated opencl kernel.

Fixes MFDNN-12589

Checklist

General

• [X] Do all unit and benchdnn tests (make test and make test_benchdnn_*) pass locally for each commit?
• [X] Have you formatted the code using clang-format?

[yehudaorel]

make test perf-gpu disable arch_gpu_xe-hpc disable arch_gpu_xe-lp disable arch_gpu_xe2-lpg