Add "disasm sym_info.py" script (#2054)

* Add "disasm sym_info.py" script for investigating the expected memory layout

* review

* add friendly error if context.csv is missing

* rom as hexadecimal

* require spimdisasm 1.28.1 (technically not *required* but may as well)

* some fixups on symbol lookup logic
This commit is contained in:
Dragorn421 2024-08-19 23:58:41 +02:00 committed by GitHub
parent d191e8714e
commit ebbc820ef6
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 239 additions and 2 deletions

View File

@ -18,4 +18,4 @@ toml
mapfile-parser>=2.3.5,<3.0.0 mapfile-parser>=2.3.5,<3.0.0
pyelftools==0.30 pyelftools==0.30
rabbitizer>=1.0.0,<2.0.0 rabbitizer>=1.0.0,<2.0.0
spimdisasm>=1.21.0,<2.0.0 spimdisasm>=1.28.1,<2.0.0

View File

@ -85,6 +85,11 @@ def main():
args = parser.parse_args() args = parser.parse_args()
if spimdisasm.__version_info__ < (1, 28, 1):
print(f"Error: spimdisasm>=1.28.1 is required (you have {spimdisasm.__version__})")
print("Hint: run `make setup` to update the venv.")
exit(1)
context = spimdisasm.common.Context() context = spimdisasm.common.Context()
context.parseArgs(args) context.parseArgs(args)
context.changeGlobalSegmentRanges(0x00000000, 0x01000000, 0x8000000, 0x81000000) context.changeGlobalSegmentRanges(0x00000000, 0x01000000, 0x8000000, 0x81000000)
@ -117,9 +122,12 @@ def main():
print() print()
print("Analyzing done.") print("Analyzing done.")
print("Writing disassembled sections...")
output_dir: Path = args.output_dir output_dir: Path = args.output_dir
output_dir.mkdir(parents=True, exist_ok=True) output_dir.mkdir(parents=True, exist_ok=True)
context.saveContextToFile(output_dir / "context.csv")
print("Writing disassembled sections...")
for i, file_splits in enumerate(all_file_splits): for i, file_splits in enumerate(all_file_splits):
f = i / len(all_file_splits) f = i / len(all_file_splits)
spimdisasm.common.Utils.printQuietless( spimdisasm.common.Utils.printQuietless(

