FPGA Intellectual Property
PCI Express*, Networking and Connectivity, Memory Interfaces, DSP IP, and Video IP
Intel Support hours are Monday-Fridays, 8am-5pm PST, except Holidays. Thanks to our community members who provide support during our down time or before we get to your questions. We appreciate you!

Need Forum Guidance? Click here
Search our FPGA Knowledge Articles here.
5988 Discussions

Lost Interrupts with Cyclone IV PCIe HIP

Honored Contributor II


I have a problem with receiving Interrupts from an external bus to my host computer via a Cyclone IV GX FPGA. 


Short description of my System: 


- I have an external Bus-System (About 10MHz) with 8 interrupt lines. 

- The Interrupt lines are directly connected to the FPGA. (Well, with Schmidt-Triggers) 

- The Interrupt-Lines are then routed to my QSYS-System which only holds The PCIe HIP and a Interrupt-Bridge with 8 Inputs to route the interrupts to PCIe HIP. 

- The FPGA is connected with PCIe Gen 1, one Lane to an Intel Atom-Processor 

- My OS is Linux Debian 7 

- I wrote a small Kernel Module which counts the Incoming MSIs so I can read the number with <cat /proc/interrupts>. 


Now I generate some Interrupts on my external bus just by pulling the lines to ground with a burst generator (self made on a DE0nano). 

And everything works (nearly) fine. The problem is: about 0,5 to 2 percent of the interrupts get lost on the way. I checked the Interrupt-Signals even on the Interrupt-Port on the PCIe Core with LAI. The Signals don't get lost until they reach the PCIe Core. 

I send only one Interrupt on the external bus in about 500us. I don't think this is too fast, right? 


Do you need further information? Or am I doing something terribly wrong? Is there an Issue with the PCIe Core?
0 Kudos
2 Replies
Honored Contributor II

Anyone out there??????

Honored Contributor II

OK, sorry for that. It was a problematic configuration in BIOS which caused a massive CPU-Load. 

Afterreconfiguring BIOS the Interrupts are functioning in normal parameters.