blob: 6d30062a2578fce54fb3e64f9cab71afa39ea174 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
|
# Ghidra bridge scripts
Two Java scripts run inside Ghidra (GUI or headless):
| Script | Direction | Purpose |
|---|---|---|
| `ExportForGreenEye.java` | Ghidra → risc-v-gsp-rm-debugger | write `addr<TAB>name<TAB>size` for every function |
| `MarkGreenEyeHits.java` | risc-v-gsp-rm-debugger → Ghidra | annotate functions with runtime hit counts + bookmarks |
Typical loop:
```bash
# 1. Export the decompiler's view of the firmware.
$GHIDRA/support/analyzeHeadless <proj_dir> <project> -process <program> -noanalysis \
-scriptPath debug/risc-v-gsp-rm-vm/ghidra \
-postScript ExportForGreenEye.java /tmp/ghidra-symbols.tsv
# 2. Build a risc-v-gsp-rm-debugger symbol DB from it.
./risc-v-gsp-rm-debugger.sh symbols export --index /tmp/ghidra-symbols.tsv -o symbols.json
# 3. Run a session, then export its hits.
python3 - <<'PY'
from greeneye.ghidra import export_hits
from greeneye.symbols import SymbolDB
db = SymbolDB(); db.load_json("symbols.json")
print(export_hits("dumps/<run>", "/tmp/greeneye_hits.tsv", symbols=db))
PY
# 4. Annotate the project.
$GHIDRA/support/analyzeHeadless <proj_dir> <project> -process <program> -noanalysis \
-scriptPath debug/risc-v-gsp-rm-vm/ghidra \
-postScript MarkGreenEyeHits.java /tmp/greeneye_hits.tsv
```
Address convention: the exported `addr` is the **program address** (what
Ghidra shows). `./risc-v-gsp-rm-debugger.sh symbols export --index` understands the 3-column
form and keeps file/program addresses separate when present (4/7-column
project indexes). Pass `--bias` only when the exported TSV does not already
carry runtime addresses.
|
