esp-idf/tools/test_apps/system/panic/test_panic_util/panic_dut.py

# SPDX-FileCopyrightText: 2022-2023 Espressif Systems (Shanghai) CO LTD
# SPDX-License-Identifier: Unlicense OR CC0-1.0
import logging
import os
import re
import subprocess
import sys
from typing import Any, Dict, List, Optional, TextIO, Union

import pexpect
from panic_utils import NoGdbProcessError, attach_logger, quote_string, sha256, verify_valid_gdb_subprocess
from pygdbmi.gdbcontroller import GdbController
from pytest_embedded_idf.app import IdfApp
from pytest_embedded_idf.dut import IdfDut
from pytest_embedded_idf.serial import IdfSerial


class PanicTestDut(IdfDut):
    BOOT_CMD_ADDR = 0x9000
    BOOT_CMD_SIZE = 0x1000
    DEFAULT_EXPECT_TIMEOUT = 10
    COREDUMP_UART_START = '================= CORE DUMP START ================='
    COREDUMP_UART_END = '================= CORE DUMP END ================='

    app: IdfApp
    serial: IdfSerial

    def __init__(self, *args: Any, **kwargs: Any) -> None:
        super().__init__(*args, **kwargs)

        self.gdbmi: Optional[GdbController] = None
        # record this since pygdbmi is using logging.debug to generate some single character mess
        self.log_level = logging.getLogger().level
        # pygdbmi is using logging.debug to generate some single character mess
        if self.log_level <= logging.DEBUG:
            logging.getLogger().setLevel(logging.INFO)

        self.coredump_output: Optional[TextIO] = None

    def close(self) -> None:
        if self.gdbmi:
            logging.info('Waiting for GDB to exit')
            self.gdbmi.exit()

        super().close()

    def revert_log_level(self) -> None:
        logging.getLogger().setLevel(self.log_level)

    @property
    def is_xtensa(self) -> bool:
        return self.target in self.XTENSA_TARGETS

    def run_test_func(self, test_func_name: str) -> None:
        self.expect_exact('Enter test name:')
        self.write(test_func_name)
        self.expect_exact('Got test name: ' + test_func_name)

    def expect_none(self, pattern, **kwargs) -> None:  # type: ignore
        """like dut.expect_all, but with an inverse logic"""
        if 'timeout' not in kwargs:
            kwargs['timeout'] = 1

        try:
            res = self.expect(pattern, **kwargs)
            raise AssertionError(f'Unexpected: {res.group().decode("utf8")}')
        except pexpect.TIMEOUT:
            pass

    def expect_backtrace(self) -> None:
        assert self.is_xtensa, 'Backtrace can be printed only on Xtensa'
        match = self.expect(r'Backtrace:( 0x[0-9a-fA-F]{8}:0x[0-9a-fA-F]{8})+(?P<corrupted> \|<-CORRUPTED)?')
        assert not match.group('corrupted')

    def expect_corrupted_backtrace(self) -> None:
        assert self.is_xtensa, 'Backtrace can be printed only on Xtensa'
        self.expect_exact('Backtrace:')
        self.expect_exact('CORRUPTED')

    def expect_stack_dump(self) -> None:
        assert not self.is_xtensa, 'Stack memory dump is only printed on RISC-V'
        self.expect_exact('Stack memory:')

    def expect_gme(self, reason: str) -> None:
        """Expect method for Guru Meditation Errors"""
        self.expect_exact(f"Guru Meditation Error: Core  0 panic'ed ({reason})")

    def expect_reg_dump(self, core: int = 0) -> None:
        """Expect method for the register dump"""
        self.expect(r'Core\s+%d register dump:' % core)

    def expect_cpu_reset(self) -> None:
        # no digital system reset for panic handling restarts (see IDF-7255)
        self.expect(r'.*rst:.*(RTC_SW_CPU_RST|SW_CPU_RESET|SW_CPU)')

    def expect_elf_sha256(self) -> None:
        """Expect method for ELF SHA256 line"""
        elf_sha256 = sha256(self.app.elf_file)
        elf_sha256_len = int(
            self.app.sdkconfig.get('CONFIG_APP_RETRIEVE_LEN_ELF_SHA', '9')
        )
        self.expect_exact('ELF file SHA256: ' + elf_sha256[0:elf_sha256_len])

