[Devel] Re: [PATCH 3/6][v5] Define siginfo_from_ancestor_ns()

[Roland McGrath]

Drop the second hunk of the patch (stray whitespace change).

The treatment of SI_ASYNCIO is wrong, as I described in an earlier posting.
(It is indeed a user-generated signal.)  You'll need to avoid using this
function in the path from the USB driver case (interrupt level), since you
can't use SI_ASYNCIO to avoid unsafe use of pid_ns data structures.  If you
do as I suggested in that posting, this is no problem since you just never
do this check in the kill_pid_info_as_uid (perhaps renamed) case.


Thanks,
Roland
_______________________________________________
Containers mailing list
Containers at lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers