A few topics in one posting.
We have multiple SBC projects based upon circa 2010 netcruzer project code. We are trying to track down some annoying bugs in some of the code branches that manifests in various ways:
1) if the SBC is powered up with an active/ready Ethernet connection already present it will hang completely in the first ~1 second. In this situation power cycling the SBC with Ethernet disconnected, waiting ~5 seconds and then plugged in the Ethernet cable all is well. (SBC configured for dynamic IP)
2) SBC and a *router* power up at the same time. As router boots after ~20 seconds the SBC is seen at default IP of 192.168.1.54. After a further ~30 seconds the routers DHCP server assigns a dynamic address to the SBC. SBC can then become "lost" all attempts to reach it over IP fail, however it's system led is still flashing indicating main cooperative loop is still active (We flash LED in main cooperative loop, not an interrupt).
Are these know issues with older libraries? Anywhere specific to check for a bug or a work around?
Which brings me to my 3rd question as a short term solution for some of these issues may be to force a reboot on hang:
3) Is there some example code for the hardware watchdog feature? and what is the maximum length of time that can be configured for the watchdog before a board reset (we probably want to use 1-10 seconds) ?