- Software Version
DRIVE OS 6.0.6
- Target OS
- SDK Manager Version
- Host Machine Version
native Ubuntu Linux 20.04 Host installed with DRIVE OS DOCKER Containers
Describe the bug
From the doc of cgf and stm, we can see that cgf uses the time-triggered-policy of schedule, and stm is
a static, centrally monitored, OS agnostic, non-preemptive scheduler that manages work across hardware engines on NVIDIA DRIVE Orin™ SoC
But in the context of safety of AV workload, there should be many emergency events, I didn’t see any consideration of event handling of CGF schedule.
my qst is, how to handle events in dwCGF?
I am checking on this internally and update you.
Could you give any example of actual examples of events/emergency events/ and expected event handling?
How about creating a separate node to deal with interrupts? Based on your requirement, you can keep the node in beginning of the graph or you can keep it middle of the graph(if the action has to be taken based on some node output results) and propagate the error state across the nodes and have them react accordingly.
You mean do it in the app level rather than framework level?
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.