Dear Sir,
Recently, on our project, we encountered an issue related to FlexCAN interrupts. Here's a description of the phenomenon:
On CAN instance 2, we have configured 14 message buffers (MBs) for receiving messages. After a period of testing, it was discovered that the entire RTOS task scheduling had become unresponsive.
Through breakpoints, we found that the system was continuously processing the FLEXCAN_IRQHandler. Further debugging revealed that the corresponding message buffer interrupt was active, but the status of the message buffer was FLEXCAN_MB_IDLE. This resulted in the inability of FLEXCAN_IRQHandler to clear the interrupt flag for that specific message buffer, which potentially caused the continuous invocation of FLEXCAN_IRQHandler and prevented the system from handling other interrupts.
Here's an overview of how the message buffers were being used during the issue reproduction:
Could you please provide some debugging suggestions?
Thanks！
Hi,
later SDK versions (from SDK4.0.x) fixed this behavior with possible trapping in ISR loop, as far as I know. Below is code from 4.0.3 version...
So please update to latest SDK.
BR, Petr
Hi,
later SDK versions (from SDK4.0.x) fixed this behavior with possible trapping in ISR loop, as far as I know. Below is code from 4.0.3 version...
So please update to latest SDK.
BR, Petr
