From: Aleksandr Nogikh <[email protected]>
Currently it is possible to craft a special netlink RTM_NEWQDISC
command that can result in jitter being equal to 0x80000000. It is
enough to set the 32 bit jitter to 0x02000000 (it will later be
multiplied by 2^6) or just set the 64 bit jitter via
TCA_NETEM_JITTER64. This causes an overflow during the generation of
uniformly distributed numbers in tabledist(), which in turn leads to
division by zero (sigma != 0, but sigma * 2 is 0).
The related fragment of code needs 32-bit division - see commit
9b0ed89 ("netem: remove unnecessary 64 bit modulus"), so switching to
64 bit is not an option.
Fix the issue by keeping the value of jitter within the range that can
be adequately handled by tabledist() - [0;INT_MAX]. As negative std
deviation makes no sense, take the absolute value of the passed value
and cap it at INT_MAX. Inside tabledist(), switch to unsigned 32 bit
arithmetic in order to prevent overflows.
Signed-off-by: Aleksandr Nogikh <[email protected]>
Reported-by: [email protected]
---
v2:
* Capping the value when receiving it from the userspace instead of
checking it each time when a new skb is enqueued.
v1:
http://lkml.kernel.org/r/[email protected]
---
net/sched/sch_netem.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/net/sched/sch_netem.c b/net/sched/sch_netem.c
index 84f82771cdf5..0c345e43a09a 100644
--- a/net/sched/sch_netem.c
+++ b/net/sched/sch_netem.c
@@ -330,7 +330,7 @@ static s64 tabledist(s64 mu, s32 sigma,
/* default uniform distribution */
if (dist == NULL)
- return ((rnd % (2 * sigma)) + mu) - sigma;
+ return ((rnd % (2 * (u32)sigma)) + mu) - sigma;
t = dist->table[rnd % dist->size];
x = (sigma % NETEM_DIST_SCALE) * t;
@@ -812,6 +812,10 @@ static void get_slot(struct netem_sched_data *q, const struct nlattr *attr)
q->slot_config.max_packets = INT_MAX;
if (q->slot_config.max_bytes == 0)
q->slot_config.max_bytes = INT_MAX;
+
+ /* capping dist_jitter to the range acceptable by tabledist() */
+ q->slot_config.dist_jitter = min_t(__s64, INT_MAX, abs(q->slot_config.dist_jitter));
+
q->slot.packets_left = q->slot_config.max_packets;
q->slot.bytes_left = q->slot_config.max_bytes;
if (q->slot_config.min_delay | q->slot_config.max_delay |
@@ -1037,6 +1041,9 @@ static int netem_change(struct Qdisc *sch, struct nlattr *opt,
if (tb[TCA_NETEM_SLOT])
get_slot(q, tb[TCA_NETEM_SLOT]);
+ /* capping jitter to the range acceptable by tabledist() */
+ q->jitter = min_t(s64, abs(q->jitter), INT_MAX);
+
return ret;
get_table_failure:
base-commit: 1c86f90a16d413621918ae1403842b43632f0b3d
--
2.29.0.rc2.309.g374f81d7ae-goog
On Wed, 28 Oct 2020 17:07:31 +0000
Aleksandr Nogikh <[email protected]> wrote:
> From: Aleksandr Nogikh <[email protected]>
>
> Currently it is possible to craft a special netlink RTM_NEWQDISC
> command that can result in jitter being equal to 0x80000000. It is
> enough to set the 32 bit jitter to 0x02000000 (it will later be
> multiplied by 2^6) or just set the 64 bit jitter via
> TCA_NETEM_JITTER64. This causes an overflow during the generation of
> uniformly distributed numbers in tabledist(), which in turn leads to
> division by zero (sigma != 0, but sigma * 2 is 0).
>
> The related fragment of code needs 32-bit division - see commit
> 9b0ed89 ("netem: remove unnecessary 64 bit modulus"), so switching to
> 64 bit is not an option.
>
> Fix the issue by keeping the value of jitter within the range that can
> be adequately handled by tabledist() - [0;INT_MAX]. As negative std
> deviation makes no sense, take the absolute value of the passed value
> and cap it at INT_MAX. Inside tabledist(), switch to unsigned 32 bit
> arithmetic in order to prevent overflows.
>
> Signed-off-by: Aleksandr Nogikh <[email protected]>
> Reported-by: [email protected]
Acked-by: Stephen Hemminger <[email protected]>
On Wed, 28 Oct 2020 11:19:59 -0700 Stephen Hemminger wrote:
> On Wed, 28 Oct 2020 17:07:31 +0000
> Aleksandr Nogikh <[email protected]> wrote:
>
> > From: Aleksandr Nogikh <[email protected]>
> >
> > Currently it is possible to craft a special netlink RTM_NEWQDISC
> > command that can result in jitter being equal to 0x80000000. It is
> > enough to set the 32 bit jitter to 0x02000000 (it will later be
> > multiplied by 2^6) or just set the 64 bit jitter via
> > TCA_NETEM_JITTER64. This causes an overflow during the generation of
> > uniformly distributed numbers in tabledist(), which in turn leads to
> > division by zero (sigma != 0, but sigma * 2 is 0).
> >
> > The related fragment of code needs 32-bit division - see commit
> > 9b0ed89 ("netem: remove unnecessary 64 bit modulus"), so switching to
> > 64 bit is not an option.
> >
> > Fix the issue by keeping the value of jitter within the range that can
> > be adequately handled by tabledist() - [0;INT_MAX]. As negative std
> > deviation makes no sense, take the absolute value of the passed value
> > and cap it at INT_MAX. Inside tabledist(), switch to unsigned 32 bit
> > arithmetic in order to prevent overflows.
> >
> > Signed-off-by: Aleksandr Nogikh <[email protected]>
> > Reported-by: [email protected]
>
> Acked-by: Stephen Hemminger <[email protected]>
Applied, thanks!