system/hardware: refactor thermal config (#34177)

* system/hardware: refactor thermal config

* fix

* fixup

---------

Co-authored-by: Comma Device <device@comma.ai>

cereal/log.capnp

@@ -488,6 +488,7 @@ struct DeviceState @0xa4d8b5af2aa492eb {
   pmicTempC @39 :List(Float32);
   intakeTempC @46 :Float32;
   exhaustTempC @47 :Float32;
+  caseTempC @48 :Float32;
   maxTempC @44 :Float32;  # max of other temps, used to control fan
   thermalZones @38 :List(ThermalZone);
   thermalStatus @14 :ThermalStatus;

system/hardware/base.py

@@ -1,11 +1,54 @@
+import os
 from abc import abstractmethod, ABC
-from collections import namedtuple
+from dataclasses import dataclass, fields
 
 from cereal import log
 
-ThermalConfig = namedtuple('ThermalConfig', ['cpu', 'gpu', 'mem', 'pmic', 'intake', 'exhaust'])
 NetworkType = log.DeviceState.NetworkType
 
+@dataclass
+class ThermalZone:
+  # a zone from /sys/class/thermal/thermal_zone*
+  name: str             # a.k.a type
+  scale: float = 1000.  # scale to get degrees in C
+  zone_number = -1
+
+  def read(self) -> float:
+    if self.zone_number < 0:
+      for n in os.listdir("/sys/devices/virtual/thermal"):
+        if not n.startswith("thermal_zone"):
+          continue
+        with open(os.path.join("/sys/devices/virtual/thermal", n, "type")) as f:
+          if f.read().strip() == self.name:
+            self.zone_number = int(n.removeprefix("thermal_zone"))
+            break
+
+    try:
+      with open(f"/sys/devices/virtual/thermal/thermal_zone{self.zone_number}/temp") as f:
+        return int(f.read()) / self.scale
+    except FileNotFoundError:
+      return 0
+
+@dataclass
+class ThermalConfig:
+  cpu: list[ThermalZone] | None = None
+  gpu: list[ThermalZone] | None = None
+  pmic: list[ThermalZone] | None = None
+  memory: ThermalZone | None = None
+  intake: ThermalZone | None = None
+  exhaust: ThermalZone | None = None
+  case: ThermalZone | None = None
+
+  def get_msg(self):
+    ret = {}
+    for f in fields(ThermalConfig):
+      v = getattr(self, f.name)
+      if v is not None:
+        if isinstance(v, list):
+          ret[f.name + "TempC"] = [x.read() for x in v]
+        else:
+          ret[f.name + "TempC"] = v.read()
+    return ret
 
 class HardwareBase(ABC):
   @staticmethod
@@ -84,9 +127,8 @@ class HardwareBase(ABC):
   def shutdown(self):
     pass
 
-  @abstractmethod
   def get_thermal_config(self):
-    pass
+    return ThermalConfig()
 
   @abstractmethod
   def set_screen_brightness(self, percentage):

system/hardware/hardwared.py

@@ -51,41 +51,6 @@ OFFROAD_DANGER_TEMP = 75
 
 prev_offroad_states: dict[str, tuple[bool, str | None]] = {}
 
-tz_by_type: dict[str, int] | None = None
-def populate_tz_by_type():
-  global tz_by_type
-  tz_by_type = {}
-  for n in os.listdir("/sys/devices/virtual/thermal"):
-    if not n.startswith("thermal_zone"):
-      continue
-    with open(os.path.join("/sys/devices/virtual/thermal", n, "type")) as f:
-      tz_by_type[f.read().strip()] = int(n.removeprefix("thermal_zone"))
-
-def read_tz(x):
-  if x is None:
-    return 0
-
-  if isinstance(x, str):
-    if tz_by_type is None:
-      populate_tz_by_type()
-    x = tz_by_type[x]
-
-  try:
-    with open(f"/sys/devices/virtual/thermal/thermal_zone{x}/temp") as f:
-      return int(f.read())
-  except FileNotFoundError:
-    return 0
-
-
-def read_thermal(thermal_config):
-  dat = messaging.new_message('deviceState', valid=True)
-  dat.deviceState.cpuTempC = [read_tz(z) / thermal_config.cpu[1] for z in thermal_config.cpu[0]]
-  dat.deviceState.gpuTempC = [read_tz(z) / thermal_config.gpu[1] for z in thermal_config.gpu[0]]
-  dat.deviceState.memoryTempC = read_tz(thermal_config.mem[0]) / thermal_config.mem[1]
-  dat.deviceState.pmicTempC = [read_tz(z) / thermal_config.pmic[1] for z in thermal_config.pmic[0]]
-  dat.deviceState.intakeTempC = read_tz(thermal_config.intake[0]) / thermal_config.intake[1]
-  dat.deviceState.exhaustTempC = read_tz(thermal_config.exhaust[0]) / thermal_config.exhaust[1]
-  return dat
 
