[device/celestica] Implement Watchdog APIs based on the new platform API (#3123)

This commit is contained in:
Wirut Getbamrung 2019-07-09 01:26:27 +07:00 committed by Joe LeVeque
parent 8329ce1d63
commit bd672000ab
6 changed files with 574 additions and 27 deletions

View File

@ -20,12 +20,15 @@ try:
from sonic_platform.psu import Psu
from sonic_platform.device import Device
from sonic_platform.component import Component
from sonic_platform.watchdog import Watchdog
except ImportError as e:
raise ImportError(str(e) + "- required module not found")
CONFIG_DB_PATH = "/etc/sonic/config_db.json"
NUM_FAN = 3
NUM_PSU = 2
RESET_REGISTER = "0x112"
REBOOT_CAUSE_PATH = "/host/reboot-cause/previous-reboot-cause.txt"
class Chassis(ChassisBase):
@ -42,6 +45,7 @@ class Chassis(ChassisBase):
ChassisBase.__init__(self)
self._component_device = Device("component")
self._component_name_list = self._component_device.get_name_list()
self._watchdog = Watchdog()
def __read_config_db(self):
try:
@ -51,6 +55,14 @@ class Chassis(ChassisBase):
except IOError:
raise IOError("Unable to open config_db file !")
def __read_txt_file(self, file_path):
try:
with open(file_path, 'r') as fd:
data = fd.read()
return data.strip()
except IOError:
raise IOError("Unable to open %s file !" % file_path)
def get_base_mac(self):
"""
Retrieves the base MAC address for the chassis
@ -63,7 +75,7 @@ class Chassis(ChassisBase):
base_mac = self.config_data["DEVICE_METADATA"]["localhost"]["mac"]
return str(base_mac)
except KeyError:
raise KeyError("Base MAC not found")
return str(None)
def get_firmware_version(self, component_name):
"""
@ -94,3 +106,32 @@ class Chassis(ChassisBase):
if component_name not in self._component_name_list:
return False
return self.component.upgrade_firmware(image_path)
def get_reboot_cause(self):
"""
Retrieves the cause of the previous reboot
Returns:
A tuple (string, string) where the first element is a string
containing the cause of the previous reboot. This string must be
one of the predefined strings in this class. If the first string
is "REBOOT_CAUSE_HARDWARE_OTHER", the second string can be used
to pass a description of the reboot cause.
"""
self.component = Component("SMC_CPLD")
description = 'None'
reboot_cause = self.REBOOT_CAUSE_HARDWARE_OTHER
hw_reboot_cause = self.component.get_register_value(RESET_REGISTER)
sw_reboot_cause = self.__read_txt_file(REBOOT_CAUSE_PATH)
if sw_reboot_cause != "Unexpected reboot":
reboot_cause = self.REBOOT_CAUSE_NON_HARDWARE
elif hw_reboot_cause == "0x11":
reboot_cause = self.REBOOT_CAUSE_POWER_LOSS
elif hw_reboot_cause == "0x33" or hw_reboot_cause == "0x55":
reboot_cause = self.REBOOT_CAUSE_WATCHDOG
else:
reboot_cause = self.REBOOT_CAUSE_HARDWARE_OTHER
description = 'Unknown reason'
return (reboot_cause, description)

View File

@ -23,7 +23,7 @@ MMC_CPLD_ADDR = '0x100'
BIOS_VERSION_PATH = "/sys/class/dmi/id/bios_version"
CONFIG_DB_PATH = "/etc/sonic/config_db.json"
SMC_CPLD_PATH = "/sys/devices/platform/e1031.smc/version"
MMC_CPLD_PATH = "/sys/devices/platform/e1031.smc/getreg"
GETREG_PATH = "/sys/devices/platform/e1031.smc/getreg"
class Component(DeviceBase):
@ -51,16 +51,6 @@ class Component(DeviceBase):
return False
return True
def __get_register_value(self, path, register):
# Retrieves the cpld register value
cmd = "echo {1} > {0}; cat {0}".format(path, register)
p = subprocess.Popen(
cmd, shell=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
raw_data, err = p.communicate()
if err is not '':
return None
return raw_data.strip()
def __get_bios_version(self):
# Retrieves the BIOS firmware version
try:
@ -70,6 +60,16 @@ class Component(DeviceBase):
except Exception as e:
return None
def get_register_value(self, register):
# Retrieves the cpld register value
cmd = "echo {1} > {0}; cat {0}".format(GETREG_PATH, register)
p = subprocess.Popen(
cmd, shell=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
raw_data, err = p.communicate()
if err is not '':
return None
return raw_data.strip()
def __get_cpld_version(self):
# Retrieves the CPLD firmware version
cpld_version = dict()
@ -78,8 +78,7 @@ class Component(DeviceBase):
smc_cpld_version = 'None' if smc_cpld_version is 'None' else "{}.{}".format(
int(smc_cpld_version[2], 16), int(smc_cpld_version[3], 16))
mmc_cpld_version = self.__get_register_value(
MMC_CPLD_PATH, MMC_CPLD_ADDR)
mmc_cpld_version = self.get_register_value(MMC_CPLD_ADDR)
mmc_cpld_version = 'None' if mmc_cpld_version is 'None' else "{}.{}".format(
int(mmc_cpld_version[2], 16), int(mmc_cpld_version[3], 16))

