Metal MSM V2

What's Changed

• github actions pipeline updated by @yaroslavyaroslav in #18
• Update README.md by @0xalizk in #15
• Slightly fixed shaders building pipeline by @yaroslavyaroslav in #19
• Bigint impl for refactored metal EC backend by @moven0831 in #21
• Field impl for refactored metal EC backend by @moven0831 in #22
• Montgomery Multiplication for refactored metal EC backend by @moven0831 in #23
• fix(ci): temporary commented precomputed msm part for the fix of msm correctness by @moven0831 in #24
• ci(metal_msm): Temporarily disable GPU benchmarks by @moven0831 in #27
• Test/metal/bigint by @moven0831 in #26
• Doc/add license by @FoodChain1028 in #28
• Fix ff.metal and refactor rust test by @moven0831 in #29
• Test/metal/mont by @moven0831 in #30
• Feat/metal/ec by @moven0831 in #31
• fix: add #pragma once as the header guard to prevent redefinition error by @FoodChain1028 in #35
• Feat/utils/limbs conversion by @moven0831 in #36
• Feat/metal/get constant by @moven0831 in #37
• Feat/curve/jacobian scalar mul by @FoodChain1028 in #38
• Feat/metal/ff reduce by @moven0831 in #41
• refactor limb conversion: update GenericLimbConversion to BigInt (incl. BigInt<8> support) by @moven0831 in #42
• Feat/metal/cuzk/barrett_reduction by @moven0831 in #44
• Fix mont_mul_cios with 16-limbs inputs by @moven0831 in #47
• Metal shared logging added by @yaroslavyaroslav in #48
• Feat/cuzk/convert point coords and scalar decompose by @moven0831 in #50
• Feat/cuzk/transpose by @moven0831 in #51
• Bpr enhancements by @yaroslavyaroslav in #53
• Feat/cuzk/new bucket point reduction by @FoodChain1028 in #52
• Feat/cuzk/smvp by @moven0831 in #54
• refactor(metal): remove bn254 p parameter from curve and field ops by @moven0831 in #55
• feat: add big int as input for pbpr by @FoodChain1028 in #56
• refactor(metal): introduce FieldElement struct and abstract modulus p by @moven0831 in #57
• Refactor/test/resource cleanup by @moven0831 in #58
• Ci/metal shaders by @moven0831 in #61
• Perf/mont mul by @moven0831 in #62
• Refactor/move helper structures to utils by @FoodChain1028 in #63
• fix(metal): fix bit shift calc in point conversion shader by @moven0831 in #65
• fix(metal): update last chunk bit shift calculation and improve tests by @moven0831 in #66
• Feat/cuzk/reproduce on cpu by @moven0831 in #67
• fix(metal): Correct Phase 3 of cuzk/transpose by @moven0831 in #68
• chore: update dependencies and related usage changes by @moven0831 in #69
• fix(metal): update pbpr shaders from single-kernel approach to two-stage pipeline by @moven0831 in #70
• Test/gpu pipeline e2e by @moven0831 in #71
• Refactor Metal MSM pipeline with improved architecture by @moven0831 in #76
• Optimize pack_affine_and_scalars with Parallel Execution by @moven0831 in #78
• chore(metal_msm): clarify dispatch params management by @moven0831 in #80
• chore(metal_msm): add timing benchmarks by @moven0831 in #81
• Refactor Metal MSM with ShaderManager and Improved Buffer Management by @moven0831 in #82
• Shaders precompile instead of compiling on the fly by @yaroslavyaroslav in #83
• Refactor shader build logic by @moven0831 in #84
• Build iOS example app for metal MSM benchmark by @moven0831 in #85
• Optimize Performance with Dynamic Thread and Threadgroup Config by @moven0831 in #86
• Fine-tune Window Size with Input Size by @moven0831 in #87
• Refactor/msl level optimization by @moven0831 in #88
• docs: add benchmark for performance comparisons by @moven0831 in #90
• Profiling results by @yaroslavyaroslav in #91
• Bump version to v2 by @moven0831 in #92

Full Changelog: v0.1.0...v0.2.0