RE: [PATCH] x86/entry/64: randomize kernel stack offset upon syscall
From: Reshetova, Elena
Date: Thu May 02 2019 - 04:07:49 EST
> From: Reshetova, Elena
> > Sent: 30 April 2019 18:51
> ...
> > I guess this is true, so I did a quick implementation now to estimate the
> > performance hits.
> > Here are the preliminary numbers (proper ones will take a bit more time):
> >
> > base: Simple syscall: 0.1761 microseconds
> > get_random_bytes (4096 bytes per-cpu buffer): 0.1793 microsecons
> > get_random_bytes (64 bytes per-cpu buffer): 0.1866 microsecons
> >
> > It does not make sense to go less than 64 bytes since this seems to be
> > Chacha20 block size, so if we go lower, we will trash useful bits.
> > You can go even higher than 4096 bytes, but even this looks like
> > okish performance to me.
> >
> > Below is a snip of what I quickly did (relevant parts) to get these numbers.
> > I do initial population of per-cpu buffers in late_initcall, but
> > practice shows that rng might not always be in good state by then.
> > So, we might not have really good randomness then, but I am not sure
> > if this is a practical problem since it only applies to system boot and by
> > the time it booted, it already issued enough syscalls that buffer gets refilled
> > with really good numbers.
> > Alternatively we can also do it on the first syscall that each cpu gets, but I
> > am not sure if that is always guaranteed to have a good randomness.
> ...
> > +unsigned char random_get_byte(void)
> > +{
> > + struct rnd_buffer *buffer = &get_cpu_var(stack_rand_offset);
> > + unsigned char res;
> > +
> > + if (buffer->byte_counter >= RANDOM_BUFFER_SIZE) {
> > + get_random_bytes(&(buffer->buffer), sizeof(buffer->buffer));
> > + buffer->byte_counter = 0;
> > + }
> > +
> > + res = buffer->buffer[buffer->byte_counter];
> > + buffer->buffer[buffer->byte_counter] = 0;
> > + buffer->byte_counter ++;
> > + put_cpu_var(stack_rand_offset);
> > + return res;
> > +}
> > +EXPORT_SYMBOL(random_get_byte);
>
> You'll almost certainly get better code if you copy buffer->byte_counter
> to a local 'unsigned long' variable.
>
Noted, will fix, thank you for the suggestion!
I haven't yet worked on this code for a "proper version", so I think many
things would need polishing for both speed and code size.
Best Regards,
Elena.