[PATCH 5.13 052/113] net/sched: ets: fix crash when flipping from strict to quantum

From: Greg Kroah-Hartman
Date: Wed Sep 01 2021 - 08:50:02 EST


From: Davide Caratti <dcaratti@xxxxxxxxxx>

[ Upstream commit cd9b50adc6bb9ad3f7d244590a389522215865c4 ]

While running kselftests, Hangbin observed that sch_ets.sh often crashes,
and splats like the following one are seen in the output of 'dmesg':

BUG: kernel NULL pointer dereference, address: 0000000000000000
#PF: supervisor read access in kernel mode
#PF: error_code(0x0000) - not-present page
PGD 159f12067 P4D 159f12067 PUD 159f13067 PMD 0
Oops: 0000 [#1] SMP NOPTI
CPU: 2 PID: 921 Comm: tc Not tainted 5.14.0-rc6+ #458
Hardware name: Red Hat KVM, BIOS 1.11.1-4.module+el8.1.0+4066+0f1aadab 04/01/2014
RIP: 0010:__list_del_entry_valid+0x2d/0x50
Code: 48 8b 57 08 48 b9 00 01 00 00 00 00 ad de 48 39 c8 0f 84 ac 6e 5b 00 48 b9 22 01 00 00 00 00 ad de 48 39 ca 0f 84 cf 6e 5b 00 <48> 8b 32 48 39 fe 0f 85 af 6e 5b 00 48 8b 50 08 48 39 f2 0f 85 94
RSP: 0018:ffffb2da005c3890 EFLAGS: 00010217
RAX: 0000000000000000 RBX: ffff9073ba23f800 RCX: dead000000000122
RDX: 0000000000000000 RSI: 0000000000000008 RDI: ffff9073ba23fbc8
RBP: ffff9073ba23f890 R08: 0000000000000001 R09: 0000000000000001
R10: 0000000000000001 R11: 0000000000000001 R12: dead000000000100
R13: ffff9073ba23fb00 R14: 0000000000000002 R15: 0000000000000002
FS: 00007f93e5564e40(0000) GS:ffff9073bba00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000000 CR3: 000000014ad34000 CR4: 0000000000350ee0
Call Trace:
ets_qdisc_reset+0x6e/0x100 [sch_ets]
qdisc_reset+0x49/0x1d0
tbf_reset+0x15/0x60 [sch_tbf]
qdisc_reset+0x49/0x1d0
dev_reset_queue.constprop.42+0x2f/0x90
dev_deactivate_many+0x1d3/0x3d0
dev_deactivate+0x56/0x90
qdisc_graft+0x47e/0x5a0
tc_get_qdisc+0x1db/0x3e0
rtnetlink_rcv_msg+0x164/0x4c0
netlink_rcv_skb+0x50/0x100
netlink_unicast+0x1a5/0x280
netlink_sendmsg+0x242/0x480
sock_sendmsg+0x5b/0x60
____sys_sendmsg+0x1f2/0x260
___sys_sendmsg+0x7c/0xc0
__sys_sendmsg+0x57/0xa0
do_syscall_64+0x3a/0x80
entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x7f93e44b8338
Code: 89 02 48 c7 c0 ff ff ff ff eb b5 0f 1f 80 00 00 00 00 f3 0f 1e fa 48 8d 05 25 43 2c 00 8b 00 85 c0 75 17 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 58 c3 0f 1f 80 00 00 00 00 41 54 41 89 d4 55
RSP: 002b:00007ffc0db737a8 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
RAX: ffffffffffffffda RBX: 0000000061255c06 RCX: 00007f93e44b8338
RDX: 0000000000000000 RSI: 00007ffc0db73810 RDI: 0000000000000003
RBP: 0000000000000000 R08: 0000000000000001 R09: 0000000000000000
R10: 000000000000000b R11: 0000000000000246 R12: 0000000000000001
R13: 0000000000687880 R14: 0000000000000000 R15: 0000000000000000
Modules linked in: sch_ets sch_tbf dummy rfkill iTCO_wdt iTCO_vendor_support intel_rapl_msr intel_rapl_common joydev i2c_i801 pcspkr i2c_smbus lpc_ich virtio_balloon ip_tables xfs libcrc32c crct10dif_pclmul crc32_pclmul crc32c_intel ahci libahci ghash_clmulni_intel libata serio_raw virtio_blk virtio_console virtio_net net_failover failover sunrpc dm_mirror dm_region_hash dm_log dm_mod
CR2: 0000000000000000

When the change() function decreases the value of 'nstrict', we must take
into account that packets might be already enqueued on a class that flips
from 'strict' to 'quantum': otherwise that class will not be added to the
bandwidth-sharing list. Then, a call to ets_qdisc_reset() will attempt to
do list_del(&alist) with 'alist' filled with zero, hence the NULL pointer
dereference.
For classes flipping from 'strict' to 'quantum', initialize an empty list
and eventually add it to the bandwidth-sharing list, if there are packets
already enqueued. In this way, the kernel will:
a) prevent crashing as described above.
b) avoid retaining the backlog packets (for an arbitrarily long time) in
case no packet is enqueued after a change from 'strict' to 'quantum'.

Reported-by: Hangbin Liu <liuhangbin@xxxxxxxxx>
Fixes: dcc68b4d8084 ("net: sch_ets: Add a new Qdisc")
Signed-off-by: Davide Caratti <dcaratti@xxxxxxxxxx>
Signed-off-by: David S. Miller <davem@xxxxxxxxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
---
net/sched/sch_ets.c | 7 +++++++
1 file changed, 7 insertions(+)

diff --git a/net/sched/sch_ets.c b/net/sched/sch_ets.c
index c1e84d1eeaba..c76701ac35ab 100644
--- a/net/sched/sch_ets.c
+++ b/net/sched/sch_ets.c
@@ -660,6 +660,13 @@ static int ets_qdisc_change(struct Qdisc *sch, struct nlattr *opt,
sch_tree_lock(sch);

q->nbands = nbands;
+ for (i = nstrict; i < q->nstrict; i++) {
+ INIT_LIST_HEAD(&q->classes[i].alist);
+ if (q->classes[i].qdisc->q.qlen) {
+ list_add_tail(&q->classes[i].alist, &q->active);
+ q->classes[i].deficit = quanta[i];
+ }
+ }
q->nstrict = nstrict;
memcpy(q->prio2band, priomap, sizeof(priomap));

--
2.30.2