- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi
I`m trying to perform a benchmark to get an idea how much overhead FreeRTOS brings when compared to Altera library,roughly. For this purpose i have implemented two tests which uses timer interrupts. Test 1: uses alt_alarm_start to start an alarm timer and toggles a output pin when alt_avalon_timer_sc_irq enters. And timeout_pulse is enabled in fgpda hw desing so I know when interrupt occurs and when its handled. Measuring via scope yields 2.02us Test 2: uses xTimerCreate start an timer and toggles a output pin when alt_irq_handler enters. It`s being called from port_asm.S. And timeout_pulse is enabled in fgpda hw desing so I know when interrupt occurs and when its handled. Measuring via scope yields 1.27us Problem is I would expect that test1 would have significant smaller delta when compared to Test2 due to FreeRTOS overhead but apparently results do indicate opposite. So I thought may be alt_irq_handler is too late to start of measurement. Do you have any idea where are interrupts first handled when FreeRtos is present ? Thanks in advanceLink Copied
1 Reply
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi again
Apparently I have a mistake in my code. After I corrected that , I can observe test2 =2.10us![](/skins/images/C4F7392A5C68ACA60290C9576F107044/responsive_peak/images/icon_anonymous_message.png)
Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page