Other Parts Discussed in Thread: MCU-PLUS-SDK-AM263X
The following symptoms are occurring due to changes in SysCfg.
I'm in trouble because of that.
"example.syscfg" is set when opened in CCS.
The arrow part is the changed part, the □ is check OFF / ■ is check ON.
[Clock]
instance : RTI0
Input Clock Frequency(Hz) : 25000000
Input Clock Source : WUCPUCLK
Tick Period (usecs) : 1000 → 100 (*)
[TIMER] Initially, it is empty, so add one with the ADD button
Name CONFIG_TIMER : 0
Input Clock Frequency(Hz) : 25000000
Input Clock Source : WCPUCLK
TickPeriod (usecs) : 1000 → 100
Actual Tick Period (usecs) : 100 (non-editable and contains the above as is)
DMA Trigger Enable : □
Start Timer After Setup : □
Interrupt Mode Enable : ■
Interrupt Priority : 4
Timer Callback : timer0_ISR (creates a callback of the same name in code)
Timer Instance : Any(RTI2)
■Signals : Pins
If the Tick Period of [Clock] remains at 1000, there are the following inconveniences.
1. 0.1ms cycle timer interrupts may not be picked up by the thread
2. The system freezes after about 10 seconds.
->When it stops in the debugger, it flies to the HwiP_data_abort_handler() and has an infinite loop and the cause is unknown.
Since it is fatal, setting 1000usec to (*) is probably not possible(?)
By setting the Tick Period of [Clock] to 100, the above inconvenience is eliminated, but there is another inconvenience.
1. The Ethernet interface does not link up.
->If stop and resume in the debugger, it will link up.
The tasks added to the sample enet_lwip_cpsw() are the same even if the launch process is removed.
The same phenomenon occurs in the enet_lwip_cpsw() sample itself when the * part is set to 100.
