Re: [PATCH]: exec: avoid propagating PF_NO_SETAFFINITY intouserspace child

[Tejun Heo]

On Thu, Nov 28, 2013 at 03:47:14PM +0100, Peter Zijlstra wrote:
> Bah, windows mentality of we know better.
> 
> If they do stupid they get stupid; the only thing we should be concerned
> about is correctness, they shouldn't be able to crash the system.

It isn't about that we know better but more that we do not want to
leak implementation details to userland.  "Whatever isn't crashing is
okay" doesn't work because we aren't supposed to change behaviors
which are depended upon from userland.  Unless what's exposed is
actively managed, we end up being locked into specific unintended
characteristics of the current implementation.  It hurts both the
kernel and userland.  We end up having to hack around silly
constraints in the future and userland gets surprised when such hacks
inevitably falls apart in corner cases.

So, no, it's not about which part knows better or windows mentaility.
It's about having proper separation between implementation details and
exposed interface.

Thanks.

-- 
tejun
--
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/