Feat/cuzk/bucket point reduction

[FoodChain1028]

This PR introduce bpr (parallel) Bucket point redution algorithm in gpu.

New Utility Functions:

• Added bpr.metal, with 2 kernels: bpr_stage_1 and bpr_stage_2.

Refactoring and Enhancements:

• Refactored Jacobian addition (jacobian_add_2007_bl) to handle special cases where one of the points is zero or both points are equal (mopro-msm/src/msm/metal_msm/shader/curve/jacobian.metal). [1] [2]
• Moved the jacobian_dbl_2009_l function to the top of the file and added it to the jacobian_add_2007_bl function to handle doubling when points are equal (mopro-msm/src/msm/metal_msm/shader/curve/jacobian.metal). [1] [2]

New Functions:

• Added jacobian_scalar_mul function to perform scalar multiplication on Jacobian points (mopro-msm/src/msm/metal_msm/shader/curve/jacobian.metal).
• Added a new Metal kernel run for scalar multiplication in jacobian_scalar_mul.metal (mopro-msm/src/msm/metal_msm/shader/curve/jacobian_scalar_mul.metal).

Tests:

• Added bpr_stage_1 test in rust side (mopro-msm/src/msm/metal_msm/shader/cuzk/bpr.rs)
• Adding bpr_stage_2 test in rust side.

TODOs

• Check the result of bpr_stage_1 is correct.
• Check the result of bpr_stage_2 is correct.