tty: hvc: fix buffer overflow during hvc_alloc().
authorAndrew Melnychenko <andrew@daynix.com>
Tue, 14 Apr 2020 19:15:03 +0000 (22:15 +0300)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 23 Apr 2020 14:02:07 +0000 (16:02 +0200)
commit9a9fc42b86c06120744555fea43fdcabe297c656
tree3b280e34181592e254e7d8355bb8232faedf46ad
parent7127d24372bf23675a36edc64d092dc7fd92ebe8
tty: hvc: fix buffer overflow during hvc_alloc().

If there is a lot(more then 16) of virtio-console devices
or virtio_console module is reloaded
- buffers 'vtermnos' and 'cons_ops' are overflowed.
In older kernels it overruns spinlock which leads to kernel freezing:
https://bugzilla.redhat.com/show_bug.cgi?id=1786239

To reproduce the issue, you can try simple script that
loads/unloads module. Something like this:
while [ 1 ]
do
  modprobe virtio_console
  sleep 2
  modprobe -r virtio_console
  sleep 2
done

Description of problem:
Guest get 'Call Trace' when loading module "virtio_console"
and unloading it frequently - clearly reproduced on kernel-4.18.0:

[   81.498208] ------------[ cut here ]------------
[   81.499263] pvqspinlock: lock 0xffffffff92080020 has corrupted value 0xc0774ca0!
[   81.501000] WARNING: CPU: 0 PID: 785 at kernel/locking/qspinlock_paravirt.h:500 __pv_queued_spin_unlock_slowpath+0xc0/0xd0
[   81.503173] Modules linked in: virtio_console fuse xt_CHECKSUM ipt_MASQUERADE xt_conntrack ipt_REJECT nft_counter nf_nat_tftp nft_objref nf_conntrack_tftp tun bridge stp llc nft_fib_inet nft_fib_ipv4 nft_fib_ipv6 nft_fib nft_reject_inet nf_reject_ipv4 nf_reject_ipv6 nft_reject nft_ct nf_tables_set nft_chain_nat_ipv6 nf_conntrack_ipv6 nf_defrag_ipv6 nf_nat_ipv6 nft_chain_route_ipv6 nft_chain_nat_ipv4 nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack nft_chain_route_ipv4 ip6_tables nft_compat ip_set nf_tables nfnetlink sunrpc bochs_drm drm_vram_helper ttm drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops drm i2c_piix4 pcspkr crct10dif_pclmul crc32_pclmul joydev ghash_clmulni_intel ip_tables xfs libcrc32c sd_mod sg ata_generic ata_piix virtio_net libata crc32c_intel net_failover failover serio_raw virtio_scsi dm_mirror dm_region_hash dm_log dm_mod [last unloaded: virtio_console]
[   81.517019] CPU: 0 PID: 785 Comm: kworker/0:2 Kdump: loaded Not tainted 4.18.0-167.el8.x86_64 #1
[   81.518639] Hardware name: Red Hat KVM, BIOS 1.12.0-5.scrmod+el8.2.0+5159+d8aa4d83 04/01/2014
[   81.520205] Workqueue: events control_work_handler [virtio_console]
[   81.521354] RIP: 0010:__pv_queued_spin_unlock_slowpath+0xc0/0xd0
[   81.522450] Code: 07 00 48 63 7a 10 e8 bf 64 f5 ff 66 90 c3 8b 05 e6 cf d6 01 85 c0 74 01 c3 8b 17 48 89 fe 48 c7 c7 38 4b 29 91 e8 3a 6c fa ff <0f> 0b c3 0f 0b 90 90 90 90 90 90 90 90 90 90 90 0f 1f 44 00 00 48
[   81.525830] RSP: 0018:ffffb51a01ffbd70 EFLAGS: 00010282
[   81.526798] RAX: 0000000000000000 RBX: 0000000000000010 RCX: 0000000000000000
[   81.528110] RDX: ffff9e66f1826480 RSI: ffff9e66f1816a08 RDI: ffff9e66f1816a08
[   81.529437] RBP: ffffffff9153ff10 R08: 000000000000026c R09: 0000000000000053
[   81.530732] R10: 0000000000000000 R11: ffffb51a01ffbc18 R12: ffff9e66cd682200
[   81.532133] R13: ffffffff9153ff10 R14: ffff9e6685569500 R15: ffff9e66cd682000
[   81.533442] FS:  0000000000000000(0000) GS:ffff9e66f1800000(0000) knlGS:0000000000000000
[   81.534914] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   81.535971] CR2: 00005624c55b14d0 CR3: 00000003a023c000 CR4: 00000000003406f0
[   81.537283] Call Trace:
[   81.537763]  __raw_callee_save___pv_queued_spin_unlock_slowpath+0x11/0x20
[   81.539011]  .slowpath+0x9/0xe
[   81.539585]  hvc_alloc+0x25e/0x300
[   81.540237]  init_port_console+0x28/0x100 [virtio_console]
[   81.541251]  handle_control_message.constprop.27+0x1c4/0x310 [virtio_console]
[   81.542546]  control_work_handler+0x70/0x10c [virtio_console]
[   81.543601]  process_one_work+0x1a7/0x3b0
[   81.544356]  worker_thread+0x30/0x390
[   81.545025]  ? create_worker+0x1a0/0x1a0
[   81.545749]  kthread+0x112/0x130
[   81.546358]  ? kthread_flush_work_fn+0x10/0x10
[   81.547183]  ret_from_fork+0x22/0x40
[   81.547842] ---[ end trace aa97649bd16c8655 ]---
[   83.546539] general protection fault: 0000 [#1] SMP NOPTI
[   83.547422] CPU: 5 PID: 3225 Comm: modprobe Kdump: loaded Tainted: G        W        --------- -  - 4.18.0-167.el8.x86_64 #1
[   83.549191] Hardware name: Red Hat KVM, BIOS 1.12.0-5.scrmod+el8.2.0+5159+d8aa4d83 04/01/2014
[   83.550544] RIP: 0010:__pv_queued_spin_lock_slowpath+0x19a/0x2a0
[   83.551504] Code: c4 c1 ea 12 41 be 01 00 00 00 4c 8d 6d 14 41 83 e4 03 8d 42 ff 49 c1 e4 05 48 98 49 81 c4 40 a5 02 00 4c 03 24 c5 60 48 34 91 <49> 89 2c 24 b8 00 80 00 00 eb 15 84 c0 75 0a 41 0f b6 54 24 14 84
[   83.554449] RSP: 0018:ffffb51a0323fdb0 EFLAGS: 00010202
[   83.555290] RAX: 000000000000301c RBX: ffffffff92080020 RCX: 0000000000000001
[   83.556426] RDX: 000000000000301d RSI: 0000000000000000 RDI: 0000000000000000
[   83.557556] RBP: ffff9e66f196a540 R08: 000000000000028a R09: ffff9e66d2757788
[   83.558688] R10: 0000000000000000 R11: 0000000000000000 R12: 646e61725f770b07
[   83.559821] R13: ffff9e66f196a554 R14: 0000000000000001 R15: 0000000000180000
[   83.560958] FS:  00007fd5032e8740(0000) GS:ffff9e66f1940000(0000) knlGS:0000000000000000
[   83.562233] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   83.563149] CR2: 00007fd5022b0da0 CR3: 000000038c334000 CR4: 00000000003406e0

Signed-off-by: Andrew Melnychenko <andrew@daynix.com>
Cc: stable <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20200414191503.3471783-1-andrew@daynix.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/tty/hvc/hvc_console.c