gpu-unlock: Unlocking the FP/BF16 instruction issue rate of Nvidia 3000 series graphics cards

The Linux kernel module gpu-unlock 1.0.0 can double the FP/BF16 FMA instruction issue rate of GA102, GA103, GA104, and GA106 graphics cards, bringing them up to professional card levels. On the 3090 Ti, BF16 rises from 87.50 TFLOPS to 157.41 TFLOPS, FP16 from 87.46 to 118.39 TFLOPS, and TF32 from 42.58 to 82.61 TFLOPS. The module completes the unlock before the NVIDIA driver binds the graphics card, so there is no need to manually patch after each driver update.