Skip to contentSkip to conversation

Renumber AUDIT_TTY_[GS]ET

PATCH In review Al Viro 2 participants 3 messages Aug 23, 2007 Audit

lore
Share & tools
lore

Al Viro

author
#0 lore raw

PATCH 1/1 Renumber AUDIT_TTY_[GS]ET

From: Miloslav Trmac <[email protected]>

Renumber AUDIT_TTY_[GS]ET to avoid a conflict with netlink message types already used in the wild.

---
Please, apply. That really ought to go in before .final ;-/

 audit.h |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
include/linux/audit.h +2 -2
index 4bbd860..d6579df 100644
@@ -63,8 +63,8 @@
 #define AUDIT_ADD_RULE		1011	/* Add syscall filtering rule */
 #define AUDIT_DEL_RULE		1012	/* Delete syscall filtering rule */
 #define AUDIT_LIST_RULES	1013	/* List syscall filtering rules */
-#define AUDIT_TTY_GET		1014	/* Get TTY auditing status */
-#define AUDIT_TTY_SET		1015	/* Set TTY auditing status */
+#define AUDIT_TTY_GET		1016	/* Get TTY auditing status */
+#define AUDIT_TTY_SET		1017	/* Set TTY auditing status */
 
 #define AUDIT_FIRST_USER_MSG	1100	/* Userspace messages mostly uninteresting to kernel */
 #define AUDIT_USER_AVC		1107	/* We filter this differently */
#1 lore raw
Quoted 4 lines · On Thu, 23 Aug 2007 10:19:53 +0100 Al Viro <[email protected]> wrote:

From: Miloslav Trmac <[email protected]>

Renumber AUDIT_TTY_[GS]ET to avoid a conflict with netlink message types already used in the wild.

which are these? Were they in mainline or -mm before?

also, if this is so sensitive, at least add something that says to never ever use 1014/1015...

author
#2 lore raw
Quoted 11 lines · Arjan van de Ven
Quoted 4 lines · On Thu, 23 Aug 2007 10:19:53 +0100 Al Viro <[email protected]> wrote:

From: Miloslav Trmac <[email protected]>

Renumber AUDIT_TTY_[GS]ET to avoid a conflict with netlink message types already used in the wild.

which are these? Were they in mainline or -mm before?

also, if this is so sensitive, at least add something that says to never ever use 1014/1015...

AUDIT_TTY_*ET went into mainline directly and I'd missed the conflict with the numbers already in use by audit-subtree patches. I *had* noticed when pushing audit-subtree to -mm, swore at myself and renumbered those. Alas, there's userland to deal with.
* auditctl has subtree support merged
* auditctl still doesn't have tty... support merged
* it's built with numbers matching the original patchset (1014/1015)
and binaries are released
* audit-subtree patchset is in RH beta kernel, also released

Vote of auditctl maintainer had been "renumber tty_...* until it's too late, since that's the one userland doesn't support yet". The author of tty patches agrees (see above) and I also think that this is the best way out of that mess.

Again, it's largely my fault - I'd missed the conflict in the first place and hadn't reacted immediately.