This is an archive of the discontinued LLVM Phabricator instance.

Differential D90665

[crashlog] Modularize parser
ClosedPublic

Authored by JDevlieghere on Nov 2 2020, 8:39 PM.

Details

Reviewers
aprantl
kastiglione
clayborg
Commits
rG16dd69347dfc: [crashlog] Modularize parser
Summary

Instead of parsing the crashlog in one big loop, use methods that correspond to the different parsing modes.

Diff Detail

Event Timeline

JDevlieghere requested review of this revision.Nov 2 2020, 8:39 PM
JDevlieghere created this revision.
JDevlieghere added a parent revision: D90664: [crashlog] Move crash log parsing into its own class.
clayborg accepted this revision.Nov 2 2020, 8:57 PM
This revision is now accepted and ready to land.Nov 2 2020, 8:57 PM
Closed by commit rG16dd69347dfc: [crashlog] Modularize parser (authored by JDevlieghere). · Explain WhyNov 3 2020, 10:21 AM
This revision was automatically updated to reflect the committed changes.
JDevlieghere added a commit: rG16dd69347dfc: [crashlog] Modularize parser.
Herald added a project: Restricted Project. · View Herald TranscriptNov 3 2020, 10:21 AM

Revision Contents

PathSize
lldb/
examples/
python/
288 lines

Diff 302610

lldb/examples/python/crashlog.py

Show First 20 Lines • Show All 408 Lines • ▼ Show 20 Lines image_regex_uuid = re.compile(r'(0x[0-9a-fA-F]+)' # img_lo
r'(<([-0-9a-fA-F]+)>\s+)?' # img_uuid r'(<([-0-9a-fA-F]+)>\s+)?' # img_uuid
r'(/.*)' # img_path r'(/.*)' # img_path
) )
def __init__(self, path, verbose): def __init__(self, path, verbose):
self.path = os.path.expanduser(path) self.path = os.path.expanduser(path)
self.verbose = verbose self.verbose = verbose
self.parse_mode = CrashLogParseMode.NORMAL
self.thread = None self.thread = None
self.app_specific_backtrace = False self.app_specific_backtrace = False
self.crashlog = CrashLog(self.path, self.verbose) self.crashlog = CrashLog(self.path, self.verbose)
self.parse_mode = CrashLogParseMode.NORMAL
self.parsers = {
CrashLogParseMode.NORMAL : self.parse_normal,
CrashLogParseMode.THREAD : self.parse_thread,
CrashLogParseMode.IMAGES : self.parse_images,
CrashLogParseMode.THREGS : self.parse_thread_registers,
CrashLogParseMode.SYSTEM : self.parse_system,
CrashLogParseMode.INSTRS : self.parse_instructions,
}
def parse(self): def parse(self):
with open(self.path,'r') as f: with open(self.path,'r') as f:
lines = f.read().splitlines() lines = f.read().splitlines()
for line in lines: for line in lines:
line_len = len(line) line_len = len(line)
if line_len == 0: if line_len == 0:
Show All 11 Lines def parse(self):
self.crashlog.threads.append(self.thread) self.crashlog.threads.append(self.thread)
self.thread = None self.thread = None
else: else:
# only append an extra empty line if the previous line # only append an extra empty line if the previous line
# in the info_lines wasn't empty # in the info_lines wasn't empty
if len(self.crashlog.info_lines) > 0 and len(self.crashlog.info_lines[-1]): if len(self.crashlog.info_lines) > 0 and len(self.crashlog.info_lines[-1]):
self.crashlog.info_lines.append(line) self.crashlog.info_lines.append(line)
self.parse_mode = CrashLogParseMode.NORMAL self.parse_mode = CrashLogParseMode.NORMAL
elif self.parse_mode == CrashLogParseMode.NORMAL: else:
self.parsers[self.parse_mode](line)
return self.crashlog
def parse_normal(self, line):
if line.startswith('Process:'): if line.startswith('Process:'):
(self.crashlog.process_name, pid_with_brackets) = line[ (self.crashlog.process_name, pid_with_brackets) = line[
8:].strip().split(' [') 8:].strip().split(' [')
self.crashlog.process_id = pid_with_brackets.strip('[]') self.crashlog.process_id = pid_with_brackets.strip('[]')
elif line.startswith('Path:'): elif line.startswith('Path:'):
self.crashlog.process_path = line[5:].strip() self.crashlog.process_path = line[5:].strip()
elif line.startswith('Identifier:'): elif line.startswith('Identifier:'):
self.crashlog.process_identifier = line[11:].strip() self.crashlog.process_identifier = line[11:].strip()
elif line.startswith('Version:'): elif line.startswith('Version:'):
version_string = line[8:].strip() version_string = line[8:].strip()
matched_pair = re.search("(.+)\((.+)\)", version_string) matched_pair = re.search("(.+)\((.+)\)", version_string)
if matched_pair: if matched_pair:
self.crashlog.process_version = matched_pair.group(1) self.crashlog.process_version = matched_pair.group(1)
self.crashlog.process_compatability_version = matched_pair.group( self.crashlog.process_compatability_version = matched_pair.group(
