Repository navigation

Vectorized scanning for ARM Neon - #40

Merged

ZeroMemes merged 70 commits into

masterfrom

feat/neon

Jul 3, 2026

Merged

Vectorized scanning for ARM Neon#40
ZeroMemes merged 70 commits into
masterfrom
feat/neon

Commits on Jun 14, 2026
- Initial Neon implementation using sse2neon
  ZeroMemes
  committed
  c1605f9
- Ignore sse2neon warnings
  ZeroMemes
  committed
  39f9201
- Missing Mode name
  ZeroMemes
  committed
  8df8dfc
Commits on Jun 15, 2026
- Implementation throughput comparison benchmark
  ZeroMemes
  committed
  17ae035
- Log processor name
  ZeroMemes
  committed
  ee66f43
- Only build `libhat_benchmark_compare_impl`
  ZeroMemes
  committed
  77fbf3e
- Neon substitutions
  ZeroMemes
  committed
  71fd67a
- Architecture guard + use time defaults
  ZeroMemes
  committed
  d23f6da
- Avoid double movemask
  ZeroMemes
  committed
  c002976
- Use 4-bit mask
  ZeroMemes
  committed
  aa6d5cc
- Fix warning
  ZeroMemes
  committed
  c7a5b5e
- Increase benchmark time
  ZeroMemes
  committed
  4ff162a
- Remove sse2neon
  ZeroMemes
  committed
  2013e92
- Missing neon header include
  ZeroMemes
  committed
  2fa461d
- Improve neon intrinsic type conformance
  ZeroMemes
  committed
  1e42ef4
- Hopefully resolve remaining macOS compile errors
  ZeroMemes
  committed
  826b640
- Compile Windows on ARM benchmark using Clang
  ZeroMemes
  committed
  82a1517
- Test manual bitwise optimization
  ZeroMemes
  committed
  ce4768e
- Merge branch 'master' into feat/neon
  ZeroMemes
  committed
  02dd07d
- Benchmark ARM Neon against Chromium
  ZeroMemes
  committed
  16f2eed
- Use cache for benchmark workflow
  ZeroMemes
  committed
  af7e14f
- Perhaps
  ZeroMemes
  committed
  45a85a2
Commits on Jun 17, 2026
- Extract masks using `vget_lane_u64`
  ZeroMemes
  committed
  af24489
- Merge branch 'master' into feat/neon
  ZeroMemes
  committed
  bab92dc
- Debug mode optimizations for NEON
  ZeroMemes
  committed
  3699104
- Revert unrolling
  ZeroMemes
  committed
  2169688
Commits on Jun 18, 2026
- Add macOS to ARM64 testing
  ZeroMemes
  committed
  df8bb55
- Partial mach-o Process
  
  Imrglop
  authored and
  ZeroMemes
  committed
  36b05c7
- Missing source file
  ZeroMemes
  committed
  ac4712e
- Missing include for `std::abort`
  ZeroMemes
  committed
  7e7c29d
- Possibly fix macOS process APIs
  ZeroMemes
  committed
  afc7c1d
- Add Linux to ARM64 testing
  ZeroMemes
  committed
  3cec491
- Fix Linux compile
  ZeroMemes
  committed
  7663ec0
- Actually fix compile + adjust workflow
  ZeroMemes
  committed
  bf19f53
- Implement `system_info_arm` on Linux
  ZeroMemes
  committed
  3671c7f
Commits on Jun 19, 2026
- Adjust mask iteration
  ZeroMemes
  committed
  b747018
- ALLEGEDLY I can't use `_CountTrailingZeros64`
  ZeroMemes
  committed
  9ae6cfc
Commits on Jun 20, 2026
- Optimize `frequency.py` using NumPy
  ZeroMemes
  committed
  5ddba36
- Avoid `bytes` copy with `memoryview`
  ZeroMemes
  committed
  3d1e9ea
Commits on Jun 22, 2026
- Merge branch 'master' into feat/neon
  ZeroMemes
  committed
  6343504
- Add new NTTP to `segment_scan` call
  ZeroMemes
  committed
  95d2437
- Replace `u64` extract with `vmaxvq_u32`
  ZeroMemes
  committed
  d280b25
Commits on Jun 24, 2026
- Merge branch 'master' into feat/neon
  ZeroMemes
  committed
  18e5111
Commits on Jun 26, 2026
- Update README with macOS support
  ZeroMemes
  committed
  c0aa557
- Add test that validates presence of default module segments
  ZeroMemes
  committed
  3c4cfda
- Missing include for `std::abort`
  ZeroMemes
  committed
  9c613eb
Commits on Jun 27, 2026
- Recognize `PT_GNU_RELRO` segment type
  ZeroMemes
  committed
  769bfd1
- Fix some includes on macOS
  ZeroMemes
  committed
  61c7673
Commits on Jun 28, 2026
- Add Android build to CMake workflow
  ZeroMemes
  committed
  32cdfa4
- Adjust the minimum Android SDK to `android-21`
  ZeroMemes
  committed
  998580d
Commits on Jun 30, 2026
- Build all currently supported Android ABIs
  ZeroMemes
  committed
  457bba0
- `vmaxvq_u32` is not supported on ARMv7
  ZeroMemes
  committed
  b51f235
Commits on Jul 1, 2026
- Attempt android emulation for testing
  ZeroMemes
  committed
  8fec0d8
- Cache AVD snapshot
  ZeroMemes
  committed
  004966f
- Exclude `armeabi-v7a` from AVD cache steps
  ZeroMemes
  committed
  453c100
- Include ABI in AVD cache key
  ZeroMemes
  committed
  1f5c8a2
- Fix options for Android test step
  ZeroMemes
  committed
  3b572dd
- Fix `get_module` on Android
  ZeroMemes
  committed
  57d79a5
- Exclude `armeabi-v7a` and `arm64-v8a` from testing
  ZeroMemes
  committed
  4f2a7a0
Commits on Jul 2, 2026
- Add 4-byte alignment
  ZeroMemes
  committed
  44beb87
- Switch benchmark to manual dispatch
  ZeroMemes
  committed
  d74b8c4
- Add AArch64 scan hint
  ZeroMemes
  committed
  eb3690b
- Actually update the scan_hint enum
  ZeroMemes
  committed
  4a8e7db
Commits on Jul 3, 2026
- Update README to reflect Android support
  ZeroMemes
  committed
  f7dd712
- Update README to include optimizations specific to ARM
  ZeroMemes
  committed
  9ee9705
- Adjust formatting
  ZeroMemes
  committed
  f7bede0
- Remove `-Wno-unused-command-line-argument` workaround
  ZeroMemes
  committed
  053791e
- Detect ISA extensions at compile time
  ZeroMemes
  committed
  57561b5
- Theoretical 32-bit Mach-O support
  ZeroMemes
  committed
  946a81d
- space
  ZeroMemes
  committed
  4f62dff

Commits

Commits on Jun 14, 2026

Commits on Jun 15, 2026

Commits on Jun 17, 2026

Commits on Jun 18, 2026

Commits on Jun 19, 2026

Commits on Jun 20, 2026

Commits on Jun 22, 2026

Commits on Jun 24, 2026

Commits on Jun 26, 2026

Commits on Jun 27, 2026

Commits on Jun 28, 2026

Commits on Jun 30, 2026

Commits on Jul 1, 2026

Commits on Jul 2, 2026

Commits on Jul 3, 2026