[PATCH v2] fork: check exit_signal passed in clone3() call

[Eugene Syromiatnikov <esyr@redhat.com>]

Hello.

After some consideration, I've decided to utilise Oleg's proposal[1]
"(args.exit_signal & ~((u64)CSIGNAL))" as a check. I still don't like
it, as it mixes argument copy check (I'm not sure if it's ever needed,
however, as I'm not sure if there's a reason for exit_signal field
of struct kernel_clone_args to have int type) with argument sanity
check; moreover, it covers only clone3 case, and the code in
copy_process is still error-prone in the long run.  Ideally, the check
should be somewhere in the one place, but as of now this one place
is likely _do_fork, but it's kinda weir to have argument check there
as of now.

Changes since v1[2]:
 - Check changed to comparison against negated CSIGNAL to address
   the bug reported by Oleg[3].
 - Added a comment to _do_fork that exit_signal has to be checked
   by the caller.

[1] https://lkml.org/lkml/2019/9/10/581
[2] https://lkml.org/lkml/2019/9/10/411
[3] https://lkml.org/lkml/2019/9/10/467

Eugene Syromiatnikov (1):
  fork: check exit_signal passed in clone3() call

 kernel/fork.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

-- 
2.1.4