View File

@ -0,0 +1,233 @@
#!/usr/bin/env python
#############################################################################
# Celestica
#
# Watchdog contains an implementation of SONiC Platform Base API
#
#############################################################################
import ctypes
import fcntl
import os
import subprocess
import time
import array
try:
from sonic_platform_base.watchdog_base import WatchdogBase
except ImportError as e:
raise ImportError(str(e) + "- required module not found")
""" ioctl constants """
IO_WRITE = 0x40000000
IO_READ = 0x80000000
IO_READ_WRITE = 0xC0000000
IO_SIZE_INT = 0x00040000
IO_SIZE_40 = 0x00280000
IO_TYPE_WATCHDOG = ord('W') << 8
WDR_INT = IO_READ | IO_SIZE_INT | IO_TYPE_WATCHDOG
WDR_40 = IO_READ | IO_SIZE_40 | IO_TYPE_WATCHDOG
WDWR_INT = IO_READ_WRITE | IO_SIZE_INT | IO_TYPE_WATCHDOG
""" Watchdog ioctl commands """
WDIOC_GETSUPPORT = 0 | WDR_40
WDIOC_GETSTATUS = 1 | WDR_INT
WDIOC_GETBOOTSTATUS = 2 | WDR_INT
WDIOC_GETTEMP = 3 | WDR_INT
WDIOC_SETOPTIONS = 4 | WDR_INT
WDIOC_KEEPALIVE = 5 | WDR_INT
WDIOC_SETTIMEOUT = 6 | WDWR_INT
WDIOC_GETTIMEOUT = 7 | WDR_INT
WDIOC_SETPRETIMEOUT = 8 | WDWR_INT
WDIOC_GETPRETIMEOUT = 9 | WDR_INT
WDIOC_GETTIMELEFT = 10 | WDR_INT
""" Watchdog status constants """
WDIOS_DISABLECARD = 0x0001
WDIOS_ENABLECARD = 0x0002
WDT_COMMON_ERROR = -1
WD_MAIN_IDENTITY = "iTCO_wdt"
WDT_SYSFS_PATH = "/sys/class/watchdog/"
class Watchdog(WatchdogBase):
def __init__(self):
self.watchdog, self.wdt_main_dev_name = self._get_wdt()
self.status_path = "/sys/class/watchdog/%s/status" % self.wdt_main_dev_name
self.state_path = "/sys/class/watchdog/%s/state" % self.wdt_main_dev_name
self.timeout_path = "/sys/class/watchdog/%s/timeout" % self.wdt_main_dev_name
# Set default value
self._disable()
self.armed = False
self.timeout = self._gettimeout(self.timeout_path)
def _is_wd_main(self, dev):
"""
Checks watchdog identity
"""
identity = self._read_file(
"{}/{}/identity".format(WDT_SYSFS_PATH, dev))
return identity == WD_MAIN_IDENTITY
def _get_wdt(self):
"""
Retrieves watchdog device
"""
wdt_main_dev_list = [dev for dev in os.listdir(
"/dev/") if dev.startswith("watchdog") and self._is_wd_main(dev)]
if not wdt_main_dev_list:
return None
wdt_main_dev_name = wdt_main_dev_list[0]
watchdog_device_path = "/dev/{}".format(wdt_main_dev_name)
watchdog = os.open(watchdog_device_path, os.O_RDWR)
return watchdog, wdt_main_dev_name
def _read_file(self, file_path):
"""
Read text file
"""
try:
with open(file_path, "r") as fd:
txt = fd.read()
except IOError:
return WDT_COMMON_ERROR
return txt.strip()
def _enable(self):
"""
Turn on the watchdog timer
"""
req = array.array('h', [WDIOS_ENABLECARD])
fcntl.ioctl(self.watchdog, WDIOC_SETOPTIONS, req, False)
def _disable(self):
"""
Turn off the watchdog timer
"""
req = array.array('h', [WDIOS_DISABLECARD])
fcntl.ioctl(self.watchdog, WDIOC_SETOPTIONS, req, False)
def _keepalive(self):
"""
Keep alive watchdog timer
"""
fcntl.ioctl(self.watchdog, WDIOC_KEEPALIVE)
def _settimeout(self, seconds):
"""
Set watchdog timer timeout
@param seconds - timeout in seconds
@return is the actual set timeout
"""
req = array.array('I', [seconds])
fcntl.ioctl(self.watchdog, WDIOC_SETTIMEOUT, req, True)
return int(req[0])
def _gettimeout(self, timeout_path):
"""
Get watchdog timeout
@return watchdog timeout
"""
req = array.array('I', [0])
fcntl.ioctl(self.watchdog, WDIOC_GETTIMEOUT, req, True)
return int(req[0])
def _gettimeleft(self):
"""
Get time left before watchdog timer expires
@return time left in seconds
"""
req = array.array('I', [0])
fcntl.ioctl(self.watchdog, WDIOC_GETTIMELEFT, req, True)
return int(req[0])
#################################################################
def arm(self, seconds):
"""
Arm the hardware watchdog with a timeout of <seconds> seconds.
If the watchdog is currently armed, calling this function will
simply reset the timer to the provided value. If the underlying
hardware does not support the value provided in <seconds>, this
method should arm the watchdog with the *next greater* available
value.
Returns:
An integer specifying the *actual* number of seconds the watchdog
was armed with. On failure returns -1.
"""
ret = WDT_COMMON_ERROR
if seconds < 0:
return ret
try:
if self.timeout != seconds:
self.timeout = self._settimeout(seconds)
if self.armed:
self._keepalive()
else:
self._enable()
self.armed = True
ret = self.timeout
except IOError as e:
pass
return ret
def disarm(self):
"""
Disarm the hardware watchdog
Returns:
A boolean, True if watchdog is disarmed successfully, False if not
"""
disarmed = False
if self.is_armed():
try:
self._disable()
self.armed = False
disarmed = True
except IOError:
pass
return disarmed
def is_armed(self):
"""
Retrieves the armed state of the hardware watchdog.
Returns:
A boolean, True if watchdog is armed, False if not
"""
return self.armed
def get_remaining_time(self):
"""
If the watchdog is armed, retrieve the number of seconds remaining on
the watchdog timer
Returns:
An integer specifying the number of seconds remaining on thei
watchdog timer. If the watchdog is not armed, returns -1.
"""
timeleft = WDT_COMMON_ERROR
if self.armed:
try:
timeleft = self._gettimeleft()
except IOError:
pass
return timeleft
def __del__(self):
"""
Close watchdog
"""
os.close(self.watchdog)

