On Wed, May 31, 2017 at 9:50 AM, Majd Dibbiny <majd@xxxxxxxxxxxx> wrote: > Is this the same issue you were facing with rdma-cm segfaults? The '0' GUID was causing a segfault in rping, udaddy, ib_send_bw in rdma_cm mode, etc. When both nodes have a GUID, it works fine. It could be argued that it should be a bit more helpful and exit with a nice message rather than a segfault. It also could be argued that a random GUID/MAC should be assigned at VF creation time and if you want something static to go through the hoops (it sounds like a good default so it works out of the box, I even though I thought I read that was the behavior but I can't find that document again). I can work around it by adding the random generation to my udev rules, but it is not convenient. I think I'll have to check if a MAC is already assigned or I could get into a loop, I'll have to look at it more later. I need to get source address binding working in iSER first. Thanks again, ---------------- Robert LeBlanc PGP Fingerprint 79A2 9CA4 6CC4 45DD A904 C70E E654 3BB2 FA62 B9F1 -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
