net/sched: fq_pie: fix OOB access in the traffic path

net/sched: fq_pie: fix OOB access in the traffic path

the following script:

  # tc qdisc add dev eth0 handle 0x1 root fq_pie flows 2
  # tc qdisc add dev eth0 clsact
  # tc filter add dev eth0 egress matchall action skbedit priority 0x10002
  # ping 192.0.2.2 -I eth0 -c2 -w1 -q

produces the following splat:

 BUG: KASAN: slab-out-of-bounds in fq_pie_qdisc_enqueue+0x1314/0x19d0 [sch_fq_pie]
 Read of size 4 at addr ffff888171306924 by task ping/942

 CPU: 3 PID: 942 Comm: ping Not tainted 5.12.0+ #441
 Hardware name: Red Hat KVM, BIOS 1.11.1-4.module+el8.1.0+4066+0f1aadab 04/01/2014
 Call Trace:
  dump_stack+0x92/0xc1
  print_address_description.constprop.7+0x1a/0x150
  kasan_report.cold.13+0x7f/0x111
  fq_pie_qdisc_enqueue+0x1314/0x19d0 [sch_fq_pie]
  __dev_queue_xmit+0x1034/0x2b10
  ip_finish_output2+0xc62/0x2120
  __ip_finish_output+0x553/0xea0
  ip_output+0x1ca/0x4d0
  ip_send_skb+0x37/0xa0
  raw_sendmsg+0x1c4b/0x2d00
  sock_sendmsg+0xdb/0x110
  __sys_sendto+0x1d7/0x2b0
  __x64_sys_sendto+0xdd/0x1b0
  do_syscall_64+0x3c/0x80
  entry_SYSCALL_64_after_hwframe+0x44/0xae
 RIP: 0033:0x7fe69735c3eb
 Code: 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 f3 0f 1e fa 48 8d 05 75 42 2c 00 41 89 ca 8b 00 85 c0 75 14 b8 2c 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 75 c3 0f 1f 40 00 41 57 4d 89 c7 41 56 41 89
 RSP: 002b:00007fff06d7fb38 EFLAGS: 00000246 ORIG_RAX: 000000000000002c
 RAX: ffffffffffffffda RBX: 000055e961413700 RCX: 00007fe69735c3eb
 RDX: 0000000000000040 RSI: 000055e961413700 RDI: 0000000000000003
 RBP: 0000000000000040 R08: 000055e961410500 R09: 0000000000000010
 R10: 0000000000000000 R11: 0000000000000246 R12: 00007fff06d81260
 R13: 00007fff06d7fb40 R14: 00007fff06d7fc30 R15: 000055e96140f0a0

 Allocated by task 917:
  kasan_save_stack+0x19/0x40
  __kasan_kmalloc+0x7f/0xa0
  __kmalloc_node+0x139/0x280
  fq_pie_init+0x555/0x8e8 [sch_fq_pie]
  qdisc_create+0x407/0x11b0
  tc_modify_qdisc+0x3c2/0x17e0
  rtnetlink_rcv_msg+0x346/0x8e0
  netlink_rcv_skb+0x120/0x380
  netlink_unicast+0x439/0x630
  netlink_sendmsg+0x719/0xbf0
  sock_sendmsg+0xe2/0x110
  ____sys_sendmsg+0x5ba/0x890
  ___sys_sendmsg+0xe9/0x160
  __sys_sendmsg+0xd3/0x170
  do_syscall_64+0x3c/0x80
  entry_SYSCALL_64_after_hwframe+0x44/0xae

 The buggy address belongs to the object at ffff888171306800
  which belongs to the cache kmalloc-256 of size 256
 The buggy address is located 36 bytes to the right of
  256-byte region [ffff888171306800, ffff888171306900)
 The buggy address belongs to the page:
 page:00000000bcfb624e refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x171306
 head:00000000bcfb624e order:1 compound_mapcount:0
 flags: 0x17ffffc0010200(slab|head|node=0|zone=2|lastcpupid=0x1fffff)
 raw: 0017ffffc0010200 dead000000000100 dead000000000122 ffff888100042b40
 raw: 0000000000000000 0000000000100010 00000001ffffffff 0000000000000000
 page dumped because: kasan: bad access detected

 Memory state around the buggy address:
  ffff888171306800: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
  ffff888171306880: 00 00 00 00 00 00 00 00 00 00 00 00 fc fc fc fc
 >ffff888171306900: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
                                ^
  ffff888171306980: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
  ffff888171306a00: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb

fix fq_pie traffic path to avoid selecting 'q->flows + q->flows_cnt' as a
valid flow: it's an address beyond the allocated memory.

Fixes: ec97ecf1ebe4 ("net: sched: add Flow Queue PIE packet scheduler")
CC: stable@vger.kernel.org
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>