So I’ve been at this a while and I have been making steady progress on my project however I seem to keep running into issues with the board’s tool-chain. With NetFPGA release 3.0.1 they moved their testing suite to python which is nice, except that it doesn’t seem like the documentation has nearly caught up yet. Fortunately, I’ve already had considerable experience with python so stepping through the code is fairly easy.
Platform
Some of the issues I ran into were both platform and software based. I just moved the board into a new computer this past weekend since I needed more PCI-E slots to fit some additional Ethernet modules for regression tests. This went fairly smoothly but when I booted, I only saw one of the devices. So I go into the bios and see nothing about PCI devices or Ethernet devices (besides boot order) so I feel out of luck there. Just before I’m at the point where I think I need to flash an update, I find one single field in a remote menu which the description reads that you an modify the bandwidth of certain slots. So I go ahead and increase the bandwidth on the x8/x16 slots in exchange for disabling the x1 slots entirely. Sure enough this fixed the issue.
Software
Next I was having trouble with getting the regression tests working with the board. Going off the NetFPGA site, it’s still somewhat hard to tell which scripts you’re supposed to be using since all the documentation has the different versions of the NetFPGA somewhat mixed up with notes here and there mentioning small things changing from version to version. Anyway, after I found I was supposed to be using a python script when I was using the Perl script, I was able to work with that. It didn’t really work the first run through, but after I stepped through the code with pdb, it was clear what was going wrong. (For those interested, it was nf_test.py and the Global Setup test was failing because the output was mismatched, even though the setup itself succeeded.)
You can never read too much
Anyway, I am in no means bashing on the documentation of either the NetFPGA maintainers or the motherboard manufacturers for these issues. Docs are difficult to maintain, and software often moves faster than docs. However I continue to be reminded that thorough reading will often save you a lot more time than trying to fix things outright using your own intuition.
Edit:
I forgot to mention that when I initially did the selftest on the NetFPGA, there were issues with the test passing the DMA test after moving it to the new computer (it worked fine on the last one). I wasn’t actually able to track down the specific issue, but by swapping out the patch cables, I was able to get the test passing. I’m not sure why the cables would affect DMA, but that’s something to investigate later.