32 bits is still too small. I wouldn't be comfortable relying on the size of the PID space to avoid collisions until we made it 128 bits or so. I think you're still seriously overestimating the danger of a DoS here: whatever limits apply to forked processes can apply to process handles. Whatever mitigates fork bombs will also mitigate handle-based attacks.
The advantages of process handles outweigh this small risk.
Re:Signalfd is useless
by Dylan16807 on Saturday May 23, 2015 @04:32PM
(#9594275)
In what scenario would you run out of 64 bit PIDs? How many per second for how many centuries?
It's not a matter of running out of PIDs: it's about the probability of accidental collision.
Re:Signalfd is useless
by Dylan16807 on Saturday May 23, 2015 @04:31PM
(#9594271)
A workable limit is trivial, how about 100 zombie process IDs per user.
I don't think a ulimit would be very effective here at preventing denial-of-service. Let's say I set it to 100... I can just have my 100 children each spawn and hold on to 100 children of their own, and so on and so forth. If I just go with a bigger process ID space all these headaches go away, plus existing software works without modification.