Application Acceleration With FPGAs
Programmable Acceleration Cards (PACs), DCP, DLA, Software Stack, and Reference Designs
Announcements
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.
425 Discussions

'aocl program ...' cause system lockdown on Intel PAC D5005 with OpenCL

HKim27
New Contributor I
298 Views

Hi, all.

 

Problem scenario:

 

1) After cold reboot, FPGA is uninitialized.

$ aocl diagnose ... Physical Dev Name Status Information pac_ec00000 Uninitialized ...

2) Program arbitrary bitstream. (e.g., vector add) It always succeeds.

$ aocl program acl0 $OPAE_PLATFORM_ROOT/opencl/vector_add.aocx ... SUCCESS ... $ $OPAE_PLATFORM_ROOT/opencl/exm_opencl_vector_add_x64_linux/vector_add/bin/host ... PASSED ...

3) Program another bitstream. (e.g., hello_world) Now, the whole system stops(ssh disconnects, tty dies, ...) and automatically reboots after minutes.

$ aocl program acl0 $OPAE_PLATFORM_ROOT/opencl/hello_world.aocx (system dies)

4) Although the system dies, bitstream seems to be successfully programmed. I can run hello_world application without partial reconfiguration time (several seconds).

 

My guess: During partial reconfiguration, FPGA emits wrong signal (or jams) on PCIe bus, causing system down. Any solution to this?

0 Kudos
0 Replies
Reply