Most system utilities that work with jails allow the jail to be specified by name or ID, but a few are still ID-only. Use jail_getid(3) to fix the laggards.
Details
Details
- Reviewers
eadler - Group Reviewers
manpages - Commits
- rS336040: MFC r335921:
rS336039: MFC r335921:
rS335921: Allow jail names (not just IDs) to be specified for: cpuset(1), ipfw(8),
Diff Detail
Diff Detail
- Repository
- rS FreeBSD src repository - subversion
- Lint
Lint Not Applicable - Unit
Tests Not Applicable
Event Timeline
Comment Actions
I had a memory of some utilities supporting any or -1 for "in a jail" (i.e. jid != 0). Might be worth looking at doing this too
| lib/libugidfw/ugidfw.c | ||
|---|---|---|
| 615 ↗ | (On Diff #44578) | I'm not a fan of this, though I understand its a copy |
| sbin/ipfw/ipfw.8 | ||
|---|---|---|
| 1540 ↗ | (On Diff #44578) | suggestion: rephrase prison to jail, and use whose (change prison in surrounding lines, too) |
| lib/libugidfw/ugidfw.c | ||
|---|---|---|
| 615 ↗ | (On Diff #44578) | Yes, it's kind of ugly (same ugliness in libjail). But it's a limitation of struct iovec, which is geared toward bi-directional traffic even though the parameter names are always read-only. I didn't want to make a new data type, or to copy every name that gets passed to jail_get(2) or jail_set(2). |
| sbin/ipfw/ipfw.8 | ||
| 1540 ↗ | (On Diff #44578) | Sounds good - "prison" is now "jail". Except I went with "jail whose ID or name is" instead of "jail whose jail ID or name is" |