tici: check that peripherals are fully booted (#30942)

* tici: check that peripherals are fully booted

* default

* typo

* fix

---------

Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: 5f191321fd

selfdrive/thermald/thermald.py

@@ -312,6 +312,9 @@ def thermald_thread(end_event, hw_queue) -> None:
     # must be at an engageable thermal band to go onroad
     startup_conditions["device_temp_engageable"] = thermal_status < ThermalStatus.red
 
+    # ensure device is fully booted
+    startup_conditions["device_booted"] = startup_conditions.get("device_booted", False) or HARDWARE.booted()
+
     # if the temperature enters the danger zone, go offroad to cool down
     onroad_conditions["device_temp_good"] = thermal_status < ThermalStatus.danger
     extra_text = f"{offroad_comp_temp:.1f}C"

system/hardware/base.py

@@ -23,6 +23,9 @@ class HardwareBase(ABC):
     except Exception:
       return default
 
+  def booted(self) -> bool:
+    return True
+
   @abstractmethod
   def reboot(self, reason=None):
     pass

system/hardware/tici/hardware.py

@@ -74,6 +74,11 @@ def sudo_write(val, path):
       # fallback for debugfs files
       os.system(f"sudo su -c 'echo {val} > {path}'")
 
+def sudo_read(path: str) -> str:
+  try:
+    return subprocess.check_output(f"sudo cat {path}", shell=True, encoding='utf8')
+  except Exception:
+    return ""
 
 def affine_irq(val, action):
   irqs = get_irqs_for_action(action)
@@ -554,6 +559,12 @@ class Tici(HardwareBase):
     time.sleep(0.5)
     gpio_set(GPIO.STM_BOOT0, 0)
 
+  def booted(self):
+    # this normally boots within 8s, but on rare occasions takes 30+s
+    encoder_state = sudo_read("/sys/kernel/debug/msm_vidc/core0/info")
+    if "Core state: 0" in encoder_state and (time.monotonic() < 60*2):
+      return False
+    return True
 
 if __name__ == "__main__":
   t = Tici()