229
tools/disasm/sym_info.py Executable file
View File

@ -0,0 +1,229 @@
#!/usr/bin/env python3
# SPDX-FileCopyrightText: © 2024 ZeldaRET
# SPDX-License-Identifier: CC0-1.0
import argparse
import csv
import dataclasses
from pathlib import Path
from typing import Optional
@dataclasses.dataclass
class Sym:
name: str
value: int
type: Optional[str]
size: Optional[int]
vrom: int
@dataclasses.dataclass
class FileSection:
file: str
section: str
syms: list[Sym]
vma_start: int
LABELS_TYPES = {"@branchlabel", "@jumptablelabel"}
def main():
parser = argparse.ArgumentParser(
description="Print informations on a symbol/address"
" (and possibly surrounding symbols)"
" from the spimdisasm disassembly context saved in"
" expected/build/VERSION/context.csv"
)
parser.add_argument("sym_or_vma")
default_version = "ntsc-1.2"
parser.add_argument(
"--version",
"-v",
default=default_version,
help=f"oot version (default: {default_version})",
)
parser.add_argument(
"--around",
"-n",
type=int,
default=0,
help="how many symbols to show around the target (at least)",
)
parser.add_argument(
"--range",
"-r",
type=lambda v: int(v, 0),
default=0,
help="show symbols within this range around the target (at least)",
)
parser.add_argument(
"--file",
"-f",
action="store_true",
help="show symbols within the same file and section as the target (at least)",
)
parser.add_argument(
"--labels",
"-l",
action="store_true",
help="also show branch and jump table labels symbols",
)
args = parser.parse_args()
sym_or_vma = args.sym_or_vma
if "_" in sym_or_vma:
# special case to avoid parsing e.g. `D_80123456` as hexadecimal 0xD80123456
sym_or_vma_is_sym = True
else:
try:
target_sym_name = None
target_vma = int(sym_or_vma, 16)
sym_or_vma_is_sym = False
except ValueError:
sym_or_vma_is_sym = True
if sym_or_vma_is_sym:
target_sym_name = sym_or_vma
target_vma = None
syms_by_section_by_file = dict[str, dict[str, list[Sym]]]()
context_csv_p = Path(f"expected/build/{args.version}/context.csv")
if not context_csv_p.exists():
print(f"Context file does not exist: {context_csv_p}")
print(f"Hint: run `make VERSION={args.version} disasm`")
exit(1)
with context_csv_p.open() as f:
for e in csv.DictReader(f):
if e["category"] != "symbol":
continue
sym_name = e["getName"]
sym_value = e["address"]
sym_type = e["getType"]
sym_size = e["getSize"]
sym_vrom = e["getVrom"]
sym_section = e["sectionType"]
sym_file = e["parentFileName"]
if sym_file == "None":
sym_file = None
if not sym_section or not sym_file:
continue
sym_value_int = int(sym_value, 0)
sym_size_int = int(sym_size, 0) if sym_size else None
sym_vrom_int = int(sym_vrom, 0)
syms_by_section_by_file.setdefault(sym_file, dict()).setdefault(
sym_section, list()
).append(
Sym(
sym_name,
sym_value_int,
sym_type if sym_type else None,
sym_size_int,
sym_vrom_int,
)
)
if sym_name == target_sym_name:
target_vma = sym_value_int
if target_vma is None:
parser.error(f"No symbol '{target_sym_name}'")
else:
if target_sym_name is not None:
print(f"{target_sym_name} = 0x{target_vma:08X}")
del target_sym_name
filesections = list[FileSection]()
for file, syms_by_section in syms_by_section_by_file.items():
for section, syms in syms_by_section.items():
syms.sort(key=lambda sym: sym.value)
vma_start = syms[0].value
filesections.append(FileSection(file, section, syms, vma_start))
filesections.sort(key=lambda fs: fs.vma_start)
def get_first_print_sym():
prev_syms = list[Sym]()
for fs in filesections:
for sym in fs.syms:
if not args.labels and sym.type in LABELS_TYPES:
continue
if target_vma < sym.value:
return prev_syms[0]
prev_syms.append(sym)
while (
len(prev_syms) - 1 > args.around
and prev_syms[0].value < target_vma - args.range
):
prev_syms.pop(0)
first_print_sym = get_first_print_sym()
def get_last_print_sym():
min_skip_count = args.around
for fs in filesections:
for sym in fs.syms:
if not args.labels and sym.type in LABELS_TYPES:
continue
if target_vma <= sym.value:
min_skip_count -= 1
if min_skip_count < 0 and sym.value >= args.range + target_vma:
return sym
last_print_sym = get_last_print_sym()
is_near_target = False
indent = " " * 4
for i_fs, fs in enumerate(filesections):
fs_printed = False
is_first_fs_sym = True
fs_printed_end_ellipsis = False
for sym in fs.syms:
if not args.labels and sym.type in LABELS_TYPES:
continue
if sym == first_print_sym:
is_near_target = True
print_sym = is_near_target or (
args.file
and fs.vma_start <= target_vma
and (
target_vma < filesections[i_fs + 1].vma_start
if i_fs + 1 < len(filesections)
else True
)
)
if not print_sym and fs_printed:
if not fs_printed_end_ellipsis:
print(f"{indent}...")
fs_printed_end_ellipsis = True
if print_sym:
if not fs_printed:
print(fs.file, fs.section)
fs_printed = True
if not is_first_fs_sym:
print(f"{indent}...")
print(
f"{indent}{sym.name} 0x{sym.value:X} ROM:0x{sym.vrom:X}"
+ (f" ({sym.type})" if sym.type else "")
+ (f" (sz=0x{sym.size:X})" if sym.size else "")
)
is_first_fs_sym = False
if sym == last_print_sym:
is_near_target = False
if __name__ == "__main__":
main()