On Thu, 27 Feb 1997 18:51:15 +0100,
Hauke Fath <hauke@Espresso.Rhein-Neckar.DE> wrote:
[...]
> nubus0 at mainbus0
> probing slot 9, first probe at 0x0xf9ffffff (PA 0x0).
> bytelanes of 0xa5 found for slot 0x9.
> fmt->top is 0x92b9000, that minus 0x14 puts us at 0x92b8fd8.
> Directory offset 0xffffff1c     Length 0xf8     CRC 0xe6177f0b
> Revision level 0x1      Format 0x1      Test Pattern 0x5a932bc7
> ae0 at nubus0ae0: failed to clear shared memory - check configuration
> ae0: failed to clear shared memory - check configuration
> ae0: failed to clear shared memory - check configuration
> ae0: failed to clear shared memory - check configuration
> ae0: failed to clear shared memory - check configuration

I'm getting the following result with my Kinetics EtherPort SE/30, from
a kernel built from the source tree supped on Mar. 2 (note that the "ae0"
line reports the card has 16KB memory whereas it really has only 8KB):

| NetBSD 1.2C (GENERIC) #0: Tue Mar  4 01:57:35 JST 1997
|     kenn@gon:/usr/src/sys/arch/mac68k/compile/GENERIC
| Apple Macintosh SE/30  (68030)
| real mem = 20971520
| avail mem = 17117184
| using 204 buffers containing 835584 bytes of memory
| mrg: 'Mac II class ROMs' ROM glue, tracing off, debug off, silent traps
| mrg: I/O map kludge for ROMs that use hardware addresses directly.
| adb: bus subsystem
| \^H\^Hadb: keyboard at 2
| adb: 100 dpi mouse at 3
| adb: extended mouse <LT01> 3-button 200 dpi trackball at 15
| mainbus0 (root)
| obio0 at mainbus0
| adb0 at obio0 (ADB event device)
| asc0 at obio0: Apple Sound Chip
| ncrscsi0 at obio0
| scsibus0 at ncrscsi0: 8 targets
| sd0 at scsibus0 targ 0 lun 0: <FUJITSU, M2694ES-512, 812A> SCSI2 0/direct fixed
| sd0: 1033MB, 1819 cyl, 15 head, 77 sec, 512 bytes/sec
| sd1 at scsibus0 targ 1 lun 0: <QUANTUM, FIREBALL1080S, 1Q04> SCSI2 0/direct fixed
| sd1: 1042MB, 3835 cyl, 4 head, 139 sec, 512 bytes/sec
| zsc0 at obio0 chip type 0 
| zstty0 at zsc0 channel 0
| zstty1 at zsc0 channel 1
| nubus0 at mainbus0
| ae0 at nubus0: address 08:00:89:b2:94:22, type EtherPort SE/30, 16KB memory
| macvid0 at nubus0: Macintosh SE/30 Internal Video
| macvid0: 512 x 342, monochrome
| grf0 at macvid0
| ite at grf0 not configured
| fpu0 at mainbus0 (mc68882)
| boot device: sd0
| root on sd0a swap on sd0b
| PRAM: 0x331ab26a, macos_boottime: 0x331ab257.
| root file system type: ffs
| ae0: length does not match next packet pointer
| ae0: len bf99 nlen 5399 start 0c first 2b curr 2c next 7f stop 40
| ae0: NIC memory corrupt - invalid packet length 21401
| ae0: length does not match next packet pointer
| ae0: len 8a8c nlen 0e8c start 0c first 25 curr 26 next 34 stop 40
| ae0: NIC memory corrupt - invalid packet length 3724
| ae0: length does not match next packet pointer
| ae0: len bc68 nlen 8168 start 0c first 29 curr 2a next ab stop 40
| ae0: NIC memory corrupt - invalid packet length 33128
| ae0: length does not match next packet pointer
| ae0: len 249a nlen 929a start 0c first 2a curr 2b next bd stop 40
| ae0: NIC memory corrupt - invalid packet length 37530
| ae0: length does not match next packet pointer
| ae0: len acda nlen 7cda start 0c first 28 curr 29 next a5 stop 40
| ae0: NIC memory corrupt - invalid packet length 31962
| ae0: length does not match next packet pointer
| ae0: len 91ed nlen 13ed start 0c first 2a curr 2b next 0a stop 40
| ae0: NIC memory corrupt - invalid packet length 5101
| ae0: length does not match next packet pointer
| ae0: len 9965 nlen 3865 start 0c first 22 curr 23 next 5b stop 40
| ae0: NIC memory corrupt - invalid packet length 14437
| ae0: length does not match next packet pointer
| ae0: len 5037 nlen 9337 start 0c first 21 curr 22 next b5 stop 40
| ae0: NIC memory corrupt - invalid packet length 37687
| ae0: length does not match next packet pointer
| ae0: len ebf7 nlen d1f7 start 0c first 23 curr 24 next f5 stop 40
| ae0: NIC memory corrupt - invalid packet length 53751
| ae0: length does not match next packet pointer
| ae0: len 9cca nlen c1ca start 0c first 23 curr 24 next e5 stop 40
| ae0: NIC memory corrupt - invalid packet length 49610
| 

It works, though.  I took a peek at if_ae_nubus.c, machdep.c, bus.h,
etc, but I'm not sure how to fix it w/o breaking anything else.

Ken