Just want to report another incident with a Leaf. My wife was driving to work this morning, according to her the Leaf started to display warning messages and then did a full HV power shutdown with a "Service EV System" "No Power" error. When I was able to get to her there were a number of DTC's being reported, I was not able to clear these with LeafSpy even with the OVMS module unplugged and had to disconnect the 12v battery. This behavior is consistent with the symptoms that I saw when the OVMS was crashing with a buffer overrun. I'll have a look at the logs when she gets home tonight to see if there are any indicators as to what is going on. Let me know if there are any specific investigative steps that I should be following. Thanks Wayne
Could you supply a version number? We did address the buffer issue. //. On Tue, 8 Apr 2025, 07:10 Wayne Love via OvmsDev, < ovmsdev@lists.openvehicles.com> wrote:
Just want to report another incident with a Leaf. My wife was driving to work this morning, according to her the Leaf started to display warning messages and then did a full HV power shutdown with a "Service EV System" "No Power" error.
When I was able to get to her there were a number of DTC's being reported, I was not able to clear these with LeafSpy even with the OVMS module unplugged and had to disconnect the 12v battery. This behavior is consistent with the symptoms that I saw when the OVMS was crashing with a buffer overrun.
I'll have a look at the logs when she gets home tonight to see if there are any indicators as to what is going on.
Let me know if there are any specific investigative steps that I should be following.
Thanks Wayne
_______________________________________________ OvmsDev mailing list OvmsDev@lists.openvehicles.com http://lists.openvehicles.com/mailman/listinfo/ovmsdev
The firmware that I am running was built from my fork of the OVMS master, it is essentially equivalate to the master as it stands (I am up to date with commits as of this morning) but with some additional debug logging added. I mention the buffer overruns simply because the symptoms that I am having are similar. The buffer overrun that you mention has been fixed, additionally I've also addressed the buffer overrun mentioned here [Ovmsdev] OVMS Poller module/singleton in pull request #1120. It is my theory from reading the investigations to date on this list and from personal experience, whenever the module does a hard crash (eg as above, memory corruption from a buffer overrun causes the CPU to crash) then the module has the potential of causing a number of errors on the CAN bus. Temper this with the knowledge that I know absolutely nothing about CAN busses and automotive technology. I'm posting here so that wiser minds can put forward better theories. Thanks Wayne On Apr 8, 2025, at 10:20 AM, Michael Geddes <frog@bunyip.wheelycreek.net> wrote: Could you supply a version number? We did address the buffer issue. //. On Tue, 8 Apr 2025, 07:10 Wayne Love via OvmsDev, < ovmsdev@lists.openvehicles.com > wrote: Just want to report another incident with a Leaf. My wife was driving to work this morning, according to her the Leaf started to display warning messages and then did a full HV power shutdown with a "Service EV System" "No Power" error. When I was able to get to her there were a number of DTC's being reported, I was not able to clear these with LeafSpy even with the OVMS module unplugged and had to disconnect the 12v battery. This behavior is consistent with the symptoms that I saw when the OVMS was crashing with a buffer overrun. I'll have a look at the logs when she gets home tonight to see if there are any indicators as to what is going on. Let me know if there are any specific investigative steps that I should be following. Thanks Wayne _______________________________________________ OvmsDev mailing list OvmsDev@lists.openvehicles.com http://lists.openvehicles.com/mailman/listinfo/ovmsdev
Thanks Wayne, It looks like it's all in hardware territory now which is above my pay grade too. I'm curious if the stack trace from this last crash was helpful or pointed at any area in particular? I'm obviously especially interested if it might have been due to issues with the poller. //.ichael On Tue, 8 Apr 2025 at 09:09, Wayne Love <wayne@gsjs.org> wrote:
The firmware that I am running was built from my fork of the OVMS master, it is essentially equivalate to the master as it stands (I am up to date with commits as of this morning) but with some additional debug logging added.
I mention the buffer overruns simply because the symptoms that I am having are similar. The buffer overrun that you mention has been fixed, additionally I've also addressed the buffer overrun mentioned here [Ovmsdev] OVMS Poller module/singleton <http://lists.openvehicles.com/pipermail/ovmsdev/2025-February/016685.html> in pull request #1120.