2) 2)
else: else:
self.crashlog.process = version_string self.crashlog.process = version_string
self.crashlog.process_compatability_version = version_string self.crashlog.process_compatability_version = version_string
elif self.parent_process_regex.search(line): elif self.parent_process_regex.search(line):
parent_process_match = self.parent_process_regex.search( parent_process_match = self.parent_process_regex.search(
line) line)
self.crashlog.parent_process_name = parent_process_match.group(1) self.crashlog.parent_process_name = parent_process_match.group(1)
self.crashlog.parent_process_id = parent_process_match.group(2) self.crashlog.parent_process_id = parent_process_match.group(2)
elif line.startswith('Exception Type:'): elif line.startswith('Exception Type:'):
self.crashlog.thread_exception = line[15:].strip() self.crashlog.thread_exception = line[15:].strip()
continue return
elif line.startswith('Exception Codes:'): elif line.startswith('Exception Codes:'):
self.crashlog.thread_exception_data = line[16:].strip() self.crashlog.thread_exception_data = line[16:].strip()
continue return
elif line.startswith('Exception Subtype:'): # iOS elif line.startswith('Exception Subtype:'): # iOS
self.crashlog.thread_exception_data = line[18:].strip() self.crashlog.thread_exception_data = line[18:].strip()
continue return
elif line.startswith('Crashed Thread:'): elif line.startswith('Crashed Thread:'):
self.crashlog.crashed_thread_idx = int(line[15:].strip().split()[0]) self.crashlog.crashed_thread_idx = int(line[15:].strip().split()[0])
continue return
elif line.startswith('Triggered by Thread:'): # iOS elif line.startswith('Triggered by Thread:'): # iOS
self.crashlog.crashed_thread_idx = int(line[20:].strip().split()[0]) self.crashlog.crashed_thread_idx = int(line[20:].strip().split()[0])
continue return
elif line.startswith('Report Version:'): elif line.startswith('Report Version:'):
self.crashlog.version = int(line[15:].strip()) self.crashlog.version = int(line[15:].strip())
continue return
elif line.startswith('System Profile:'): elif line.startswith('System Profile:'):
self.parse_mode = CrashLogParseMode.SYSTEM self.parse_mode = CrashLogParseMode.SYSTEM
continue return
elif (line.startswith('Interval Since Last Report:') or elif (line.startswith('Interval Since Last Report:') or
line.startswith('Crashes Since Last Report:') or line.startswith('Crashes Since Last Report:') or
line.startswith('Per-App Interval Since Last Report:') or line.startswith('Per-App Interval Since Last Report:') or
line.startswith('Per-App Crashes Since Last Report:') or line.startswith('Per-App Crashes Since Last Report:') or
line.startswith('Sleep/Wake UUID:') or line.startswith('Sleep/Wake UUID:') or
line.startswith('Anonymous UUID:')): line.startswith('Anonymous UUID:')):
# ignore these # ignore these
continue return
elif line.startswith('Thread'): elif line.startswith('Thread'):
thread_state_match = self.thread_state_regex.search(line) thread_state_match = self.thread_state_regex.search(line)
if thread_state_match: if thread_state_match:
self.app_specific_backtrace = False self.app_specific_backtrace = False
thread_state_match = self.thread_regex.search(line) thread_state_match = self.thread_regex.search(line)
thread_idx = int(thread_state_match.group(1)) thread_idx = int(thread_state_match.group(1))
self.parse_mode = CrashLogParseMode.THREGS self.parse_mode = CrashLogParseMode.THREGS
self.thread = self.crashlog.threads[thread_idx] self.thread = self.crashlog.threads[thread_idx]
continue return
thread_insts_match = self.thread_instrs_regex.search(line) thread_insts_match = self.thread_instrs_regex.search(line)
if thread_insts_match: if thread_insts_match:
self.parse_mode = CrashLogParseMode.INSTRS self.parse_mode = CrashLogParseMode.INSTRS
continue return
thread_match = self.thread_regex.search(line) thread_match = self.thread_regex.search(line)
if thread_match: if thread_match:
self.app_specific_backtrace = False self.app_specific_backtrace = False
self.parse_mode = CrashLogParseMode.THREAD self.parse_mode = CrashLogParseMode.THREAD
thread_idx = int(thread_match.group(1)) thread_idx = int(thread_match.group(1))
self.thread = self.crashlog.Thread(thread_idx, False) self.thread = self.crashlog.Thread(thread_idx, False)
continue return
continue return
elif line.startswith('Binary Images:'): elif line.startswith('Binary Images:'):
self.parse_mode = CrashLogParseMode.IMAGES self.parse_mode = CrashLogParseMode.IMAGES
continue return
