When a link-layer address changes (e.g., ifconfig ex0 link
02🇩🇪ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor
Advertisement to update the network-/link-layer address bindings
on our LAN peers.
Refuse a change of ethernet address to the address 00:00:00:00:00:00
or to any multicast/broadcast address. (Thanks matt@.)
Reorder ifnet ioctl operations so that driver ioctls may inherit
the functions of their "class"---ether_ioctl(), fddi_ioctl(), et
cetera---and the class ioctls may inherit from the generic ioctl,
ifioctl_common(), but both driver- and class-ioctls may override
the generic behavior. Make network drivers share more code.
Distinguish a "factory" link-layer address from others for the
purposes of both protecting that address from deletion and computing
EUI64.
Return consistent, appropriate error codes from network drivers.
Improve readability. KNF.
*** Details ***
In if_attach(), always initialize the interface ioctl routine,
ifnet->if_ioctl, if the driver has not already initialized it.
Delete if_ioctl == NULL tests everywhere else, because it cannot
happen.
In the ioctl routines of network interfaces, inherit common ioctl
behaviors by calling either ifioctl_common() or whichever ioctl
routine is appropriate for the class of interface---e.g., ether_ioctl()
for ethernets.
Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In
the user->kernel interface, SIOCSIFADDR's argument was an ifreq,
but on the protocol->ifnet interface, SIOCSIFADDR's argument was
an ifaddr. That was confusing, and it would work against me as I
make it possible for a network interface to overload most ioctls.
On the protocol->ifnet interface, replace SIOCSIFADDR with
SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to
invoke SIOCINITIFADDR.
In ifioctl(), give the interface the first shot at handling most
interface ioctls, and give the protocol the second shot, instead
of the other way around. Finally, let compatibility code (COMPAT_OSOCK)
take a shot.
Pull device initialization out of switch statements under
SIOCINITIFADDR. For example, pull ..._init() out of any switch
statement that looks like this:
switch (...->sa_family) {
case ...:
..._init();
...
break;
...
default:
..._init();
...
break;
}
Rewrite many if-else clauses that handle all permutations of IFF_UP
and IFF_RUNNING to use a switch statement,
switch (x & (IFF_UP|IFF_RUNNING)) {
case 0:
...
break;
case IFF_RUNNING:
...
break;
case IFF_UP:
...
break;
case IFF_UP|IFF_RUNNING:
...
break;
}
unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and
#ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4).
In ipw(4), remove an if_set_sadl() call that is out of place.
In nfe(4), reuse the jumbo MTU logic in ether_ioctl().
Let ethernets register a callback for setting h/w state such as
promiscuous mode and the multicast filter in accord with a change
in the if_flags: ether_set_ifflags_cb() registers a callback that
returns ENETRESET if the caller should reset the ethernet by calling
if_init(), 0 on success, != 0 on failure. Pull common code from
ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(),
and register if_flags callbacks for those drivers.
Return ENOTTY instead of EINVAL for inappropriate ioctls. In
zyd(4), use ENXIO instead of ENOTTY to indicate that the device is
not any longer attached.
Add to if_set_sadl() a boolean 'factory' argument that indicates
whether a link-layer address was assigned by the factory or some
other source. In a comment, recommend using the factory address
for generating an EUI64, and update in6_get_hw_ifid() to prefer a
factory address to any other link-layer address.
Add a routing message, RTM_LLINFO_UPD, that tells protocols to
update the binding of network-layer addresses to link-layer addresses.
Implement this message in IPv4 and IPv6 by sending a gratuitous
ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD
messages on a change of an interface's link-layer address.
In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address
that is broadcast/multicast or equal to 00:00:00:00:00:00.
Make ether_ioctl() call ifioctl_common() to handle ioctls that it
does not understand.
In gif(4), initialize if_softc and use it, instead of assuming that
the gif_softc and ifp overlap.
Let ifioctl_common() handle SIOCGIFADDR.
Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels
that certain invariants on a struct route are satisfied.
