blob: dc0c62bd375028af5c28ec8585a7604739373c97 [file]
import json
import re
import gdbremote_testcase
from lldbsuite.test.decorators import *
from lldbsuite.test.lldbtest import *
from lldbsuite.test import lldbutil
@requireThreadSupport
@skipIfMTE # MTE security transition shims restrict socket operations.
class TestGdbRemoteThreadsInStopReply(gdbremote_testcase.GdbRemoteTestCaseBase):
ENABLE_THREADS_IN_STOP_REPLY_ENTRIES = [
"read packet: $QListThreadsInStopReply#21",
"send packet: $OK#00",
]
def gather_stop_reply_fields(self, thread_count, field_names):
context, threads = self.launch_with_threads(thread_count)
key_vals_text = context.get("stop_reply_kv")
self.assertIsNotNone(key_vals_text)
self.reset_test_sequence()
self.add_register_info_collection_packets()
self.add_process_info_collection_packets()
context = self.expect_gdbremote_sequence()
self.assertIsNotNone(context)
hw_info = self.parse_hw_info(context)
# Parse the stop reply contents.
kv_dict = self.parse_key_val_dict(key_vals_text)
result = dict()
result["pc_register"] = hw_info["pc_register"]
result["little_endian"] = hw_info["little_endian"]
result["ptrsize"] = hw_info["ptrsize"]
for key_field in field_names:
result[key_field] = kv_dict.get(key_field)
return result
def gather_stop_reply_threads(self, thread_count):
# Pull out threads from stop response.
stop_reply_threads_text = self.gather_stop_reply_fields(
thread_count, ["threads"]
)["threads"]
if stop_reply_threads_text:
return [
int(thread_id, 16) for thread_id in stop_reply_threads_text.split(",")
]
else:
return []
def gather_stop_reply_pcs(self, thread_count):
results = self.gather_stop_reply_fields(thread_count, ["threads", "thread-pcs"])
if not results:
return []
threads_text = results["threads"]
pcs_text = results["thread-pcs"]
thread_ids = threads_text.split(",")
pcs = pcs_text.split(",")
self.assertEqual(len(thread_ids), len(pcs))
thread_pcs = dict()
for i in range(0, len(pcs)):
thread_pcs[int(thread_ids[i], 16)] = pcs[i]
result = dict()
result["thread_pcs"] = thread_pcs
result["pc_register"] = results["pc_register"]
result["little_endian"] = results["little_endian"]
return result
def switch_endian(self, egg):
return "".join(reversed(re.findall("..", egg)))
def parse_hw_info(self, context):
self.assertIsNotNone(context)
process_info = self.parse_process_info_response(context)
endian = process_info.get("endian")
reg_info = self.parse_register_info_packets(context)
(pc_lldb_reg_index, pc_reg_info) = self.find_pc_reg_info(reg_info)
hw_info = dict()
hw_info["pc_register"] = pc_lldb_reg_index
hw_info["little_endian"] = endian == "little"
hw_info["ptrsize"] = int(process_info.get("ptrsize", 0))
return hw_info
def gather_threads_info(self):
"""The parsed jThreadsInfo reply: one dict per thread."""
self.reset_test_sequence()
self.test_sequence.add_log_lines(
[
"read packet: $jThreadsInfo#c1",
{
"direction": "send",
"regex": r"^\$(.*)#[0-9a-fA-F]{2}$",
"capture": {1: "threads_info"},
},
],
True,
)
context = self.expect_gdbremote_sequence()
self.assertIsNotNone(context)
threads_info = context.get("threads_info")
# A literal '}' is escaped as '}]' on the wire.
return json.loads(re.sub(r"}]", "}", threads_info))
def gather_threads_info_pcs(self, pc_register, little_endian):
jthreads_info = self.gather_threads_info()
register = str(pc_register)
thread_pcs = dict()
for thread_info in jthreads_info:
tid = thread_info["tid"]
pc = thread_info["registers"][register]
thread_pcs[tid] = self.switch_endian(pc) if little_endian else pc
return thread_pcs
def test_QListThreadsInStopReply_supported(self):
self.build()
self.set_inferior_startup_launch()
procs = self.prep_debug_monitor_and_inferior()
self.test_sequence.add_log_lines(
self.ENABLE_THREADS_IN_STOP_REPLY_ENTRIES, True
)
context = self.expect_gdbremote_sequence()
self.assertIsNotNone(context)
@skipIfNetBSD
@expectedFailureAll(oslist=["windows"]) # Extra threads present
def test_stop_reply_reports_multiple_threads(self):
self.build()
self.set_inferior_startup_launch()
# Gather threads from stop notification when QThreadsInStopReply is
# enabled.
self.test_sequence.add_log_lines(
self.ENABLE_THREADS_IN_STOP_REPLY_ENTRIES, True
)
stop_reply_threads = self.gather_stop_reply_threads(5)
self.assertEqual(len(stop_reply_threads), 5)
@skipIfNetBSD
def test_no_QListThreadsInStopReply_supplies_no_threads(self):
self.build()
self.set_inferior_startup_launch()
