On Thu, Dec 20, 2007 at 02:59:13PM +0200, Andriy Gapon wrote:
> P.S. a PR with a terse description is already opened for the above
> behavior of acd:
> http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/118779
This is a regression since RELENG_4 (RELENG_3 was OK),
please take a look at (incorrectly
FreeBSD 6.2-RELEASE-p6 amd64
This is what "cdcontrol eject" performs (incidentally, on a mounted acd
device):
11991 cdcontrol CALL open(0x7fffe130,0,0x9)
11991 cdcontrol NAMI "/dev/acd0"
11991 cdcontrol RET open 3
11991 cdcontrol CALL ioctl(0x3,CDIOCALLOW,0)
11991 cdcontrol RET io