View File

@ -20,12 +20,15 @@ try:
from sonic_platform.psu import Psu
from sonic_platform.device import Device
from sonic_platform.component import Component
from sonic_platform.watchdog import Watchdog
except ImportError as e:
raise ImportError(str(e) + "- required module not found")
CONFIG_DB_PATH = "/etc/sonic/config_db.json"
NUM_FAN = 5
NUM_PSU = 2
RESET_REGISTER = "0x103"
REBOOT_CAUSE_PATH = "/host/reboot-cause/previous-reboot-cause.txt"
class Chassis(ChassisBase):
@ -42,6 +45,7 @@ class Chassis(ChassisBase):
ChassisBase.__init__(self)
self._component_device = Device("component")
self._component_name_list = self._component_device.get_name_list()
self._watchdog = Watchdog()
def __read_config_db(self):
try:
@ -51,6 +55,14 @@ class Chassis(ChassisBase):
except IOError:
raise IOError("Unable to open config_db file !")
def __read_txt_file(self, file_path):
try:
with open(file_path, 'r') as fd:
data = fd.read()
return data.strip()
except IOError:
raise IOError("Unable to open %s file !" % file_path)
def get_base_mac(self):
"""
Retrieves the base MAC address for the chassis
@ -63,7 +75,7 @@ class Chassis(ChassisBase):
base_mac = self.config_data["DEVICE_METADATA"]["localhost"]["mac"]
return str(base_mac)
except KeyError:
raise KeyError("Base MAC not found")
return str(None)
def get_firmware_version(self, component_name):
"""
@ -94,3 +106,33 @@ class Chassis(ChassisBase):
if component_name not in self._component_name_list:
return False
return self.component.upgrade_firmware(image_path)
def get_reboot_cause(self):
"""
Retrieves the cause of the previous reboot
Returns:
A tuple (string, string) where the first element is a string
containing the cause of the previous reboot. This string must be
one of the predefined strings in this class. If the first string
is "REBOOT_CAUSE_HARDWARE_OTHER", the second string can be used
to pass a description of the reboot cause.
"""
self.component = Component("CPLD1")
description = 'None'
reboot_cause = self.REBOOT_CAUSE_HARDWARE_OTHER
hw_reboot_cause = self.component.get_register_value(RESET_REGISTER)
sw_reboot_cause = self.__read_txt_file(REBOOT_CAUSE_PATH)
if sw_reboot_cause != "Unexpected reboot":
reboot_cause = self.REBOOT_CAUSE_NON_HARDWARE
description = sw_reboot_cause
elif hw_reboot_cause == "0x11":
reboot_cause = self.REBOOT_CAUSE_POWER_LOSS
elif hw_reboot_cause == "0x22":
reboot_cause = self.REBOOT_CAUSE_WATCHDOG,
else:
reboot_cause = self.REBOOT_CAUSE_HARDWARE_OTHER
description = 'Unknown reason'
return (reboot_cause, description)

