Description of problem: Using both 2.0 redboot and current CVS trunk, redboot on viper is unreliable over the net. Loading via GDB by itself seems okay, but it doesn't take much more traffic for things to go badly wrong. When watching with tcpdump, it seems things quickly get stuck in a rut with the viper sending a packet, and the host responding with an ack, but the viper waiting 1 sec and resending anyway. Verified with both Linux and NT hosts so it's not a host stack problem. I need to look closer at this, but need the bug entered for 2.0 known problems...
This bug has moved to http://bugs.ecos.sourceware.org/show_bug.cgi?id=90395