Skip to contentSkip to conversation

cls_matchall: avoid panic when receiving a packet before filter set

PATCH In review Matteo Croce 3 participants 5 messages May 2, 2019 – May 4, 2019

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 cls_matchall: avoid panic when receiving a packet before filter set

When a matchall classifier is added, there is a small time interval in which tp->root is NULL. If we receive a packet in this small time slice a NULL pointer dereference will happen, leading to a kernel panic:

    # tc qdisc replace dev eth0 ingress
    # tc filter add dev eth0 parent ffff: matchall action gact drop
    Unable to handle kernel NULL pointer dereference at virtual address 0000000000000034
    Mem abort info:
      ESR = 0x96000005
      Exception class = DABT (current EL), IL = 32 bits
      SET = 0, FnV = 0
      EA = 0, S1PTW = 0
    Data abort info:
      ISV = 0, ISS = 0x00000005
      CM = 0, WnR = 0
    user pgtable: 4k pages, 39-bit VAs, pgdp = 00000000a623d530
    [0000000000000034] pgd=0000000000000000, pud=0000000000000000
    Internal error: Oops: 96000005 [#1] SMP
    Modules linked in: cls_matchall sch_ingress nls_iso8859_1 nls_cp437 vfat fat m25p80 spi_nor mtd xhci_plat_hcd xhci_hcd phy_generic sfp mdio_i2c usbcore i2c_mv64xxx marvell10g mvpp2 usb_common spi_orion mvmdio i2c_core sbsa_gwdt phylink ip_tables x_tables autofs4
    Process ksoftirqd/0 (pid: 9, stack limit = 0x0000000009de7d62)
    CPU: 0 PID: 9 Comm: ksoftirqd/0 Not tainted 5.1.0-rc6 #21
    Hardware name: Marvell 8040 MACCHIATOBin Double-shot (DT)
    pstate: 40000005 (nZcv daif -PAN -UAO)
    pc : mall_classify+0x28/0x78 [cls_matchall]
    lr : tcf_classify+0x78/0x138
    sp : ffffff80109db9d0
    x29: ffffff80109db9d0 x28: ffffffc426058800
    x27: 0000000000000000 x26: ffffffc425b0dd00
    x25: 0000000020000000 x24: 0000000000000000
    x23: ffffff80109dbac0 x22: 0000000000000001
    x21: ffffffc428ab5100 x20: ffffffc425b0dd00
    x19: ffffff80109dbac0 x18: 0000000000000000
    x17: 0000000000000000 x16: 0000000000000000
    x15: 0000000000000000 x14: 0000000000000000
    x13: ffffffbf108ad288 x12: dead000000000200
    x11: 00000000f0000000 x10: 0000000000000001
    x9 : ffffffbf1089a220 x8 : 0000000000000001
    x7 : ffffffbebffaa950 x6 : 0000000000000000
    x5 : 000000442d6ba000 x4 : 0000000000000000
    x3 : ffffff8008735ad8 x2 : ffffff80109dbac0
    x1 : ffffffc425b0dd00 x0 : ffffff8010592078
    Call trace:
     mall_classify+0x28/0x78 [cls_matchall]
     tcf_classify+0x78/0x138
     __netif_receive_skb_core+0x29c/0xa20
     __netif_receive_skb_one_core+0x34/0x60
     __netif_receive_skb+0x28/0x78
     netif_receive_skb_internal+0x2c/0xc0
     napi_gro_receive+0x1a0/0x1d8
     mvpp2_poll+0x928/0xb18 [mvpp2]
     net_rx_action+0x108/0x378
     __do_softirq+0x128/0x320
     run_ksoftirqd+0x44/0x60
     smpboot_thread_fn+0x168/0x1b0
     kthread+0x12c/0x130
     ret_from_fork+0x10/0x1c
    Code: aa0203f3 aa1e03e0 d503201f f9400684 (b9403480)
    ---[ end trace fc71e2ef7b8ab5a5 ]---
    Kernel panic - not syncing: Fatal exception in interrupt
    SMP: stopping secondary CPUs
    Kernel Offset: disabled
    CPU features: 0x002,00002000
    Memory Limit: none
    Rebooting in 1 seconds..

Fix this by adding a NULL check in mall_classify().

Fixes ed76f5edccc9 ("net: sched: protect filter_chain list with filter_chain_lock mutex")
---
 net/sched/cls_matchall.c | 3 +++
 1 file changed, 3 insertions(+)
net/sched/cls_matchall.c +3 -0
index a13bc351a414..3d021f2aad1c 100644
@@ -32,6 +32,9 @@ static int mall_classify(struct sk_buff *skb, const struct tcf_proto *tp,
 {
 	struct cls_mall_head *head = rcu_dereference_bh(tp->root);
 
+	if (unlikely(!head))
+		return -1;
+
 	if (tc_skip_sw(head->flags))
 		return -1;
 
Signature · 1 line
2.21.0
#1 lore raw
Quoted 5 lines · Matteo Croce

When a matchall classifier is added, there is a small time interval in which tp->root is NULL. If we receive a packet in this small time slice a NULL pointer dereference will happen, leading to a kernel panic:

Hi,

I forgot to mark it as v2. Will someone handle it, or I have to resubmit a v2 or v3?

Regards,

Signature · 2 lines
Matteo Croce
per aspera ad upstream
#2 lore raw
Quoted 8 lines · Matteo Croce
Quoted 5 lines · Matteo Croce

When a matchall classifier is added, there is a small time interval in which tp->root is NULL. If we receive a packet in this small time slice a NULL pointer dereference will happen, leading to a kernel panic:

Quoted 4 lines

Hi,

I forgot to mark it as v2. Will someone handle it, or I have to resubmit a v2 or v3?

As long as David doesn't get confused on which one he should apply, it is okay.

Thanks.

Acked-by on v1 1/1
#3 lore raw
In reply to Matteo Croce
Quoted 4 lines · Matteo Croce

When a matchall classifier is added, there is a small time interval in which tp->root is NULL. If we receive a packet in this small time slice a NULL pointer dereference will happen, leading to a kernel panic:

...

Quoted 4 lines

Fix this by adding a NULL check in mall_classify().

Fixes ed76f5edccc9 ("net: sched: protect filter_chain list with filter_chain_lock mutex")

Thanks for the update!

maintainer
#4 lore raw
In reply to Matteo Croce

From: Matteo Croce <[email protected]>
Date: Thu, 2 May 2019 10:51:05 +0200

Quoted 3 lines

When a matchall classifier is added, there is a small time interval in which tp->root is NULL. If we receive a packet in this small time slice a NULL pointer dereference will happen, leading to a kernel panic:

...

Quoted 4 lines

Fix this by adding a NULL check in mall_classify().

Fixes ed76f5edccc9 ("net: sched: protect filter_chain list with filter_chain_lock mutex")

Applied and queued up for -stable, thanks.