View File

@ -55,16 +55,6 @@ class Component(DeviceBase):
return False
return True
def __get_register_value(self, path, register):
# Retrieves the cpld register value
cmd = "echo {1} > {0}; cat {0}".format(path, register)
p = subprocess.Popen(
cmd, shell=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
raw_data, err = p.communicate()
if err is not '':
return None
return raw_data.strip()
def __get_bios_version(self):
# Retrieves the BIOS firmware version
try:
@ -74,14 +64,23 @@ class Component(DeviceBase):
except Exception as e:
return None
def get_register_value(self, register):
# Retrieves the cpld register value
cmd = "echo {1} > {0}; cat {0}".format(GETREG_PATH, register)
p = subprocess.Popen(
cmd, shell=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
raw_data, err = p.communicate()
if err is not '':
return None
return raw_data.strip()
def __get_cpld_version(self):
# Retrieves the CPLD firmware version
cpld_version = dict()
for cpld_name in CPLD_ADDR_MAPPING:
try:
cpld_addr = CPLD_ADDR_MAPPING[cpld_name]
cpld_version_raw = self.__get_register_value(
GETREG_PATH, cpld_addr)
cpld_version_raw = self.get_register_value(cpld_addr)
cpld_version_str = "{}.{}".format(int(cpld_version_raw[2], 16), int(
cpld_version_raw[3], 16)) if cpld_version_raw is not None else 'None'
cpld_version[cpld_name] = cpld_version_str

