[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <4e8324cf-e6de-acff-5e30-373d015a3cb4@mojatatu.com>
Date: Thu, 20 Apr 2023 13:25:33 -0300
From: Pedro Tammela <pctammela@...atatu.com>
To: Davide Caratti <dcaratti@...hat.com>,
Eric Dumazet <edumazet@...gle.com>,
Jamal Hadi Salim <jhs@...atatu.com>,
Cong Wang <xiyou.wangcong@...il.com>,
Jiri Pirko <jiri@...nulli.us>, Jakub Kicinski <kuba@...nel.org>
Cc: netdev@...r.kernel.org
Subject: Re: [PATCH net v2] net/sched: sch_fq: fix integer overflow of
"credit"
On 20/04/2023 11:59, Davide Caratti wrote:
> if sch_fq is configured with "initial quantum" having values greater than
> INT_MAX, the first assignment of "credit" does signed integer overflow to
> a very negative value.
> In this situation, the syzkaller script provided by Cristoph triggers the
> CPU soft-lockup warning even with few sockets. It's not an infinite loop,
> but "credit" wasn't probably meant to be minus 2Gb for each new flow.
> Capping "initial quantum" to INT_MAX proved to fix the issue.
>
> v2: validation of "initial quantum" is done in fq_policy, instead of open
> coding in fq_change() _ suggested by Jakub Kicinski
>
> Reported-by: Christoph Paasch <cpaasch@...le.com>
> Link: https://github.com/multipath-tcp/mptcp_net-next/issues/377
> Fixes: afe4fd062416 ("pkt_sched: fq: Fair Queue packet scheduler")
> Reviewed-by: Eric Dumazet <edumazet@...gle.com>
> Signed-off-by: Davide Caratti <dcaratti@...hat.com>
> ---
> net/sched/sch_fq.c | 6 ++++-
> .../tc-testing/tc-tests/qdiscs/fq.json | 22 +++++++++++++++++++
> 2 files changed, 27 insertions(+), 1 deletion(-)
>
> diff --git a/net/sched/sch_fq.c b/net/sched/sch_fq.c
> index 48d14fb90ba0..f59a2cb2c803 100644
> --- a/net/sched/sch_fq.c
> +++ b/net/sched/sch_fq.c
> @@ -779,13 +779,17 @@ static int fq_resize(struct Qdisc *sch, u32 log)
> return 0;
> }
>
> +static struct netlink_range_validation iq_range = {
> + .max = INT_MAX,
> +};
> +
> static const struct nla_policy fq_policy[TCA_FQ_MAX + 1] = {
> [TCA_FQ_UNSPEC] = { .strict_start_type = TCA_FQ_TIMER_SLACK },
>
> [TCA_FQ_PLIMIT] = { .type = NLA_U32 },
> [TCA_FQ_FLOW_PLIMIT] = { .type = NLA_U32 },
> [TCA_FQ_QUANTUM] = { .type = NLA_U32 },
> - [TCA_FQ_INITIAL_QUANTUM] = { .type = NLA_U32 },
> + [TCA_FQ_INITIAL_QUANTUM] = NLA_POLICY_FULL_RANGE(NLA_U32, &iq_range),
> [TCA_FQ_RATE_ENABLE] = { .type = NLA_U32 },
> [TCA_FQ_FLOW_DEFAULT_RATE] = { .type = NLA_U32 },
> [TCA_FQ_FLOW_MAX_RATE] = { .type = NLA_U32 },
> diff --git a/tools/testing/selftests/tc-testing/tc-tests/qdiscs/fq.json b/tools/testing/selftests/tc-testing/tc-tests/qdiscs/fq.json
> index 8acb904d1419..3593fb8f79ad 100644
> --- a/tools/testing/selftests/tc-testing/tc-tests/qdiscs/fq.json
> +++ b/tools/testing/selftests/tc-testing/tc-tests/qdiscs/fq.json
> @@ -114,6 +114,28 @@
> "$IP link del dev $DUMMY type dummy"
> ]
> },
> + {
> + "id": "10f7",
> + "name": "Create FQ with invalid initial_quantum setting",
> + "category": [
> + "qdisc",
> + "fq"
> + ],
> + "plugins": {
> + "requires": "nsPlugin"
> + },
> + "setup": [
> + "$IP link add dev $DUMMY type dummy || /bin/true"
> + ],
> + "cmdUnderTest": "$TC qdisc add dev $DUMMY handle 1: root fq initial_quantum 0x80000000",
> + "expExitCode": "2",
> + "verifyCmd": "$TC qdisc show dev $DUMMY",
> + "matchPattern": "qdisc fq 1: root.*initial_quantum 2048Mb",
> + "matchCount": "0",
> + "teardown": [
> + "$IP link del dev $DUMMY type dummy"
> + ]
> + },
> {
> "id": "9398",
> "name": "Create FQ with maxrate setting",
You probably don't want to backport the test as well? If so I would
break this patch into two.
Powered by blists - more mailing lists