Re: [RFC][PATCH 0/6][v3] Container-init signal semantics

[Sukadev Bhattiprolu]

Eric W. Biederman [ebiederm@xxxxxxxxxxxx] wrote:
| I haven't dug in too deep but right now my concern are user space semantics,
| I don't want to wind up with something ugly there because we can not change
| it later.

The one restriction we are imposing is that SIGINT, SIGTERM etc will not
currently kill containter-inits. Only SIGKILL will.  But that is good point.
Maybe we should document that as a limitation we may remove in the future ? 
i.e. Its not a feature that container-inits should rely on.  Like sysV init,
container-init should still SIG_IGN all unhandled signals. If they don't,
they may break in the future.

| 
| So if we can write a description of what happens to signals to cinit
| that is right 100% of the time.  Something we can write a test case
| for that tests all of the corner cases and it always get the same
| results. I am happy.

Yes, I believe we can say that SIGKILL/SIGSTOP from parent are always
delivered and no fatal signal from same ns is.

| 
| I don't mind dropping signals early as an optimization, but if it
| is just an optimization we can't count on it in cinit.

Yes, you have a point. It started out as an optimization, but unwanted
signals are either ignored or dropped _always_ (or we have a bug).

| 
| So I would rather deliver less and make user space deal with it,
| then deliver more cause problems for user space.

The user-semantics appear to be clean now.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/