    def expect_coredump(self, output_file_name: str, patterns: List[Union[str, re.Pattern]]) -> None:
        with open(output_file_name, 'r') as file:
            coredump = file.read()
            for pattern in patterns:
                if isinstance(pattern, str):
                    position = coredump.find(pattern)
                    assert position != -1, f"'{pattern}' not found in the coredump output"
                elif isinstance(pattern, re.Pattern):
                    match = pattern.findall(coredump)
                    assert match, f"'{pattern.pattern}' not found in the coredump output"
                else:
                    raise ValueError(f'Unsupported input type: {type(pattern).__name__}')

    def _call_espcoredump(
        self, extra_args: List[str], coredump_file_name: str, output_file_name: str
    ) -> None:
        # no "with" here, since we need the file to be open for later inspection by the test case
        if not self.coredump_output:
            self.coredump_output = open(output_file_name, 'w')

        espcoredump_script = os.path.join(
            os.environ['IDF_PATH'], 'components', 'espcoredump', 'espcoredump.py'
        )
        espcoredump_args = [
            sys.executable,
            espcoredump_script,
            'info_corefile',
            '--core',
            coredump_file_name,
        ]
        espcoredump_args += extra_args
        espcoredump_args.append(self.app.elf_file)
        logging.info('Running %s', ' '.join(espcoredump_args))
        logging.info('espcoredump output is written to %s', self.coredump_output.name)

        subprocess.check_call(espcoredump_args, stdout=self.coredump_output)
        self.coredump_output.flush()
        self.coredump_output.seek(0)

    def process_coredump_uart(self, expected: Optional[List[Union[str, re.Pattern]]] = None) -> None:
        """Extract the core dump from UART output of the test, run espcoredump on it"""
        self.expect(self.COREDUMP_UART_START)
        res = self.expect('(.+)' + self.COREDUMP_UART_END)
        coredump_base64 = res.group(1).decode('utf8')
        with open(os.path.join(self.logdir, 'coredump_data.b64'), 'w') as coredump_file:
            logging.info('Writing UART base64 core dump to %s', coredump_file.name)
            coredump_file.write(coredump_base64)

        output_file_name = os.path.join(self.logdir, 'coredump_uart_result.txt')
        self._call_espcoredump(
            ['--core-format', 'b64'], coredump_file.name, output_file_name
        )
        if expected:
            self.expect_coredump(output_file_name, expected)

    def process_coredump_flash(self, expected: Optional[List[Union[str, re.Pattern]]] = None) -> None:
        """Extract the core dump from flash, run espcoredump on it"""
        coredump_file_name = os.path.join(self.logdir, 'coredump_data.bin')
        logging.info('Writing flash binary core dump to %s', coredump_file_name)
        self.serial.dump_flash(partition='coredump', output=coredump_file_name)

        output_file_name = os.path.join(self.logdir, 'coredump_flash_result.txt')
        self._call_espcoredump(
            ['--core-format', 'raw'], coredump_file_name, output_file_name
        )
        if expected:
            self.expect_coredump(output_file_name, expected)

    def gdb_write(self, command: str) -> Any:
        """
        Wrapper to write to gdb with a longer timeout, as test runner
        host can be slow sometimes
        """
        assert self.gdbmi, 'This function should be called only after start_gdb'
        return self.gdbmi.write(command, timeout_sec=10)

    def start_gdb(self) -> None:
        """
        Runs GDB and connects it to the "serial" port of the DUT.
        After this, the DUT expect methods can no longer be used to capture output.
        """
        gdb_args = ['--nx', '--quiet', '--interpreter=mi2']
        if self.is_xtensa:
            gdb_path = 'xtensa-esp-elf-gdb-no-python'  # TODO: GCC-311
            gdb_args = [f'--mcpu={self.target}'] + gdb_args
        else:
            gdb_path = 'riscv32-esp-elf-gdb-no-python'  # TODO: GCC-311
        try:
            from pygdbmi.constants import GdbTimeoutError
            gdb_command = [gdb_path] + gdb_args
            self.gdbmi = GdbController(command=gdb_command)
            pygdbmi_logger = attach_logger()
        except ImportError:
            # fallback for pygdbmi<0.10.0.0.
            from pygdbmi.gdbcontroller import GdbTimeoutError
            self.gdbmi = GdbController(gdb_path=gdb_path, gdb_args=gdb_args)
            pygdbmi_logger = self.gdbmi.logger