 
 def set_offroad_alert_if_changed(offroad_alert: str, show_alert: bool, extra_text: str | None=None):
@@ -234,7 +199,8 @@ def hardware_thread(end_event, hw_queue) -> None:
     if (sm.frame % round(SERVICE_LIST['pandaStates'].frequency * DT_HW) != 0) and not ign_edge:
       continue
 
-    msg = read_thermal(thermal_config)
+    msg = messaging.new_message('deviceState', valid=True)
+    msg.deviceState = thermal_config.get_msg()
     msg.deviceState.deviceType = HARDWARE.get_device_type()
 
     try:

system/hardware/pc/hardware.py

@@ -1,7 +1,7 @@
 import random
 
 from cereal import log
-from openpilot.system.hardware.base import HardwareBase, ThermalConfig
+from openpilot.system.hardware.base import HardwareBase
 
 NetworkType = log.DeviceState.NetworkType
 NetworkStrength = log.DeviceState.NetworkStrength
@@ -53,9 +53,6 @@ class Pc(HardwareBase):
   def shutdown(self):
     print("SHUTDOWN!")
 
-  def get_thermal_config(self):
-    return ThermalConfig(cpu=((None,), 1), gpu=((None,), 1), mem=(None, 1), pmic=((None,), 1), intake=(None, 1), exhaust=(None, 1))
-
   def set_screen_brightness(self, percentage):
     pass
 

system/hardware/tici/hardware.py

@@ -10,7 +10,7 @@ from pathlib import Path
 
 from cereal import log
 from openpilot.common.gpio import gpio_set, gpio_init, get_irqs_for_action
-from openpilot.system.hardware.base import HardwareBase, ThermalConfig
+from openpilot.system.hardware.base import HardwareBase, ThermalConfig, ThermalZone
 from openpilot.system.hardware.tici import iwlist
 from openpilot.system.hardware.tici.pins import GPIO
 from openpilot.system.hardware.tici.amplifier import Amplifier
@@ -323,17 +323,19 @@ class Tici(HardwareBase):
     os.system("sudo poweroff")
 
   def get_thermal_config(self):
-    intake, exhaust = (None, 1), (None, 1)
+    intake, exhaust, case = None, None, None
     if self.get_device_type() == "mici":
-      intake = ("intake", 1000)
+      case = ThermalZone("case")
-      exhaust = ("exhaust", 1000)
+      intake = ThermalZone("intake")
-    return ThermalConfig(cpu=([f"cpu{i}-silver-usr" for i in range(4)] +
+      exhaust = ThermalZone("exhaust")
-                              [f"cpu{i}-gold-usr" for i in range(4)], 1000),
+    return ThermalConfig(cpu=[ThermalZone(f"cpu{i}-silver-usr") for i in range(4)] +
-                         gpu=(("gpu0-usr", "gpu1-usr"), 1000),
+                             [ThermalZone(f"cpu{i}-gold-usr") for i in range(4)],
-                         mem=("ddr-usr", 1000),
+                         gpu=[ThermalZone("gpu0-usr"), ThermalZone("gpu1-usr")],
-                         pmic=(("pm8998_tz", "pm8005_tz"), 1000),
+                         memory=ThermalZone("ddr-usr"),
+                         pmic=[ThermalZone("pm8998_tz"), ThermalZone("pm8005_tz")],
                          intake=intake,
-                         exhaust=exhaust)
+                         exhaust=exhaust,
+                         case=case)
 
   def set_screen_brightness(self, percentage):
     try: