mirror of
https://github.com/frida/frida-rust.git
synced 2026-11-07 17:13:59 -10:00
Kirby Kuehl
7f9c5eaa17
* feat: upgrade to Frida 17.16.4 and add AVX-512/ARM64 PAC instructions Upgrade Frida from 17.16.1 to 17.16.4 across all three devkit components. Add new instruction writer support from Frida 17.16.1+: x86_64 Writer (AVX-512 instructions): - kmovq: k-register mask operations for AVX-512 - vextracti64x4: extract 256 bits from 512-bit ZMM register - vinserti64x4: insert 256 bits into 512-bit ZMM register - vmovdqu64: unaligned 512-bit moves to/from ZMM registers ARM64 Writer: - movk: move with keep for building 64-bit constants incrementally - pacia: Pointer Authentication Code for ARMv8.3-A security All changes tested with full quality gate (fmt, clippy, build, test). * fix(aarch64): correct put_movk_reg_imm parameter types The function signature expects (reg, imm: u16, shift: u32), not (reg, imm: u32, shift: u8). This was revealed after merging upstream which had updated bindings. * feat(x86_64): add comprehensive instruction writer coverage Add ~70 additional x86_64 instruction writer methods: Writer utilities: - clear, cur, offset: Writer state management - set_target_cpu, set_target_abi: Target configuration - get_cpu_register_for_nth_argument: ABI introspection Basic instructions: - put_breakpoint, put_padding, put_nop_padding: Debug/alignment - put_u8, put_s8: Raw byte insertion Call instructions: - put_call_indirect, put_call_near_label, put_call_reg - put_call_reg_offset_ptr, put_call_reg_with_arguments - put_call_*_with_aligned_arguments variants - Full unsafe argument array APIs Flag manipulation: - put_clc, put_stc, put_cld, put_std: Flag operations - put_lahf, put_sahf: Flag register access CPU instructions: - put_cpuid, put_lfence, put_rdtsc, put_pause: CPU features Comparison/test: - put_cmp_imm_ptr_imm_u32, put_cmp_reg_i32, put_cmp_reg_reg - put_cmp_reg_offset_ptr_reg, put_test_reg_u32 Conditional jumps: - put_jcc_short, put_jcc_near: Full conditional jump support Atomic operations: - put_inc_reg_ptr, put_dec_reg_ptr - put_lock_inc_imm32_ptr, put_lock_dec_imm32_ptr - put_lock_cmpxchg_reg_ptr_reg, put_lock_xadd_reg_ptr_reg - put_xchg_reg_reg_ptr Memory operations: - put_mov_reg_near_ptr, put_mov_near_ptr_reg - put_mov_reg_fs_*, put_mov_fs_*: FS segment access - put_mov_reg_gs_*, put_mov_gs_*: GS segment access - put_movdqu_*, put_movq_*: XMM register moves FPU state: - put_fxsave_reg_ptr, put_fxrstor_reg_ptr All changes tested with full quality gate (fmt, clippy, build, test). * fix(aarch64): correct doctest imports for put_pacia_reg_reg Fix imports in the put_pacia_reg_reg example to use the public API. Aarch64Register is re-exported from instruction_writer, not aarch64 submodule. * fix(x86_64): remove duplicate doc comment causing clippy error Remove duplicate doc comment for put_kmovq_kreg_reg_offset_ptr that was causing clippy::empty_line_after_doc_comments error.
frida-rust
Rust bindings for Frida.
Install
- Build Frida, or download the devkits for your system (see
frida-gumorfrida-coreREADME for version) - For crate installation:
- Move the frida-gum and frida-core devkits into
rustc-link-search, e.g.:/usr/local/{include, lib}on Unix
- Move the frida-gum and frida-core devkits into
- For local development:
- Move the frida-gum devkit into
frida-gum-sys, and the frida-core devkit intofrida-sysandcargo buildin the root
- Move the frida-gum devkit into