Hello, I am investigating the rt patch with linux kernel 2.6.22 (on i386, x86_64 and powerpc). I use a hr timer to generate interrupts. I have three questions: 1. To determine the worst-case period between this interrupt and the start of a real-time user space thread that is waiting for this interrupt I want to use the following formula: period = irq_off_time + preempt_disable_time + soft_irq_handling + sched_latency with irq_off_time measured with CONFIG_CRITICAL_IRQSOFF_TIMING with preempt_disable_time measured with CONFIG_PREEMPT_TIMING with soft_irq_handling the time seen with e.g. LTT with sched_latency measured with cyclictest Is this correct or am I missing something? Is preempt_disable_time including the irq_off_time? 2. Can a page fault (fopen()) in another application influence this period? 3. I tried to measure the CONFIG_CRITICAL_IRQSOFF_TIMING but preempt_max_latency remains 0 (I used the echo 0 >/proc/sys/kernel/preempt_max_latency) What can be wrong? With kind regards, Gerrit Binnenmars - To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
