Problem with large numbers of bpf devices

[Kevin Oberman]

A STABLE (not updated since 12/13/02) system shows very strange behavior
for devices with a minor mode greater than 256. tcpdump does not work
reliably and ls is, uh, "strange", using hex for the minor mode.

crw-------   1 root  wheel      23, 252 Apr 11 14:29 bpf252
crw-------   1 root  wheel      23, 253 Apr 11 14:29 bpf253
crw-------   1 root  wheel      23, 254 Apr 11 14:29 bpf254
crw-------   1 root  wheel      23, 255 Apr 11 14:29 bpf255
crw-------   1 root  wheel      23, 0x00010000 Apr 11 14:29 bpf256
crw-------   1 root  wheel      23, 0x00010001 Apr 11 14:29 bpf257
crw-------   1 root  wheel      23, 0x00010002 Apr 11 14:29 bpf258
crw-------   1 root  wheel      23, 0x00010003 Apr 11 14:29 bpf259
crw-------   1 root  wheel      23,  26 Apr 11 14:29 bpf26
crw-------   1 root  wheel      23, 0x00010004 Apr 11 14:29 bpf260
crw-------   1 root  wheel      23, 0x00010005 Apr 11 14:29 bpf261
crw-------   1 root  wheel      23, 0x00010006 Apr 11 14:29 bpf262
crw-------   1 root  wheel      23, 0x00010007 Apr 11 14:29 bpf263
crw-------   1 root  wheel      23, 0x00010008 Apr 11 14:29 bpf264

The minor mode is reported as "0xnnnnnnnn" on all device numbers beyond
256.

Is this problem with ls(1) only, or is there a one byte limit built in
somewhere?

Thanks,

R. Kevin Oberman, Network Engineer
Energy Sciences Network (ESnet)
Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab)
E-mail: oberman at es.net			Phone: +1 510 486-8634