Re: [Qemu-devel] [PATCH 1/2] hmp: expr_unary(): check for overflow in strtoul()

[Luiz Capitulino]

On Thu, 26 Apr 2012 15:34:50 -0600
Eric Blake <address@hidden> wrote:

> On 04/26/2012 03:10 PM, Luiz Capitulino wrote:
> > It's not checked currently, so something like:
> > 
> >   (qemu) balloon -100000000000001111114334234
> >   (qemu)
> > 
> > Will just "work" (in this case the balloon command will get a random
> > value).
> > 
> > Fix it by checking if strtout() overflowed.
> > 
> > Signed-off-by: Luiz Capitulino <address@hidden>
> > ---
> >  monitor.c |    6 ++++++
> >  1 file changed, 6 insertions(+)
> > 
> > diff --git a/monitor.c b/monitor.c
> > index 8946a10..6178f48 100644
> > --- a/monitor.c
> > +++ b/monitor.c
> > @@ -3122,8 +3122,14 @@ static int64_t expr_unary(Monitor *mon)
> >      default:
> >  #if TARGET_PHYS_ADDR_BITS > 32
> >          n = strtoull(pch, &p, 0);
> 
> strtoull and friends are evil.  That's why libvirt has a rule that all
> string-to-integer conversion must go through a libvirt wrapper call, and
> _only_ the wrapper may use strtoull (since that one use has been audited
> for correctness).

We could add a wrapper too, but I'll only fix the bug for now.

> 
> > +        if (n == ULLONG_MAX && errno == ERANGE) {
> 
> Not quite right.  ULLONG_MAX is a valid return, but you did not prime
> errno, so if errno has junk ERANGE from some earlier point in the
> program, you will have a false negative.  You are guaranteed that errno
> is unchanged on success, so prime things by setting errno to 0 before
> calling strtoull.

I thought about doing that but - hope that won't sound stupid - how can I
guarantee that ERANGE is not 0?

> > +            expr_error(mon, "error: too long number");
> 
> Grammar - this reads better as 'number too long'.

Yes, Peter suggested something similar. Will change that.