> Any reason why this shouldn't go into -current? If there aren't any
Yes, it breaks seeking to half of the address space in /dev/mem and
dev/kmem on 64-bit machines.
The maximum value for an lseek() offset depends on the file. I think
POSIX indirectly requires checking against the limit at lseek() time, at
least for regular files, since there is no Standard way for read(2) to
report failure due to an invalid offset (FreeBSD documents returning
EINVAL, but ufs actually returns EFBIG). Applications still need to
be careful about seeking to (off_t)-1 if they actually want to do this,
since (off_t)-1 may be a valid seek offset.
Bruce
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-bugs" in the body of the message