Em Thu, 28 Dec 2017 14:19:56 +0200
Sakari Ailus <sakari.ailus@xxxxxxxxxxxxxxx> escreveu:
> Hi Mauro,
>
> Thanks for the patch.
>
> On Thu, Dec 28, 2017 at 10:02:33AM -0200, Mauro Carvalho Chehab wrote:
> > Currently, there's no check if an invalid buffer range
> > is passed. However, while testing DVB memory mapped apps,
> > I got this:
> >
> > videobuf2_core: VB: num_buffers -2143943680, buffer 33, index -2143943647
> > unable to handle kernel paging request at ffff888b773c0890
> > IP: __vb2_queue_alloc+0x134/0x4e0 [videobuf2_core]
> > PGD 4142c7067 P4D 4142c7067 PUD 0
> > Oops: 0002 [#1] SMP
> > Modules linked in: xt_CHECKSUM iptable_mangle ipt_MASQUERADE nf_nat_masquerade_ipv4 iptable_nat nf_nat_ipv4 nf_nat nf_conntrack_ipv4 nf_defrag_ipv4 xt_conntrack nf_conntrack tun bridge stp llc ebtable_filter ebtables ip6table_filter ip6_tables bluetooth rfkill ecdh_generic binfmt_misc rc_dvbsky sp2 ts2020 intel_rapl x86_pkg_temp_thermal dvb_usb_dvbsky intel_powerclamp dvb_usb_v2 coretemp m88ds3103 kvm_intel i2c_mux dvb_core snd_hda_codec_hdmi crct10dif_pclmul crc32_pclmul videobuf2_vmalloc videobuf2_memops snd_hda_intel ghash_clmulni_intel videobuf2_core snd_hda_codec rc_core mei_me intel_cstate snd_hwdep snd_hda_core videodev intel_uncore snd_pcm mei media tpm_tis tpm_tis_core intel_rapl_perf tpm snd_timer lpc_ich snd soundcore kvm irqbypass libcrc32c i915 i2c_algo_bit drm_kms_helper
> > e1000e ptp drm crc32c_intel video pps_core
> > CPU: 3 PID: 1776 Comm: dvbv5-zap Not tainted 4.14.0+ #78
> > Hardware name: /NUC5i7RYB, BIOS RYBDWi35.86A.0364.2017.0511.0949 05/11/2017
> > task: ffff88877c73bc80 task.stack: ffffb7c402418000
> > RIP: 0010:__vb2_queue_alloc+0x134/0x4e0 [videobuf2_core]
> > RSP: 0018:ffffb7c40241bc60 EFLAGS: 00010246
> > RAX: 0000000080360421 RBX: 0000000000000021 RCX: 000000000000000a
> > RDX: ffffb7c40241bcf4 RSI: ffff888780362c60 RDI: ffff888796d8e130
> > RBP: ffffb7c40241bcc8 R08: 0000000000000316 R09: 0000000000000004
> > R10: ffff888780362c00 R11: 0000000000000001 R12: 000000000002f000
> > R13: ffff8887758be700 R14: 0000000000021000 R15: 0000000000000001
> > FS: 00007f2849024740(0000) GS:ffff888796d80000(0000) knlGS:0000000000000000
> > CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > CR2: ffff888b773c0890 CR3: 000000043beb2005 CR4: 00000000003606e0
> > Call Trace:
> > vb2_core_reqbufs+0x226/0x420 [videobuf2_core]
> > dvb_vb2_reqbufs+0x2d/0xc0 [dvb_core]
> > dvb_dvr_do_ioctl+0x98/0x1d0 [dvb_core]
> > dvb_usercopy+0x53/0x1b0 [dvb_core]
> > ? dvb_demux_ioctl+0x20/0x20 [dvb_core]
> > ? tty_ldisc_deref+0x16/0x20
> > ? tty_write+0x1f9/0x310
> > ? process_echoes+0x70/0x70
> > dvb_dvr_ioctl+0x15/0x20 [dvb_core]
> > do_vfs_ioctl+0xa5/0x600
> > SyS_ioctl+0x79/0x90
> > entry_SYSCALL_64_fastpath+0x1a/0xa5
> > RIP: 0033:0x7f28486f7ea7
> > RSP: 002b:00007ffc13b2db18 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
> > RAX: ffffffffffffffda RBX: 000055b10fc06130 RCX: 00007f28486f7ea7
> > RDX: 00007ffc13b2db48 RSI: 00000000c0086f3c RDI: 0000000000000007
> > RBP: 0000000000000203 R08: 000055b10df1e02c R09: 000000000000002e
> > R10: 0036b42415108357 R11: 0000000000000246 R12: 0000000000000000
> > R13: 00007f2849062f60 R14: 00000000000001f1 R15: 00007ffc13b2da54
> > Code: 74 0a 60 8b 0a 48 83 c0 30 48 83 c2 04 89 48 d0 89 48 d4 48 39 f0 75 eb 41 8b 42 08 83 7d d4 01 41 c7 82 ec 01 00 00 ff ff ff ff <4d> 89 94 c5 88 00 00 00 74 14 83 c3 01 41 39 dc 0f 85 f1 fe ff
> > RIP: __vb2_queue_alloc+0x134/0x4e0 [videobuf2_core] RSP: ffffb7c40241bc60
> > CR2: ffff888b773c0890
> >
> > So, add a sanity check in order to prevent going past array.
> >
> > Signed-off-by: Mauro Carvalho Chehab <mchehab@xxxxxxxxxxxxxxxx>
> > ---
> > drivers/media/common/videobuf/videobuf2-core.c | 4 ++++
> > 1 file changed, 4 insertions(+)
> >
> > diff --git a/drivers/media/common/videobuf/videobuf2-core.c b/drivers/media/common/videobuf/videobuf2-core.c
> > index cb115ba6a1d2..9107ffc4d808 100644
> > --- a/drivers/media/common/videobuf/videobuf2-core.c
> > +++ b/drivers/media/common/videobuf/videobuf2-core.c
> > @@ -332,6 +332,10 @@ static int __vb2_queue_alloc(struct vb2_queue *q, enum vb2_memory memory,
> > struct vb2_buffer *vb;
> > int ret;
> >
> > + /* Sanity check to avoid going past q->bufs size */
> > + if (q->num_buffers < 0 || q->num_buffers + num_buffers > VB2_MAX_FRAME)
>
> These checks already exists in vb2_core_reqbufs as well as in
> vb2_core_create_bufs. The queue_setup callback can change it though, so the
> check appears to be worth it.
>
> I'm just trying to figure out whether it's only DVB that's affected.
I suspect so. At dvb-vb2, it just calls the VB2 core, without validating
the buffers.
Without this patch, it is accepting q->num_buffers + num_buffers to be
greater than VB2_MAX_FRAME.
In any case, as this causes a buffer overflow, it is worth having
the check there, where q->bufs[q->num_buffers + index] is
used for the first time.
I still think I'll add an additional check at dvb-vb2, but, as we
could have other clients in the future, it seems worth that the core
itself to double-check.
>
> q->num_buffers is unsigned, so no need to check for < 0.
Ah, true. Will resend without this extra check.
>
> > + return 0;
> > +
> > for (buffer = 0; buffer < num_buffers; ++buffer) {
> > /* Allocate videobuf buffer structures */
> > vb = kzalloc(q->buf_struct_size, GFP_KERNEL);
> > --
> > 2.14.3
> >
>
Thanks,
Mauro
