Re: [PATCH] mm/secretmem: use refcount_t instead of atomic_t

[James Bottomley]

On Fri, 2021-08-20 at 12:38 -0400, Jordy Zomer wrote:
> Hi There!
> 
> Because this is a global variable, it appears to be exploitable.
> Either we generate a sufficient number of processes to achieve this
> counter, or you increase the open file limit with ulimit or sysctl.
> Unless the kernel has a hard restriction on the number of potential
> file descriptors that I'm not aware of.

There's no direct global counter for file descriptors, no; however,
there is an indirect limit: the number of processes per user which is
now defaulting to around 65535, so even a fork bomb opening the max
number of fds won't get you a wrap.

> In any case, it's probably a good idea to patch this to make it
> explicitly secure. If you discover a hard-limit in the kernel for
> open file descriptors, please let me know. I'm genuinely interested
> :D!

I didn't disagree it might be a useful think to update ... I just
didn't think it was currently exploitable.

James