4 ms·

Jeep Issues Emergency Recall for OTA-Bricked Wrangler 4xes

13 points by daheza 1y ago

daheza 1y ago
The specific details from nhtsa: https://static.nhtsa.gov/odi/rcl/2025/RCLRPT-25V710-7238.pdf https://static.nhtsa.gov/odi/rcl/2025/RCLRPT-25V710-7238.pdf
- jeffrallen 1y ago
  > Interim corrective action to avoid loss of motive power was implemented using an update to the backend policy Wonder if this is about a misconfigured secure CAN gateway?
duxup 1y ago
I remember working with switches and routers for the first time and I came across a Juniper router who when I did an upgrade I had the option to issue a command to just fall back ... the same applied to config changes and so on. At the time some other systems kinda had this system (sometimes not so great), but Juniper had perfected it. These vehicles really should have a similar dual partition setup where you can on your own go into the menu and say "Yo, things broke, go back." and you're good to go. Based on the last paragraph it sounds like Jeep ... might have such a system, but it isn't clear.
ilamont 1y ago
“An HCP which resets while driving may cause a loss of motive power which can cause a vehicle crash without prior warning.” I assume other modern hybrids have control processors that perform similar functions. Have there been similar incidents/updates/flaws causing in-motion resets?
AlotOfReading 1y ago
I'm not deeply familiar with stellantis's architecture and both of these are relatively new modules for them, but this is baffling to me. I get why the HCP, which orchestrates the power balance, might talk to the TCM. But that interface is inherently crossing a safety boundary and should have been identified as such by systems engineers. Why the fuck (except for disastrously buggy code on the HCP side) would anything on the TCM side cause an HCP reset? Completely wild speculation, but my initial guess is a memory safety issue that could have been caught with better languages or fuzzing. Neither are as common as they should be in automotive.

13 points by daheza 1y ago

The specific details from nhtsa: https://static.nhtsa.gov/odi/rcl/2025/RCLRPT-25V710-7238.pdf https://static.nhtsa.gov/odi/rcl/2025/RCLRPT-25V710-7238.pdf

jeffrallen 1y ago
> Interim corrective action to avoid loss of motive power was implemented using an update to the backend policy Wonder if this is about a misconfigured secure CAN gateway?

duxup 1y ago

I remember working with switches and routers for the first time and I came across a Juniper router who when I did an upgrade I had the option to issue a command to just fall back ... the same applied to config changes and so on. At the time some other systems kinda had this system (sometimes not so great), but Juniper had perfected it. These vehicles really should have a similar dual partition setup where you can on your own go into the menu and say "Yo, things broke, go back." and you're good to go. Based on the last paragraph it sounds like Jeep ... might have such a system, but it isn't clear.

ilamont 1y ago

“An HCP which resets while driving may cause a loss of motive power which can cause a vehicle crash without prior warning.” I assume other modern hybrids have control processors that perform similar functions. Have there been similar incidents/updates/flaws causing in-motion resets?

AlotOfReading 1y ago

I'm not deeply familiar with stellantis's architecture and both of these are relatively new modules for them, but this is baffling to me. I get why the HCP, which orchestrates the power balance, might talk to the TCM. But that interface is inherently crossing a safety boundary and should have been identified as such by systems engineers. Why the fuck (except for disastrously buggy code on the HCP side) would anything on the TCM side cause an HCP reset? Completely wild speculation, but my initial guess is a memory safety issue that could have been caught with better languages or fuzzing. Neither are as common as they should be in automotive.