Copyright (c) 1986 Regents of the University of California.
All rights reserved. The Berkeley software License Agreement
specifies the terms and conditions for redistribution.

@(#)np.4 6.3 (Berkeley) 05/21/86

C 5 np - Interlan Np100 10 Mb/s Ethernet interface "device np0 at uba0 csr 166000 vector npintr" The NP device provides access to an Interlan Np100 Ethernet interface for control functions.

This interface is unusual in that it requires loading firmware into the controller before it may be used as a network link-level interface. This is accomplished by opening a character special device, and writing data to it. It is also possible to do post-mortem debugging of firmware failures by reading the local memory of the device.

A program to load the image is provided in /usr/src/new/np100. The sequence of commands would be:

 # ./npload np.image [/dev/np00]
 # sleep 10
 # ifconfig ix0 ...

Multiple control processes are allowed by opening separate minor devices; secondary interfaces are specified by shifting the interface number by 4 bits.

The device also responds to commands passed through the driver by the following ioctls:

NPRESET

kills off all active network processes.

NPSTART

begins execution of the board at the specified address (usually 0x400).

NPNETBOOT

downloads the image from a server on the network. [Contact MICOM-INTERLAN for details.]

np%d: Bad Maintenance command: %x!. An invalid ioctl was passed to the np driver.

np%d: Panic NP100 bad buffer chain. An error occurred in an read or write operation causing it to run out of buffers before it finished the operation. This indicates a kernel failure rather than a device failure.

NP100 unit %d not found!. A failure occurred during initialization, such that the unibus address expected for the board was found to be bad. Probably indicates hardware problems with the board, as do the following:

NP100 Unit %d timed out!

NP100 Unit %d Failed diagnostics!
Status from CSR0: %x.

Panic from NP100 unit %d!\\nPanic Message: %s. An occurrence on the board was deemed serious enough to have the vax print it out.

NP100 unit #%d available!. The board was successfully loaded and started.

np%d: Bad Req: %x.. The board made a maintenance request to the vax that it did not understand.

np%d: No more room on Command Queue!. The np driver allowed an internal resource to be exhausted. This should never happen.

There are 110 other diagnostic messages that can be enabled by setting bits in a debugging mask. Consult the driver for details.

intro(4N), inet(4F), arp(4P), ix(4)