        # pygdbmi logs to console by default, make it log to a file instead
        pygdbmi_log_file_name = os.path.join(self.logdir, 'pygdbmi_log.txt')
        pygdbmi_logger.setLevel(logging.DEBUG)
        while pygdbmi_logger.hasHandlers():
            pygdbmi_logger.removeHandler(pygdbmi_logger.handlers[0])
        log_handler = logging.FileHandler(pygdbmi_log_file_name)
        log_handler.setFormatter(
            logging.Formatter('%(asctime)s %(levelname)s: %(message)s')
        )
        logging.info(f'Saving pygdbmi logs to {pygdbmi_log_file_name}')
        pygdbmi_logger.addHandler(log_handler)
        try:
            gdb_command = self.gdbmi.command
        except AttributeError:
            # fallback for pygdbmi < 0.10
            gdb_command = self.gdbmi.cmd

        logging.info(f'Running command: "{" ".join(quote_string(c) for c in gdb_command)}"')
        for _ in range(10):
            try:
                # GdbController creates a process with subprocess.Popen(). Is it really running? It is probable that
                # an RPI under high load will get non-responsive during creating a lot of processes.
                if not hasattr(self.gdbmi, 'verify_valid_gdb_subprocess'):
                    # for pygdbmi >= 0.10.0.0
                    verify_valid_gdb_subprocess(self.gdbmi.gdb_process)
                resp = self.gdbmi.get_gdb_response(
                    timeout_sec=10
                )  # calls verify_valid_gdb_subprocess() internally for pygdbmi < 0.10.0.0
                # it will be interesting to look up this response if the next GDB command fails (times out)
                logging.info('GDB response: %s', resp)
                break  # success
            except GdbTimeoutError:
                logging.warning(
                    'GDB internal error: cannot get response from the subprocess'
                )
            except NoGdbProcessError:
                logging.error('GDB internal error: process is not running')
                break  # failure - TODO: create another GdbController
            except ValueError:
                logging.error(
                    'GDB internal error: select() returned an unexpected file number'
                )

        # Set up logging for GDB remote protocol
        gdb_remotelog_file_name = os.path.join(self.logdir, 'gdb_remote_log.txt')
        self.gdb_write('-gdb-set remotelogfile ' + gdb_remotelog_file_name)

        # Load the ELF file
        self.gdb_write('-file-exec-and-symbols {}'.format(self.app.elf_file))

        # Connect GDB to UART
        self.serial.close()
        logging.info('Connecting to GDB Stub...')
        self.gdb_write('-gdb-set serial baud 115200')

        if sys.platform == 'darwin':
            assert '/dev/tty.' not in self.serial.port, \
                '/dev/tty.* ports can\'t be used with GDB on macOS. Use with /dev/cu.* instead.'

        # Make sure we get the 'stopped' notification
        responses = self.gdb_write('-target-select remote ' + self.serial.port)
        stop_response = self.find_gdb_response('stopped', 'notify', responses)

        retries = 3
        while not stop_response and retries > 0:
            logging.info('Sending -exec-interrupt')
            responses = self.gdb_write('-exec-interrupt')
            stop_response = self.find_gdb_response('stopped', 'notify', responses)
            retries -= 1

        frame = stop_response['payload']['frame']
        if 'file' not in frame:
            frame['file'] = '?'
        if 'line' not in frame:
            frame['line'] = '?'
        logging.info('Stopped in {func} at {addr} ({file}:{line})'.format(**frame))

        # Drain remaining responses
        self.gdbmi.get_gdb_response(raise_error_on_timeout=False)

    def gdb_backtrace(self) -> Any:
        """
        Returns the list of stack frames for the current thread.
        Each frame is a dictionary, refer to pygdbmi docs for the format.
        """
        assert self.gdbmi

        responses = self.gdb_write('-stack-list-frames')
        return self.find_gdb_response('done', 'result', responses)['payload']['stack']

    @staticmethod
    def verify_gdb_backtrace(
        gdb_backtrace: List[Any], expected_functions_list: List[Any]
    ) -> None:
        """
        Raises an assert if the function names listed in expected_functions_list do not match the backtrace
        given by gdb_backtrace argument. The latter is in the same format as returned by gdb_backtrace()
        function.
        """
        actual_functions_list = [frame['func'] for frame in gdb_backtrace]
        if actual_functions_list != expected_functions_list:
            logging.error(f'Expected backtrace: {expected_functions_list}')
            logging.error(f'Actual backtrace: {actual_functions_list}')
            assert False, 'Got unexpected backtrace'

    @staticmethod
    def find_gdb_response(
        message: str, response_type: str, responses: List[Any]
    ) -> Any:
        """
        Helper function which extracts one response from an array of GDB responses, filtering
        by message and type. Returned message is a dictionary, refer to pygdbmi docs for the format.
        """

        def match_response(response: Dict[str, Any]) -> bool:
            return response['message'] == message and response['type'] == response_type  # type: ignore

        filtered_responses = [r for r in responses if match_response(r)]
        if not filtered_responses:
            return None
        return filtered_responses[0]