wehub-resource-sync
44 lines
1.4 KiB
Markdown
44 lines
1.4 KiB
Markdown
# LLDB.md - LLDB Backend Notes for CLI-Anything
|
|
|
|
## Why LLDB Python API
|
|
|
|
This harness follows the CLI-Anything principle:
|
|
|
|
> Direct integration with actual software backends — full professional capabilities, zero compromises.
|
|
|
|
Using `import lldb` is the native backend integration route and avoids fragile
|
|
text parsing from `lldb -b -o ...` subprocess output.
|
|
|
|
## Integration Strategy
|
|
|
|
1. Try `import lldb`
|
|
2. On failure, run `lldb -P` to discover LLDB Python module directory
|
|
3. Prepend discovered path to `sys.path`
|
|
4. Retry import
|
|
|
|
This is implemented in `utils/lldb_backend.py`.
|
|
|
|
## Session Lifecycle
|
|
|
|
- Initialize LLDB debugger (`SBDebugger.Initialize`, `SBDebugger.Create`)
|
|
- Keep one process-level session object during command/repl lifetime
|
|
- Use synchronous mode (`SetAsync(False)`) for deterministic command behavior
|
|
- Destroy debugger cleanly on session close
|
|
|
|
## Data Model
|
|
|
|
All core operations return plain dictionaries:
|
|
- process info (`pid`, `state`, `num_threads`)
|
|
- stop info (`reason`, `description`, `module`, `function`, `frame`)
|
|
- frame info (`function`, `file`, `line`, `address`)
|
|
- breakpoints (`id`, `locations`, `condition`)
|
|
- expression result (`type`, `value`, `summary`, `error`)
|
|
|
|
This maps directly to `--json` output mode.
|
|
|
|
## Current Limitations
|
|
|
|
- No advanced watchpoint support yet
|
|
- Memory search is a naive byte scan over fetched range
|
|
- Multi-target workflows are not implemented yet (single active target/session)
|