Re: [Etherboot-developers] [PATCH] [tftp] Kick off TFTP RRQ from a process to avoid losing first packet
Thomas Horsten <[email protected]> Thu, 14 Jan 2010 14:20:31 +0000
| Newsgroups | gmane.network.etherboot.gpxe,gmane.network.etherboot.devel |
|---|---|
| Message-ID | <[email protected]> |
2010/1/14 Stefan Hajnoczi <[email protected]>: > Does this patch fix the issue? =A0It should be equivalent to adding a new= process. Yes, it does fix it. However, it spins an extra time before the socket is ready. With the process approach xfer_window() is true the first time round, and I'm guessing this is because the timers get scheduled ahead of the processes. > There is still an issue because UDP deliver_iob() does not fail when > the packet is dropped due to an ARP cache miss. =A0This means TFTP > cannot determine whether a packet it sent was handed to the network > device. =A0I haven't audited other UDP user code but it is possible we > could change UDP so that users know when packets are hitting the wire > versus waiting for ARP. That would probably be good. Again the wait objects I mentioned earlier, or even a simple callback in the socket structure would be an idea. Thomas