Re: ipfw: ipfw: Adding record failed: Inappropriate ioctl for device
Date: Sun, 12 Oct 2025 14:06:49 UTC
On Sat, Oct 11, 2025 at 01:49:23PM -0700, David Wolfskill wrote: > ... > g1-48(15.0-S)[87] cat t1 | /sbin/ipfw /dev/stdin > added: 1.0.1.0/24 0 > Line 1: Adding record failed: Inappropriate ioctl for device > g1-48(15.0-S)[88] ipfw table 1 list > 1.0.1.0/24 0 > g1-48(15.0-S)[89] ipfw table 1 flush > g1-48(15.0-S)[90] ipfw table 1 list > g1-48(15.0-S)[91] /sbin/ipfw /tmp/t1 > added: 1.0.1.0/24 0 > added: 1.0.2.0/23 0 > added: 1.0.8.0/21 0 > added: 1.0.32.0/19 0 > added: 1.1.0.0/24 0 > added: 1.1.2.0/23 0 > added: 1.1.4.0/22 0 > added: 1.1.9.0/24 0 > added: 1.1.10.0/23 0 > added: 1.1.12.0/22 0 > ... > So it seems that /sbin/ipfw no longer copes with reading from > /dev/stdin, but is OK reading from a regular file. (I had observed the > same behavior in main-n281059-2d9fd2c573c3, now that I know to look for > it.) > .... I have found a curious (to me) wrinkle: root@g1-48:/tmp # ipfw /dev/stdin table 1 add 1.0.1.0/24 added: 1.0.1.0/24 0 table 1 add 1.0.2.0/23 added: 1.0.2.0/23 0 table 1 add 1.0.8.0/21 added: 1.0.8.0/21 0 table 1 add 1.0.32.0/19 added: 1.0.32.0/19 0 table 1 add 1.1.0.0/24 added: 1.1.0.0/24 0 ... seems to work without a whine -- while (as noted earlier) a pipe fails. (Please note that I am neither using nore referring to ipfw's ability to pipe the commands through a (command-line identified) preprocessor.) Peace, david -- David H. Wolfskill david@catwhisker.org See https://www.catwhisker.org/~david/publickey.gpg for my public key.