* elf: os: detect Android via clang compiler .ident note
* elf: os: detect Android via dependency on liblog.so
* main: split main into a bunch of "main routines"
[wip] since there are a few references to BinExport2
that are in progress elsewhre. Next commit will remove them.
* features: add BinExport2 declarations
* BinExport2: initial skeleton of feature extraction
* main: remove references to wip BinExport2 code
* changelog
* main: rename first position argument "input_file"
closes#1946
* main: linters
* main: move rule-related routines to capa.rules
ref #1821
* main: extract routines to capa.loader module
closes#1821
* add loader module
* loader: learn to load freeze format
* freeze: use new cli arg handling
* Update capa/loader.py
Co-authored-by: Moritz <mr-tz@users.noreply.github.com>
* main: remove duplicate documentation
* main: add doc about where some functions live
* scripts: migrate to new main wrapper helper functions
* scripts: port to main routines
* main: better handle auto-detection of backend
* scripts: migrate bulk-process to main wrappers
* scripts: migrate scripts to main wrappers
* main: rename *_from_args to *_from_cli
* changelog
* cache-ruleset: remove duplication
* main: fix tag handling
* cache-ruleset: fix cli args
* cache-ruleset: fix special rule cli handling
* scripts: fix type bytes
* main: nicely format debug messages
* helpers: ensure log messages aren't very long
* flake8 config
* binexport2: formatting
* loader: learn to load BinExport2 files
* main: debug log the format and backend
* elf: add more arch constants
* binexport: parse global features
* binexport: extract file features
* binexport2: begin to enumerate function/bb/insns
* binexport: pass context to function/bb/insn extractors
* binexport: linters
* binexport: linters
* scripts: add script to inspect binexport2 file
* inspect-binexport: fix xref symbols
* inspect-binexport: factor out the index building
* binexport: move index to binexport extractor module
* binexport: implement ELF/aarch64 GOT/thunk analyzer
* binexport: implement API features
* binexport: record the full vertex for a thunk
* binexport: learn to extract numbers
* binexport: number: skipped mapped numbers
* binexport: fix basic block address indexing
* binexport: rename function
* binexport: extract operand numbers
* binexport: learn to extract calls from characteristics
* binexport: learn to extract mnemonics
* pre-commit: skip protobuf file
* binexport: better search for sample file
* loader: add file extractors for BinExport2
* binexport: remove extra parameter
* new black config
* binexport: index string xrefs
* binexport: learn to extract bytes and strings
* binexport: cache parsed PE/ELF
* binexport: handle Ghidra SYMBOL numbers
* binexport2: handle binexport#78 (Ghidra only uses SYMBOL expresssions)
* main: write error output to stderr, not stdout
* scripts: add example detect-binexport2-capabilities.py
* detect-binexport2-capabilities: more documentation/examples
* elffile: recognize more architectures
* binexport: handle read_memory errors
* binexport: index flow graphs by address
* binexport: cleanup logging
* binexport: learn to extract function names
* binexport: learn to extract all function features
* binexport: learn to extract bb tight loops
* elf: don't require vivisect just for type annotations
* main: remove unused imports
* rules: don't eagerly import ruamel until needed
* loader: avoid eager imports of some backend-related code
* changelog
* fmt
* binexport: better render optional fields
* fix merge conflicts
* fix formatting
* remove Ghidra data reference madness
* handle PermissionError when searching sample file for BinExport2 file
* handle PermissionError when searching sample file for BinExport2 file
* add Android as valid OS
* inspect-binexport: strip strings
* inspect-binexport: render operands
* fix lints
* ruff: update config layout
* inspect-binexport: better align comments/xrefs
* use explicit search paths to get sample for BinExport file
* add initial BinExport tests
* add/update BinExport tests and minor fixes
* inspect-binexport: add perf tracking
* inspect-binexport: cache rendered operands
* lints
* do not extract number features for ret instructions
* Fix BinExport's "tight loop" feature extraction.
`idx.target_edges_by_basic_block_index[basic_block_index]` is of type
`List[Edges]`. The index `basic_block_index` was definitely not an
element.
