Hex-Rays has officially rolled out IDA Pro 9.0 (build 240925), and it is anything but a minor point release. This update represents a fundamental shift in the legendary disassembler’s architecture, finally dragging its user interface into the modern era while solving a problem that has plagued reverse engineers for years: native ARM64 decompilation.
If you reverse-engineer iOS kernels, Android native libraries, or Apple Silicon malware, version 9.0 is likely an instant upgrade.
(type)expr now behaves more like C.array[index] are noticeably sharper.The decompiler uses new pattern matching to suggest better variable names: IDA Pro 9.0.240925
lpMutexAttributes → lpMutexAttrscbSize → sizeThis build is distributed exclusively to:
Warning against pirated builds: Keygens and cracked versions of IDA Pro 9.0.240925 circulating on torrent sites are known to contain remote access trojans (RATs). Given the sensitive nature of RE work (analyzing malware, proprietary software), running an untrusted disassembler is catastrophic. Always verify checksums (SHA256: e4f7a...) via the official Hex-Rays file integrity page. IDA Pro 9
Previous versions relied on the ARM32 decompiler with emulation for 64-bit. Version 9.0 introduces a native ARM64 decompiler backend.
Supported features:
__int128 operations and NEON intrinsics.With the rise of Apple Silicon (M1/M2/M3) and ARM64e (featuring pointer authentication - PAC), linear analysis was slow. The build 240925 implements threaded parallel analysis for ARM64 binaries. A typical iOS kernelcache (approx. 200MB) loads and analyzes in under 90 seconds, down from 6 minutes in IDA 8.x.
The changelog for 9.0.240925 is extensive. Below are the headline features that redefine the reverse engineering workflow. Simplified type casting: (type)expr now behaves more like
ida_typeinf.get_type_by_name() for programmatic type queries.idc.GetDisasm() – up to 30% quicker in loops.