View File

@ -0,0 +1,233 @@
#!/usr/bin/env python
#############################################################################
# Celestica
#
# Watchdog contains an implementation of SONiC Platform Base API
#
#############################################################################
import ctypes
import fcntl
import os
import subprocess
import time
import array
try:
from sonic_platform_base.watchdog_base import WatchdogBase
except ImportError as e:
raise ImportError(str(e) + "- required module not found")
""" ioctl constants """
IO_WRITE = 0x40000000
IO_READ = 0x80000000
IO_READ_WRITE = 0xC0000000
IO_SIZE_INT = 0x00040000
IO_SIZE_40 = 0x00280000
IO_TYPE_WATCHDOG = ord('W') << 8
WDR_INT = IO_READ | IO_SIZE_INT | IO_TYPE_WATCHDOG
WDR_40 = IO_READ | IO_SIZE_40 | IO_TYPE_WATCHDOG
WDWR_INT = IO_READ_WRITE | IO_SIZE_INT | IO_TYPE_WATCHDOG
""" Watchdog ioctl commands """
WDIOC_GETSUPPORT = 0 | WDR_40
WDIOC_GETSTATUS = 1 | WDR_INT
WDIOC_GETBOOTSTATUS = 2 | WDR_INT
WDIOC_GETTEMP = 3 | WDR_INT
WDIOC_SETOPTIONS = 4 | WDR_INT
WDIOC_KEEPALIVE = 5 | WDR_INT
WDIOC_SETTIMEOUT = 6 | WDWR_INT
WDIOC_GETTIMEOUT = 7 | WDR_INT
WDIOC_SETPRETIMEOUT = 8 | WDWR_INT
WDIOC_GETPRETIMEOUT = 9 | WDR_INT
WDIOC_GETTIMELEFT = 10 | WDR_INT
""" Watchdog status constants """
WDIOS_DISABLECARD = 0x0001
WDIOS_ENABLECARD = 0x0002
WDT_COMMON_ERROR = -1
WD_MAIN_IDENTITY = "iTCO_wdt"
WDT_SYSFS_PATH = "/sys/class/watchdog/"
class Watchdog(WatchdogBase):
def __init__(self):
self.watchdog, self.wdt_main_dev_name = self._get_wdt()
self.status_path = "/sys/class/watchdog/%s/status" % self.wdt_main_dev_name
self.state_path = "/sys/class/watchdog/%s/state" % self.wdt_main_dev_name
self.timeout_path = "/sys/class/watchdog/%s/timeout" % self.wdt_main_dev_name
# Set default value
self._disable()
self.armed = False
self.timeout = self._gettimeout(self.timeout_path)
def _is_wd_main(self, dev):
"""
Checks watchdog identity
"""
identity = self._read_file(
"{}/{}/identity".format(WDT_SYSFS_PATH, dev))
return identity == WD_MAIN_IDENTITY
def _get_wdt(self):
"""
Retrieves watchdog device
"""
wdt_main_dev_list = [dev for dev in os.listdir(
"/dev/") if dev.startswith("watchdog") and self._is_wd_main(dev)]
if not wdt_main_dev_list:
return None
wdt_main_dev_name = wdt_main_dev_list[0]
watchdog_device_path = "/dev/{}".format(wdt_main_dev_name)
watchdog = os.open(watchdog_device_path, os.O_RDWR)
return watchdog, wdt_main_dev_name
def _read_file(self, file_path):
"""
Read text file
"""
try:
with open(file_path, "r") as fd:
txt = fd.read()
except IOError:
return WDT_COMMON_ERROR
return txt.strip()
def _enable(self):
"""
Turn on the watchdog timer
"""
req = array.array('h', [WDIOS_ENABLECARD])
fcntl.ioctl(self.watchdog, WDIOC_SETOPTIONS, req, False)
def _disable(self):
"""
Turn off the watchdog timer
"""
req = array.array('h', [WDIOS_DISABLECARD])
fcntl.ioctl(self.watchdog, WDIOC_SETOPTIONS, req, False)
def _keepalive(self):
"""
Keep alive watchdog timer
"""
fcntl.ioctl(self.watchdog, WDIOC_KEEPALIVE)
def _settimeout(self, seconds):
"""
Set watchdog timer timeout
@param seconds - timeout in seconds
@return is the actual set timeout
"""
req = array.array('I', [seconds])
fcntl.ioctl(self.watchdog, WDIOC_SETTIMEOUT, req, True)
return int(req[0])
def _gettimeout(self, timeout_path):
"""
Get watchdog timeout
@return watchdog timeout
"""
req = array.array('I', [0])
fcntl.ioctl(self.watchdog, WDIOC_GETTIMEOUT, req, True)
return int(req[0])
def _gettimeleft(self):
"""
Get time left before watchdog timer expires
@return time left in seconds
"""
req = array.array('I', [0])
fcntl.ioctl(self.watchdog, WDIOC_GETTIMELEFT, req, True)
return int(req[0])
#################################################################
def arm(self, seconds):
"""
Arm the hardware watchdog with a timeout of <seconds> seconds.
If the watchdog is currently armed, calling this function will
simply reset the timer to the provided value. If the underlying
hardware does not support the value provided in <seconds>, this
method should arm the watchdog with the *next greater* available
value.
Returns:
An integer specifying the *actual* number of seconds the watchdog
was armed with. On failure returns -1.
"""
ret = WDT_COMMON_ERROR
if seconds < 0:
return ret
try:
if self.timeout != seconds:
self.timeout = self._settimeout(seconds)
if self.armed:
self._keepalive()
else:
self._enable()
self.armed = True
ret = self.timeout
except IOError as e:
pass
return ret
def disarm(self):
"""
Disarm the hardware watchdog
Returns:
A boolean, True if watchdog is disarmed successfully, False if not
"""
disarmed = False
if self.is_armed():
try:
self._disable()
self.armed = False
disarmed = True
except IOError:
pass
return disarmed
def is_armed(self):
"""
Retrieves the armed state of the hardware watchdog.
Returns:
A boolean, True if watchdog is armed, False if not
"""
return self.armed
def get_remaining_time(self):
"""
If the watchdog is armed, retrieve the number of seconds remaining on
the watchdog timer
Returns:
An integer specifying the number of seconds remaining on thei
watchdog timer. If the watchdog is not armed, returns -1.
"""
timeleft = WDT_COMMON_ERROR
if self.armed:
try:
timeleft = self._gettimeleft()
except IOError:
pass
return timeleft
def __del__(self):
"""
Close watchdog
"""
os.close(self.watchdog)