* inspect-binexport: better render data section
* linters
* main: accept --format=binexport2
* binexport: insn: add support for parsing bare immediate int operands
* binexport2: bb: fix tight loop detection
ref #2050
* binexport: api: generate variations of Win32 APIs
* lints
* binexport: index: don't assume instruction index is 1:1 with address
* be2: index instruction addresses
* be2: temp remove bytes feature processing
* binexport: read memory from an address space extracted from PE/ELF
closes#2061
* be2: resolve thunks to imported functions
* be2: check for be2 string reference before bytes/string extraction overhead
* be2: remove unneeded check
* be2: do not process thunks
* be2: insn: polish thunk handling a bit
* be2: pre-compute thunk targets
* parse negative numbers
* update tests to use Ghidra-generated BinExport file
* remove unused import
* black reformat
* run tests always (for now)
* binexport: tests: fix test case
* binexport: extractor: fix insn lint
* binexport: addressspace: use base address recovered from binexport file
* Add nzxor charecteristic in BinExport extractor.
by referencing vivisect implementation.
* add tests, fix stack cookie detection
* test BinExport feature PRs
* reformat and fix
* complete TODO descriptions
* wip tests
* binexport: add typing where applicable (#2106)
* binexport2: revert import names from BinExport2 proto
binexport2_pb.BinExport2 isnt a package so we can't import it like:
from ...binexport2_pb.BinExport2 import CallGraph
* fix stack offset numbers and disable offset tests
* xfail OperandOffset
* generate symbol variants
* wip: read negative numbers
* update tight loop tests
* binexport: fix function loop feature detection
* binexport: update binexport function loop tests
* binexport: fix lints and imports
* binexport: add back assert statement to thunk calculation
* binexport: update tests to use Ghidra binexport file
* binexport: add additional debug info to thunk calculation assert
* binexport: update unit tests to focus on Ghidra
* binexport: fix lints
* binexport: remove Ghidra symbol madness and fix x86/amd64 stack offset number tests
* binexport: use masking for Number features
* binexport: ignore call/jmp immediates for intel architecture
* binexport: check if immediate is a mapped address
* binexport: emit offset features for immediates likely structure offsets
* binexport: add twos complement wrapper insn.py
* binexport: add support for x86 offset features
* binexport: code refactor
* binexport: init refactor for multi-arch instruction feature parsing
* binexport: intel: emit indirect call characteristic
* binexport: use helper method for instruction mnemonic
* binexport: arm: emit offset features from stp instruction
* binexport: arm: emit indirect call characteristic
* binexport: arm: improve offset feature extraction
* binexport: add workaroud for Ghidra bug that results in empty operands (no expressions)
* binexport: skip x86 stack string tests
* binexport: update mimikatz.exe_ feature count tests for Ghidra
* core: loader: update binja import
* core: loader: update binja imports
* binexport: arm: ignore number features for add instruction manipulating stack
* binexport: update unit tests
* binexport: arm: ignore number features for sub instruction manipulating stack
* binexport: arm: emit offset features for add instructions
* binexport: remove TODO from tests workflow
* binexport: update CHANGELOG
* binexport: remove outdated TODOs
* binexport: re-enable support for data references in inspect-binexport2.py
* binexport: skip data references to code
* binexport: remove outdated TODOs
* Update scripts/inspect-binexport2.py
* Update CHANGELOG.md
* Update capa/helpers.py
* Update capa/features/extractors/common.py
* Update capa/features/extractors/binexport2/extractor.py
* Update capa/features/extractors/binexport2/arch/arm/insn.py
Co-authored-by: Moritz <mr-tz@users.noreply.github.com>
* initial add
* test binexport scripts
* add tests using small ARM ELF
* add method to get instruction by address
* index instructions by address
* adjust and extend tests
* handle operator with no children bug
* binexport: use instruction address index
ref: https://github.com/mandiant/capa/pull/1950/files#r1728570811
* inspect binexport: handle lsl with no children
add pruning phase to expression tree building
to remove known-bad branches. This might address
some of the data we're seeing due to:
https://github.com/NationalSecurityAgency/ghidra/issues/6821
Also introduces a --instruction optional argument
to dump the details of a specific instruction.
