Interrupts and interrupt service routines. As the same equipment was used with the same RTOS, the measurements regarding the ISR are leading to the same results results. Therefore they would be worthless.
Interrupts and interrupt service routines. At this point interrupts measurements are presented. The results are available here: • [5]/serial_publisher/isr/20200831_07_33_1598938415_isr.json [5], • [5]/serial_publisher/isr/xxxxxxxxx.xxxx [5]. Figure 7: Interrupts number for Serial As shown on the chart above, different interrupts are: • Interrupt 3: Corresponds to the hard fault. Of course, this is not a hard fault. In NuttX it’s used for context switching • Interrupt 15: Corresponds to the system tick • Interrupt 83: Corresponds to the OTG USB. This correspond to the USB uart console, • Interrupt 87: Corresponds to the USART6 interrupt. It has the highest number because of the transmission happening over UART on each byte transmitted and received.
Interrupts and interrupt service routines. Interrupts may be critical for an RTOS application. Thus analysing them would be a great way to solve some performance issue that could be cause by a faulty sensor’s driver configuration/code. Here the approach taken was to measure the time from the moment the RTOS is entering into the IRQs upper half dispatcher to the moment when the RTOS is leaving the upper half dispatcher. For this purpose, two ctf events were created: • isr_enter: which is in charge of timestamping when the RTOS is entering the dispatcher. The interrupt number is provided, • isr_exit: which is is charge of timestamping when the RTOS is leaving the dispatcher. The interrupt number is provided as well, this can be useful for preemptive interrupts.
Interrupts and interrupt service routines. At this point interrupts measurements are presented. The results are available here: • [5]/ethernet_publisher/isr/20200831_07_32_1598938336_isr.json [5], • [5]/ethernet_publisher/isr/xxxxxxxxx.xxxx [5].