elif line.startswith('Application Specific Backtrace'): elif line.startswith('Application Specific Backtrace'):
app_backtrace_match = self.app_backtrace_regex.search(line) app_backtrace_match = self.app_backtrace_regex.search(line)
if app_backtrace_match: if app_backtrace_match:
self.parse_mode = CrashLogParseMode.THREAD self.parse_mode = CrashLogParseMode.THREAD
self.app_specific_backtrace = True self.app_specific_backtrace = True
idx = int(app_backtrace_match.group(1)) idx = int(app_backtrace_match.group(1))
self.thread = self.crashlog.Thread(idx, True) self.thread = self.crashlog.Thread(idx, True)
elif line.startswith('Last Exception Backtrace:'): # iOS elif line.startswith('Last Exception Backtrace:'): # iOS
self.parse_mode = CrashLogParseMode.THREAD self.parse_mode = CrashLogParseMode.THREAD
self.app_specific_backtrace = True self.app_specific_backtrace = True
idx = 1 idx = 1
self.thread = self.crashlog.Thread(idx, True) self.thread = self.crashlog.Thread(idx, True)
self.crashlog.info_lines.append(line.strip()) self.crashlog.info_lines.append(line.strip())
elif self.parse_mode == CrashLogParseMode.THREAD:
def parse_thread(self, line):
if line.startswith('Thread'): if line.startswith('Thread'):
continue return
if self.null_frame_regex.search(line): if self.null_frame_regex.search(line):
print('warning: thread parser ignored null-frame: "%s"' % line) print('warning: thread parser ignored null-frame: "%s"' % line)
continue return
frame_match = self.frame_regex.search(line) frame_match = self.frame_regex.search(line)
if frame_match: if frame_match:
(frame_id, frame_img_name, _, frame_img_version, _, (frame_id, frame_img_name, _, frame_img_version, _,
frame_addr, frame_ofs) = frame_match.groups() frame_addr, frame_ofs) = frame_match.groups()
ident = frame_img_name ident = frame_img_name
self.thread.add_ident(ident) self.thread.add_ident(ident)
if ident not in self.crashlog.idents: if ident not in self.crashlog.idents:
self.crashlog.idents.append(ident) self.crashlog.idents.append(ident)
self.thread.frames.append(self.crashlog.Frame(int(frame_id), int( self.thread.frames.append(self.crashlog.Frame(int(frame_id), int(
frame_addr, 0), frame_ofs)) frame_addr, 0), frame_ofs))
else: else:
print('error: frame regex failed for line: "%s"' % line) print('error: frame regex failed for line: "%s"' % line)
elif self.parse_mode == CrashLogParseMode.IMAGES:
def parse_images(self, line):
image_match = self.image_regex_uuid.search(line) image_match = self.image_regex_uuid.search(line)
if image_match: if image_match:
(img_lo, img_hi, img_name, _, img_version, _, (img_lo, img_hi, img_name, _, img_version, _,
_, img_uuid, img_path) = image_match.groups() _, img_uuid, img_path) = image_match.groups()
image = self.crashlog.DarwinImage(int(img_lo, 0), int(img_hi, 0), image = self.crashlog.DarwinImage(int(img_lo, 0), int(img_hi, 0),
img_name.strip(), img_name.strip(),
img_version.strip() img_version.strip()
if img_version else "", if img_version else "",
uuid.UUID(img_uuid), img_path, uuid.UUID(img_uuid), img_path,
self.verbose) self.verbose)
self.crashlog.images.append(image) self.crashlog.images.append(image)
else: else:
print("error: image regex failed for: %s" % line) print("error: image regex failed for: %s" % line)
elif self.parse_mode == CrashLogParseMode.THREGS:
def parse_thread_registers(self, line):
stripped_line = line.strip() stripped_line = line.strip()
# "r12: 0x00007fff6b5939c8 r13: 0x0000000007000006 r14: 0x0000000000002a03 r15: 0x0000000000000c00" # "r12: 0x00007fff6b5939c8 r13: 0x0000000007000006 r14: 0x0000000000002a03 r15: 0x0000000000000c00"
reg_values = re.findall( reg_values = re.findall(
'([a-zA-Z0-9]+: 0[Xx][0-9a-fA-F]+) *', stripped_line) '([a-zA-Z0-9]+: 0[Xx][0-9a-fA-F]+) *', stripped_line)
for reg_value in reg_values: for reg_value in reg_values:
(reg, value) = reg_value.split(': ') (reg, value) = reg_value.split(': ')
self.thread.registers[reg.strip()] = int(value, 0) self.thread.registers[reg.strip()] = int(value, 0)
elif self.parse_mode == CrashLogParseMode.SYSTEM:
self.crashlog.system_profile.append(line)
elif self.parse_mode == CrashLogParseMode.INSTRS:
pass
return self.crashlog def parse_system(self, line):
self.crashlog.system_profile.append(line)
def parse_instructions(self, line):
pass
def usage(): def usage():
print("Usage: lldb-symbolicate.py [-n name] executable-image") print("Usage: lldb-symbolicate.py [-n name] executable-image")
sys.exit(0) sys.exit(0)
class Interactive(cmd.Cmd): class Interactive(cmd.Cmd):
▲ Show 20 Lines • Show All 461 Lines • Show Last 20 Lines