mirror of
https://gitlab.isc.org/isc-projects/bind9
synced 2025-08-22 10:10:06 +00:00
7f0790c82f
util/parse_tsan.py builds tables of mutexes, threads, and pointers it
finds in the TSAN report provided to it as a command-line argument and
then replaces all mentions of each of these entities so that they are
numbered sequentially in the processed report. For example, this line:
Cycle in lock order graph: M0 (...) => M5 (...) => M9 (...) => M0
is expected to become:
Cycle in lock order graph: M1 (...) => M2 (...) => M3 (...) => M1
Problems arise when the gaps between mutex/thread identifiers present on
a single line are smaller than the total number of mutexes/threads found
by the script so far. For example, the following line:
Cycle in lock order graph: M0 (...) => M1 (...) => M2 (...) => M0
first gets turned into:
Cycle in lock order graph: M1 (...) => M1 (...) => M2 (...) => M1
and then into:
Cycle in lock order graph: M2 (...) => M2 (...) => M2 (...) => M2
In other words, lines like this become garbled due to information loss.
The problem stems from the fact that the numbering scheme the script
uses for identifying mutexes and threads is exactly the same as the one
used by TSAN itself. Update util/parse_tsan.py so that it uses
zero-padded numbers instead, making the "overlapping" demonstrated above
impossible.
129 lines
3.8 KiB
Python
Executable File
129 lines
3.8 KiB
Python
Executable File
#!/usr/bin/env python3
|
|
############################################################################
|
|
# Copyright (C) Internet Systems Consortium, Inc. ("ISC")
|
|
#
|
|
# SPDX-License-Identifier: MPL-2.0
|
|
#
|
|
# This Source Code Form is subject to the terms of the Mozilla Public
|
|
# License, v. 2.0. If a copy of the MPL was not distributed with this
|
|
# file, you can obtain one at https://mozilla.org/MPL/2.0/.
|
|
#
|
|
# See the COPYRIGHT file distributed with this work for additional
|
|
# information regarding copyright ownership.
|
|
############################################################################
|
|
|
|
"""Parse the ThreadSanizer reports, unify them and put them into unique dirs."""
|
|
|
|
import sys
|
|
import os
|
|
import os.path
|
|
import re
|
|
from hashlib import sha256
|
|
|
|
|
|
class State:
|
|
"""Class that holds state of the TSAN parser."""
|
|
|
|
# pylint: disable=too-many-instance-attributes
|
|
# pylint: disable=too-few-public-methods
|
|
|
|
inside = False
|
|
block = ""
|
|
last_line = None
|
|
|
|
mutexes = {}
|
|
m_index = 1
|
|
threads = {}
|
|
t_index = 1
|
|
pointers = {}
|
|
p_index = 1
|
|
|
|
def __init__(self):
|
|
self.reset()
|
|
|
|
def reset(self):
|
|
"""Reset the object to initial state"""
|
|
|
|
self.inside = False
|
|
self.block = ""
|
|
|
|
self.mutexes = {}
|
|
self.threads = {}
|
|
self.pointers = {}
|
|
self.pointers["0x000000000000"] = 0
|
|
|
|
self.m_index = 1
|
|
self.t_index = 1
|
|
self.p_index = 1
|
|
|
|
|
|
TOP = os.path.dirname(os.path.dirname(os.path.realpath(__file__)))
|
|
|
|
OUT = os.path.join(TOP, "tsan")
|
|
|
|
if not os.path.isdir(OUT):
|
|
os.mkdir(OUT)
|
|
|
|
# Regular Expressions
|
|
MUTEX = re.compile(r"M\d+")
|
|
THREAD = re.compile(r"T\d+")
|
|
STACK = re.compile(r"\s\(\S+\+0x\S+\)")
|
|
POINTER = re.compile(r"0x[0-9a-f]+")
|
|
PID = re.compile(r"\(pid=\d+,?\)")
|
|
TID = re.compile(r"tid=\d+,?\s*")
|
|
WORKER = re.compile(r"\s+'(isc-worker|isc-net-)\d+'")
|
|
PATH = re.compile(TOP + "/")
|
|
|
|
S = State()
|
|
|
|
with open(sys.argv[1], "r", encoding="utf-8") as f:
|
|
for line in f.readlines():
|
|
if line == "==================\n":
|
|
if not S.inside:
|
|
S.inside = True
|
|
else:
|
|
DNAME = sha256(S.last_line.encode("utf-8")).hexdigest()
|
|
DNAME = os.path.join(OUT, DNAME)
|
|
if not os.path.isdir(DNAME):
|
|
os.mkdir(DNAME)
|
|
FNAME = sha256(S.block.encode("utf-8")).hexdigest() + ".txt"
|
|
FNAME = os.path.join(DNAME, FNAME)
|
|
if not os.path.isfile(FNAME):
|
|
with open(FNAME, "w", encoding="utf-8") as w:
|
|
w.write(S.block)
|
|
S.reset()
|
|
else:
|
|
for m in MUTEX.finditer(line):
|
|
k = m.group()
|
|
if k not in S.mutexes:
|
|
S.mutexes[k] = S.m_index
|
|
S.m_index += 1
|
|
for m in THREAD.finditer(line):
|
|
k = m.group()
|
|
if k not in S.threads:
|
|
S.threads[k] = S.t_index
|
|
S.t_index += 1
|
|
for m in POINTER.finditer(line):
|
|
k = m.group()
|
|
if k not in S.pointers:
|
|
S.pointers[k] = S.p_index
|
|
S.p_index += 1
|
|
for k, v in S.mutexes.items():
|
|
r = re.compile(k)
|
|
line = r.sub("M{:04d}".format(v), line)
|
|
for k, v in S.threads.items():
|
|
r = re.compile(k)
|
|
line = r.sub("T{:04d}".format(v), line)
|
|
for k, v in S.pointers.items():
|
|
r = re.compile(k)
|
|
line = r.sub("0x{:012d}".format(v), line)
|
|
|
|
line = STACK.sub("", line)
|
|
line = PID.sub("", line)
|
|
line = TID.sub("", line)
|
|
line = WORKER.sub("", line)
|
|
line = PATH.sub("", line)
|
|
|
|
S.block += line
|
|
S.last_line = line
|