In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit
about the ioctls that we do not allow on an agr(4) member interface.
bzero -> memset. Delete unnecessary casts to void *. Use
sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with
NULL instead of "testing truth". Replace some instances of (type
*)0 with NULL. Change some K&R prototypes to ANSI C, and join
lines.
NCR 53C80/53C400 driver
BACKGROUND
----------
The NCR 53C80 SCSI Bus Controller (SBC) is an early single-chip solution
which formed the basis of many early SCSI host adapters for both the
i386 and m68k platforms. The NCR 53C400 is a slightly more advanced
chip which retains backward compatibility with the 53C80.
On the PC, the NCR 53C80 was most commonly used to implement simple, cheap
SCSI host adapters that were bundled with tape and CD-ROM drives. Since
these controllers were not bus-mastering (and in some cases were not even
interrupt-driven), they (like IDE adapters) required the CPU to perform
much of the actual processing. These days, these controllers are cheap
and plentiful since many are not supported by Windows 95.
Similarly, NetBSD, although it has had an MI 53C80 driver (used by the
Sun3 and Mac68k ports) for some time, has not had a i386 driver.
Until now, that is...
OVERVIEW
--------
The NCR 53C80/53C400 driver (the 'nca' device) consists of two pieces:
1) Patches for the 53C80 MI driver to make it use bus_space()
functions. (This requires an optional define. By default,
the driver will compile in "legacy" memory-mapped mode.
2) A machine-dependent driver (nca) containing probe and
attachment routines.
This driver has bene tested with the following adapters:
NCS-250 (Chinon) 53C80, port-mapped, polled-mode
(This is used in my primary development
box to drive an external Zip drive.)
Sumo SCSI-AT 53C80, port-mapped, interrupt driven
(Note: This is an odd card in that its
own firmware seems to have trouble detecting
attached drives. Under NetBSD, however,
it operates with no problems.)
Trantor T-160 53C400, port-mapped, interrupt driven
This card was often bundled with NEC
CD-ROM drives. (My standalone test box
is using this as its primary adapter.)
DTC 3150V 53C400, memory-mapped, interrupt driven
This a simple card designed to drive
a CD-ROM.
CONFIGURATION
-------------
To setup the nca driver, the configuration file must contain the following:
options NCR5380_USE_BUS_SPACE
This line is required to add bus_space() compatibility to the MI driver.
Next you need to add one or more configuration lines for the nca devices:
nca0 at isa? port 0x360 irq 15
nca1 at isa? iomem 0xd8000 irq 5
The first is for a port-mapped controller at 0x360, IRQ 15. The second line
is for a memory-mapped controller (Trantor T128 or equivalent) at
0xd800-0xdff, IRQ 5.
You can also set up the driver in "polled" mode (i.e., no interrupts) by
leaving off the "irq" portion of the line:
nca0 at isa? port 0x360
nca1 at isa? iomem 0xd8000
Lastly, you need to add a scsibus attachment line for the nca device:
scsibus* at nca?
The following is the probe output from my test system:
Copyright (c) 1996, 1997, 1998
The NetBSD Foundation, Inc. All rights reserved.
Copyright (c) 1982, 1986, 1989, 1991, 1993
The Regents of the University of California. All rights reserved.
