Re: Motion 46: finalise interval literals, amendments
On 2013-07-09 21:37:12 -0700, Dmitry Nadezhin wrote:
> I imagine only how a hacker sipplies long digit sequences as DoS attack.
That's the problem (or try to trigger a buffer overflow to breach
the system).
> An implementation may fail if it can't store input stream in memory
> reporting OutOfMemory error.
In practice, several times the size of the input stream with fast
algorithms. But anyway a DoS may occur *before* an OutOfMemory
error, just because too many resources are consumed. This is what
happened with XML entity expansion[*] for instance.
[*] http://seclists.org/oss-sec/2013/q1/391
Here the problem is rather limited, but it still may occur, in
particular if I/O related memory (disk, network...) if much larger
than RAM.
> I don't think that standard should specifiy this.
Some standards do specify limits, e.g. ISO C (5.2.4.1).
If P1788 doesn't specify limits, an implementation aware of this
problem may attempt to fix its own limit, possibly too low compared
to what can be regarded as acceptable.
--
Vincent Lefèvre <vincent@xxxxxxxxxx> - Web: <http://www.vinc17.net/>
100% accessible validated (X)HTML - Blog: <http://www.vinc17.net/blog/>
Work: CR INRIA - computer arithmetic / AriC project (LIP, ENS-Lyon)