# Gather threads from stop notification when QThreadsInStopReply is not
# enabled.
stop_reply_threads = self.gather_stop_reply_threads(5)
self.assertEqual(len(stop_reply_threads), 0)
@skipIfNetBSD
def test_stop_reply_reports_correct_threads(self):
self.build()
self.set_inferior_startup_launch()
# Gather threads from stop notification when QThreadsInStopReply is
# enabled.
thread_count = 5
self.test_sequence.add_log_lines(
self.ENABLE_THREADS_IN_STOP_REPLY_ENTRIES, True
)
stop_reply_threads = self.gather_stop_reply_threads(thread_count)
# Gather threads from q{f,s}ThreadInfo.
self.reset_test_sequence()
self.add_threadinfo_collection_packets()
context = self.expect_gdbremote_sequence()
self.assertIsNotNone(context)
threads = self.parse_threadinfo_packets(context)
self.assertIsNotNone(threads)
self.assertGreaterEqual(len(threads), thread_count)
# Ensure each thread in q{f,s}ThreadInfo appears in stop reply threads
for tid in threads:
self.assertIn(tid, stop_reply_threads)
@add_test_categories(["debugserver"])
def test_stop_reply_expedites_frame_pointer_backchain(self):
"""The stop reply expedites at most the two innermost backchain
entries, one `memory:` entry each of 2 * ptrsize bytes."""
self.build()
self.set_inferior_startup_launch()
results = self.gather_stop_reply_fields(1, ["memory"])
memory = results["memory"]
self.assertIsNotNone(memory, "the stop reply carried no memory: entry")
# parse_key_val_dict promotes a repeated key to a list.
entries = memory if isinstance(memory, list) else [memory]
# A backchain entry is 2 pointers at $fp: the previous FP and PC.
backchain_entry_size = 2 * results["ptrsize"]
for entry in entries:
addr, sep, hex_bytes = entry.partition("=")
self.assertEqual(sep, "=", "malformed memory entry %r" % entry)
self.assertEqual(
len(hex_bytes),
2 * backchain_entry_size,
"memory:%s= should carry %d bytes (2 * ptrsize), got %d: %r"
% (addr, backchain_entry_size, len(hex_bytes) // 2, entry),
)
# The stop reply caps the backchain at 2 entries. How many it actually
# walks depends on where the backchain terminates.
self.assertLessEqual(len(entries), 2, "expected at most 2 backchain entries")
# Frame 0's stack memory is expedited only by a debugserver built from this
# tree; an older system debugserver sends the backchain alone.
@skipIfOutOfTreeDebugserver
@add_test_categories(["debugserver"])
def test_threads_info_expedites_stopped_frame_stack(self):
"""jThreadsInfo expedites every thread's frame pointer backchain, plus
frame 0's stack memory for the stopped thread only."""
self.build()
self.set_inferior_startup_launch()
results = self.gather_stop_reply_fields(5, ["thread"])
backchain_entry_size = 2 * results["ptrsize"]
stopped_tid = int(results["thread"], 16)
# Chunk sizes tell the two apart: a backchain entry is 2 pointers,
# frame 0's stack memory is a larger range.
def classify(thread_info):
sizes = [len(m["bytes"]) // 2 for m in thread_info.get("memory", [])]
stack_backchain = [n for n in sizes if n == backchain_entry_size]
frame_0_stack_memory = [n for n in sizes if n != backchain_entry_size]
return sizes, stack_backchain, frame_0_stack_memory
saw_stopped = False
for thread_info in self.gather_threads_info():
sizes, stack_backchain, frame_0_stack_memory = classify(thread_info)
where = "thread %x, chunk sizes %s" % (thread_info["tid"], sizes)
self.assertGreaterEqual(
len(stack_backchain), 1, "no backchain entry for %s" % where
)
if thread_info["tid"] == stopped_tid:
saw_stopped = True
# One chunk, or two when $fp is usable.
self.assertIn(
len(frame_0_stack_memory),
(1, 2),
"no frame 0 stack memory for stopped %s" % where,
)
else:
self.assertEqual(
frame_0_stack_memory,
[],
"unexpected frame 0 stack memory for %s" % where,
)
self.assertTrue(saw_stopped, "no jThreadsInfo entry for the stopped thread")
@skipIfNetBSD
@skipIfWindows # Flaky on Windows
def test_stop_reply_contains_thread_pcs(self):
self.build()
self.set_inferior_startup_launch()
thread_count = 5
self.test_sequence.add_log_lines(
self.ENABLE_THREADS_IN_STOP_REPLY_ENTRIES, True
)
results = self.gather_stop_reply_pcs(thread_count)
stop_reply_pcs = results["thread_pcs"]
pc_register = results["pc_register"]
little_endian = results["little_endian"]
self.assertGreaterEqual(len(stop_reply_pcs), thread_count)
threads_info_pcs = self.gather_threads_info_pcs(pc_register, little_endian)
self.assertEqual(len(threads_info_pcs), len(stop_reply_pcs))
for thread_id in stop_reply_pcs:
self.assertIn(thread_id, threads_info_pcs)
self.assertEqual(
int(stop_reply_pcs[thread_id], 16), int(threads_info_pcs[thread_id], 16)
)