* binexport: consolidate expression tree logic into helpers
* binexport: index instruction indices by address
* binexport: introduce instruction pattern matching
Introduce intruction pattern matching to declaratively
describe the instructions and operands that we want to
extract. While there's a bit more code, its much more
thoroughly tested, and is less brittle than the prior
if/else/if/else/if/else implementation.
* binexport: helpers: fix missing comment words
* binexport: update tests to reflect updated test files
* remove testing of feature branch
---------
Co-authored-by: Moritz <mr-tz@users.noreply.github.com>
Co-authored-by: Mike Hunhoff <mike.hunhoff@gmail.com>
Co-authored-by: mr-tz <moritz.raabe@mandiant.com>
Co-authored-by: Lin Chen <larch.lin.chen@gmail.com>
* web: index: add gif of capa running
* index: add screencast of running capa
produced via:
```
asciinema capa.cast
./capa Practical\ Malware\ Analysis\ Lab\ 01-01.dll_
<ctrl-d>
agg --no-loop --theme solarized-light capa.cast capa.gif
```
* web: index: start to sketch out style
* web: landing page
* web: merge rules website
* web: rules: update bootstrap and integrate rules
* web: rules: use pygments to syntax highlight rules
Use the Pygments syntax-highlighting library to parse
and render the YAML rule content. This way we don't have
to manually traverse the rule nodes and emit lists; instead,
we rely on the fact that YAML is pretty easy for humans
to read and let them consume it directly, with some text
formatting to help hint at the types/structure.
* web: rules: use capa to load rule content
capa (the library) has routines for deserializing the YAML
content into structured objects, which means we can use tools
like mypy to find bugs. So, prefer to use those routines instead
of parsing YAML ourselves.
* web: rules: linters
Run and fix the issues identified by the following linters:
- isort
- black
- ruff
- mypy
* web: rules: add some links to rule page
Add links to the following external resources:
- GitHub rule source in capa-rules repo
- VirusTotal search for matching samples
* web: rules: accept ?q= parameter for initial search
Update the rules landing page to accept a HTTP
query parameter named "q" that specifies an initial
search term to to pass to pagefind. This enables
external pages link to rule searches.
* web: rules: add link to namespace search
* web: rules: use consistent header
Import header from root capa landing page.
* web: rules: add umami script
* web: add initial whats new section, TODOs
* web: rules: remove old images
* changelog
* CI: remove temporary branch push event triggers
* Delete web/rules/public/css/bootstrap-4.5.2.min.css
* Delete web/rules/public/js/bootstrap-4.5.2.min.js
* Delete web/public/img/capa.cast
* Rename readme.md to README.md
* web: rules: add scripts to pre-commit configs
* web: rules: add scripts to pre-commit configs
* lints
* ci: add temporary branch push trigger to get incremental builds
* web: rules: assert start_dir must exist
* ci: web: rules: deep checkout so we can get rule history
* web: rules: check output of subprocess
* web: rules: factor out common CSS
* web: rules: fix header links
* web: rules: only index rule content, not surrounding text
* ci: web: remote temporary branch push trigger
* elf: read segment memory size
* elf: add routine to read mapped memory
* elf: better detect OS for binaries compiled by Go
* elf: guess OS from Go source filenames
* changelog
* elf: mypy
* merge
* elf: add OS detection based on vDSO strings
* elf: document VTGrep searches
* elf: describe further technique to identify Go binaries
* elf: search for `.go.buildinfo` section via @yelhamer
* black
* elf: detect Alpine Linux ident
* elf: log interest symtab entries
* tests: add test for OS detection by Go buildinfo
* loader: handle missing viv modules
* pre-commit: run deptry before tests (which are slow)
* loader: describe removing viv symbolic switch solver
* pyproject: add PyGithub for deptry
* black
* Add deptry support
This commit resolves#1497.
Note: known_first_party refers to modules that are supposed to be
local, i.e. idaapi, ghidra, java, binaryninja, ... etc.
* adjust running stages for deptry hook
* adjust deptry exclusions, and humanize dependency
---------
Co-authored-by: Soufiane Fariss <soufiane.fariss@um5s.net.ma>
