In the Linux kernel, the following vulnerability has been resolved:
net/sched: taprio: avoid disabling offload when it was never enabled In an
incredibly strange API design decision, qdisc->destroy() gets called even
if qdisc->init() never succeeded, not exclusively since commit 87b60cfacf9f
(“net_sched: fix error recovery at qdisc creation”), but apparently also
earlier (in the case of qdisc_create_dflt()). The taprio qdisc does not
fully acknowledge this when it attempts full offload, because it starts off
with q->flags = TAPRIO_FLAGS_INVALID in taprio_init(), then it replaces
q->flags with TCA_TAPRIO_ATTR_FLAGS parsed from netlink (in
taprio_change(), tail called from taprio_init()). But in taprio_destroy(),
we call taprio_disable_offload(), and this determines what to do based on
FULL_OFFLOAD_IS_ENABLED(q->flags). But looking at the implementation of
FULL_OFFLOAD_IS_ENABLED() (a bitwise check of bit 1 in q->flags), it is
invalid to call this macro on q->flags when it contains
TAPRIO_FLAGS_INVALID, because that is set to U32_MAX, and therefore
FULL_OFFLOAD_IS_ENABLED() will return true on an invalid set of flags. As a
result, it is possible to crash the kernel if user space forces an error
between setting q->flags = TAPRIO_FLAGS_INVALID, and the calling of
taprio_enable_offload(). This is because drivers do not expect the offload
to be disabled when it was never enabled. The error that we force here is
to attach taprio as a non-root qdisc, but instead as child of an mqprio
root qdisc: $ tc qdisc add dev swp0 root handle 1: \ mqprio num_tc 8 map 0
1 2 3 4 5 6 7 \ queues 1@0 1@1 1@2 1@3 1@4 1@5 1@6 1@7 hw 0 $ tc qdisc
replace dev swp0 parent 1:1 \ taprio num_tc 8 map 0 1 2 3 4 5 6 7 \ queues
1@0 1@1 1@2 1@3 1@4 1@5 1@6 1@7 base-time 0 \ sched-entry S 0x7f 990000
sched-entry S 0x80 100000 \ flags 0x0 clockid CLOCK_TAI Unable to handle
kernel paging request at virtual address fffffffffffffff8
[fffffffffffffff8] pgd=0000000000000000, p4d=0000000000000000 Internal
error: Oops: 96000004 [#1] PREEMPT SMP Call trace: taprio_dump+0x27c/0x310
vsc9959_port_setup_tc+0x1f4/0x460 felix_port_setup_tc+0x24/0x3c
dsa_slave_setup_tc+0x54/0x27c taprio_disable_offload.isra.0+0x58/0xe0
taprio_destroy+0x80/0x104 qdisc_create+0x240/0x470
tc_modify_qdisc+0x1fc/0x6b0 rtnetlink_rcv_msg+0x12c/0x390
netlink_rcv_skb+0x5c/0x130 rtnetlink_rcv+0x1c/0x2c Fix this by keeping
track of the operations we made, and undo the offload only if we actually
did it. I’ve added “bool offloaded” inside a 4 byte hole between “int
clockid” and “atomic64_t picos_per_byte”. Now the first cache line looks
like below: $ pahole -C taprio_sched net/sched/sch_taprio.o struct
taprio_sched { struct Qdisc * * qdiscs; /* 0 8 / struct Qdisc * root; / 8
8 / u32 flags; / 16 4 / enum tk_offsets tk_offset; / 20 4 / int
clockid; / 24 4 / bool offloaded; / 28 1 / / XXX 3 bytes hole, try to
pack / atomic64_t picos_per_byte; / 32 0 / / XXX 8 bytes hole, try to
pack / spinlock_t current_entry_lock; / 40 0 / / XXX 8 bytes hole, try
to pack / struct sched_entry * current_entry; / 48 8 / struct
sched_gate_list * oper_sched; / 56 8 / / — cacheline 1 boundary (64
bytes) — */
OS | Version | Architecture | Package | Version | Filename |
---|---|---|---|---|---|
ubuntu | 20.04 | noarch | linux | < any | UNKNOWN |
ubuntu | 22.04 | noarch | linux | < any | UNKNOWN |
ubuntu | 20.04 | noarch | linux-aws | < any | UNKNOWN |
ubuntu | 22.04 | noarch | linux-aws | < any | UNKNOWN |
ubuntu | 20.04 | noarch | linux-aws-5.15 | < any | UNKNOWN |
ubuntu | 18.04 | noarch | linux-aws-5.4 | < any | UNKNOWN |
ubuntu | 20.04 | noarch | linux-azure | < any | UNKNOWN |
ubuntu | 22.04 | noarch | linux-azure | < any | UNKNOWN |
ubuntu | 20.04 | noarch | linux-azure-5.15 | < any | UNKNOWN |
ubuntu | 18.04 | noarch | linux-azure-5.4 | < any | UNKNOWN |
git.kernel.org/linus/db46e3a88a09c5cf7e505664d01da7238cd56c92 (6.0-rc7)
git.kernel.org/stable/c/586def6ebed195f3594a4884f7c5334d0e1ad1bb
git.kernel.org/stable/c/c7c9c7eb305ab8b4e93e4e4e1b78d8cfcbc26323
git.kernel.org/stable/c/d12a1eb07003e597077329767c6aa86a7e972c76
git.kernel.org/stable/c/db46e3a88a09c5cf7e505664d01da7238cd56c92
git.kernel.org/stable/c/f58e43184226e5e9662088ccf1389e424a3a4cbd
launchpad.net/bugs/cve/CVE-2022-48644
nvd.nist.gov/vuln/detail/CVE-2022-48644
security-tracker.debian.org/tracker/CVE-2022-48644
www.cve.org/CVERecord?id=CVE-2022-48644