NetBSD 1.3.2 (GENERIC) #2: Sun Oct 4 17:11:43 EDT 1998
root@hefalump:/usr/src/sys/arch/i386/compile/GENERIC
cpu0: Intel 486DX (486-class)
real mem = 7995392
avail mem = 5349376
using 123 buffers containing 503808 bytes of memory
mainbus0 (root)
isa0 at mainbus0
com1 at isa0 port 0x2f8-0x2ff irq 3: ns8250 or ns16450, no fifo
com2 at isa0 port 0x3e8-0x3ef irq 5: ns8250 or ns16450, no fifo
lpt0 at isa0 port 0x378-0x37b irq 7
nca0 at isa0 port 0x360-0x36f irq 15
nca0: NCR 53C400 detected
scsibus0 at nca0: 8 targets
sd0 at scsibus0 targ 0 lun 0: <HP, C2235, 0B11> SCSI2 0/direct fixed
sd0: 402MB, 1574 cyl, 9 head, 58 sec, 512 bytes/sect x 825012 sectors
cd0 at scsibus0 targ 6 lun 0: <CHINON, CD-ROM CDS-535, Q20> SCSI2 5/cdrom removable
nca1 at isa0 iomem 0xdb878-0xdb887 irq 5
nca1: NCR 53C400 detected
scsibus1 at nca1: 8 targets
sd1 at scsibus1 targ 5 lun 0: <IOMEGA, ZIP 100, J.02> SCSI2 0/direct removable
sd1: 96MB, 96 cyl, 64 head, 32 sec, 512 bytes/sect x 196608 sectors
npx0 at isa0 port 0xf0-0xff: using exception 16
pc0 at isa0 port 0x60-0x6f irq 1: color
pc0: console
fdc0 at isa0 port 0x3f0-0x3f7 irq 6 drq 2
fd0 at fdc0 drive 0: 1.44MB, 80 cyl, 2 head, 18 sec
biomask 8060 netmask 8460 ttymask 84e2
boot device: sd0
root on sd0a dumps on sd0b
root file system type: ffs
In this output, nca0 is a Trantor T-160 and nca1 is a DTC 3150V. Both happen
to be 53C400-based controllers.
LIMITATIONS
-----------
As of this writing, the nca driver has two known limitations:
1) No DMA or pseudo-DMA support
This is unfortunate, but may be remedied in a later release. I would welcome
any help by someone more familiar with DMA, particularly in relation to
bus_space().
As it is, however, performance of the nca driver is acceptable, though some
of that may depend on one's definition of "acceptable". Remember that these
were not high speed controller under the best conditions, so much of it is
really the nature of the beast. It should be adequate for tapes, CD-ROMS,
and low-usage disk devices (e.g., Zip drives). If you want to drive a CD-R
drive, then invest in an Adaptec 154X or a PCI controller.
2) No support for the SCSI port of the Pro AudioStudio 16.
This is also unfortunate and may not be able to be remedied withing the
current framework of the bus_space() functions and the nca driver.
The problem is this: In most adapters, the eight 53C80 registers are mapped
to eight sequential locations, either ports or memory addresses. On the
PAS-16, however, the registers are mapped to two sets of ports- four
sequential ports at the base address and four sequential ports located
0x2000 higher. As I currently understand it, this is not supportable by
the current bus_space() implementation nor is it possible for the driver
to allocate a second bus_space_tag and _handle itself to accommodate the
second set of ports. Without either, it is very difficult to imagine how
a portable linkage to the MI driver could be made.
Again, I welcome suggestions.
HISTORY
-------
An nca driver first appeared in FreeBSD.
This particular one borrows a little code from it and some from the i386
'esp' and sun3 'si' drivers. It, like many things in the free unix world,
was written because it solved a problem- mine! In my case, it was a need
of a SCSI card and a lack of IRQs. The good news was that I had one
(NCS-250); the bad news was that it was not supported under NetBSD. The
rest is history.
DISCLAIMER
----------
Like most things, you should take this code with a grain of salt. I have
tried to test it sufficiently, but it is always possible that it is not
compatible with some aspect of your system. If you end up suffering
massive data loss and destruction, you have my sympathies, but I do not
and will not allow myself to be held responsible.
CREDITS
-------
My thanks to Jason Thorpe and the rest of the NetBSD team for making it
so easy to write this driver. My thanks also to the authors of the
FreeBSD nca driver for inspiration and 53C400 support.
In the end, I hope that someone else can find this driver as useful as I
have. If so, please drop me a line at jruschme@exit109.com and let me
know about it.
Share and enjoy
John Ruschmeyer (jruschme@exit109.com)
11 October 1998
de87fe677d