It is my theory from reading the investigations to date on this list and from personal experience, whenever the module does a hard crash (eg as above, memory corruption from a buffer overrun causes the CPU to crash) then the module has the potential of causing a number of errors on the CAN bus. Temper this with the knowledge that I know absolutely nothing about CAN busses and automotive technology.
I'm posting here so that wiser minds can put forward better theories.
Thanks Wayne
On Apr 8, 2025, at 10:20 AM, Michael Geddes <frog@bunyip.wheelycreek.net> wrote:
Could you supply a version number?
We did address the buffer issue.
//.
On Tue, 8 Apr 2025, 07:10 Wayne Love via OvmsDev, < ovmsdev@lists.openvehicles.com> wrote:
Just want to report another incident with a Leaf. My wife was driving to work this morning, according to her the Leaf started to display warning messages and then did a full HV power shutdown with a "Service EV System" "No Power" error.
When I was able to get to her there were a number of DTC's being reported, I was not able to clear these with LeafSpy even with the OVMS module unplugged and had to disconnect the 12v battery. This behavior is consistent with the symptoms that I saw when the OVMS was crashing with a buffer overrun.
I'll have a look at the logs when she gets home tonight to see if there are any indicators as to what is going on.
Let me know if there are any specific investigative steps that I should be following.
Thanks Wayne
_______________________________________________ OvmsDev mailing list OvmsDev@lists.openvehicles.com http://lists.openvehicles.com/mailman/listinfo/ovmsdev
On Tue, 8 Apr 2025 at 09:09, Wayne Love <wayne@gsjs.org> wrote:
It is my theory from reading the investigations to date on this list and from personal experience, whenever the module does a hard crash (eg as above, memory corruption from a buffer overrun causes the CPU to crash) then the module has the potential of causing a number of errors on the CAN bus.
Rather than the crash, it could be the subsequent boot that causes the problem. Previously I've caused the 2016 car to throw dashboard errors by typing "module reset" over SSH while charging was in progress. I assume this is an orderly shutdown, and hence it's the boot process that disturbs the CAN bus of an active car. Chris
Rather than the crash, it could be the subsequent boot that causes the problem.
Previously I've caused the 2016 car to throw dashboard errors by typing "module reset" over SSH while charging was in progress. I assume this is an orderly shutdown, and hence it's the boot process that disturbs the CAN bus of an active car.
Chris
My thinking is that the issue is caused when the module goes down. The reason that I say this is that in my instances when the module crashes it does not come back up again, you need to power cycle it for it to restart. Is there a chance that we are seeing floating voltages on the CAN connectors when they are not actively controlled by the cpu? Cheers Wayne
Chris, Wayne, you can help by testing switching off the SN65: https://github.com/openvehicles/Open-Vehicle-Monitoring-System-3/issues/980#... As Richard wrote, he could not reproduce the issue in his Leaf by that, but on his bench. If that's the culprit, there still is the Housekeeping SN65 switch-off to address, but there is probably little we can do about cold boots. Regarding floating voltages, see Richards followup post. Also I suggest asking Richard how you can help in narrowing this down. Regards, Michael Am 11.04.25 um 23:18 schrieb Wayne Love via OvmsDev:
Rather than the crash, it could be the subsequent boot that causes the problem.
Previously I've caused the 2016 car to throw dashboard errors by typing "module reset" over SSH while charging was in progress. I assume this is an orderly shutdown, and hence it's the boot process that disturbs the CAN bus of an active car.
Chris My thinking is that the issue is caused when the module goes down. The reason that I say this is that in my instances when the module crashes it does not come back up again, you need to power cycle it for it to restart.
Is there a chance that we are seeing floating voltages on the CAN connectors when they are not actively controlled by the cpu?
Cheers Wayne _______________________________________________ OvmsDev mailing list OvmsDev@lists.openvehicles.com http://lists.openvehicles.com/mailman/listinfo/ovmsdev
-- Michael Balzer * Am Rahmen 5 * D-58313 Herdecke Fon 02330 9104094 * Handy 0176 20698926
participants (4)
-
Chris Box -
Michael Balzer -
Michael Geddes -
Wayne Love