On Mon, May 04, 2020 at 07:06:33PM -0400, Thomas Walker wrote: > > Thanks, that did get me dumps that I assume are the outgoing packets. I > don't know anything really about NII, but I kind of naively assumed the debug > output (at least of the data block) would look something like the hex dump of > tshark -x but they are nowhere near alike. > Any pointers as to how to compare this with what we see on the wire would be > greatly appreciated. >
Ok, I've been unable to work out what is wrong with the NII driver on these systems, but I *have* been able to work out why the PCI device probing wasn't working correctly and causing iPXE to fall back to NII in the first place. At least the 1G Broadcom (BCM5720) driver then works as expected and is able to pull images via http without issue. I'll cleanup and submit a patch. Does ipxe have any sort of additional contributor agreement? (As this was done on work time, I have to ask and go through the motions to get approval from my employer, etc). Thanks! _______________________________________________ ipxe-devel mailing list ipxe-devel@lists.ipxe.org https://lists.ipxe.org/mailman/listinfo/ipxe-devel
