Re: [PATCH RESEND v3 0/5] media: uvcvideo: Fix race conditions

From: Laurent Pinchart
Date: Thu Sep 17 2020 - 08:51:48 EST


Hi Guenter,

On Wed, Sep 16, 2020 at 07:25:42PM -0700, Guenter Roeck wrote:
> Something seems to have gone wrong with v3 of this patch series.
> I am sure I sent it out, but I don't find it anywhere.
> Resending. Sorry for any duplicates.

I haven't checked the mailing list, but I've found it in my inbox :-)
I'm not forgetting about you, just been fairly busy recently. I still
plan to try and provide an alternative implementation in the V4L2 core
(in a form that I think should even be moved to the cdev core) that
would fix this for all drivers.

By the way, as you managed to get hold of non-UVC webcams, one thing you
could try in your tests to make the drivers misbehave is to block on a
DQBUF call, and unplug the device at that time. When blocking, DQBUF
releases the driver lock (through the vb2ops .wait_prepare() and
.wait_finis() operations for drivers based on vb2), so this may allow
unregistration to proceed without waiting for userspace calls to
complete.

> The uvcvideo code has no lock protection against USB disconnects
> while video operations are ongoing. This has resulted in random
> error reports, typically pointing to a crash in usb_ifnum_to_if(),
> called from usb_hcd_alloc_bandwidth(). A typical traceback is as
> follows.
>
> usb 1-4: USB disconnect, device number 3
> BUG: unable to handle kernel NULL pointer dereference at 0000000000000000
> PGD 0 P4D 0
> Oops: 0000 [#1] PREEMPT SMP PTI
> CPU: 0 PID: 5633 Comm: V4L2CaptureThre Not tainted 4.19.113-08536-g5d29ca36db06 #1
> Hardware name: GOOGLE Edgar, BIOS Google_Edgar.7287.167.156 03/25/2019
> RIP: 0010:usb_ifnum_to_if+0x29/0x40
> Code: <...>
> RSP: 0018:ffffa46f42a47a80 EFLAGS: 00010246
> RAX: 0000000000000000 RBX: 0000000000000000 RCX: ffff904a396c9000
> RDX: ffff904a39641320 RSI: 0000000000000001 RDI: 0000000000000000
> RBP: ffffa46f42a47a80 R08: 0000000000000002 R09: 0000000000000000
> R10: 0000000000009975 R11: 0000000000000009 R12: 0000000000000000
> R13: ffff904a396b3800 R14: ffff904a39e88000 R15: 0000000000000000
> FS: 00007f396448e700(0000) GS:ffff904a3ba00000(0000) knlGS:0000000000000000
> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 0000000000000000 CR3: 000000016cb46000 CR4: 00000000001006f0
> Call Trace:
> usb_hcd_alloc_bandwidth+0x1ee/0x30f
> usb_set_interface+0x1a3/0x2b7
> uvc_video_start_transfer+0x29b/0x4b8 [uvcvideo]
> uvc_video_start_streaming+0x91/0xdd [uvcvideo]
> uvc_start_streaming+0x28/0x5d [uvcvideo]
> vb2_start_streaming+0x61/0x143 [videobuf2_common]
> vb2_core_streamon+0xf7/0x10f [videobuf2_common]
> uvc_queue_streamon+0x2e/0x41 [uvcvideo]
> uvc_ioctl_streamon+0x42/0x5c [uvcvideo]
> __video_do_ioctl+0x33d/0x42a
> video_usercopy+0x34e/0x5ff
> ? video_ioctl2+0x16/0x16
> v4l2_ioctl+0x46/0x53
> do_vfs_ioctl+0x50a/0x76f
> ksys_ioctl+0x58/0x83
> __x64_sys_ioctl+0x1a/0x1e
> do_syscall_64+0x54/0xde
>
> While there are not many references to this problem on mailing lists, it is
> reported on a regular basis on various Chromebooks (roughly 300 reports
> per month). The problem is relatively easy to reproduce by adding msleep()
> calls into the code.
>
> I tried to reproduce the problem with non-uvcvideo webcams, but was
> unsuccessful. I was unable to get Philips (pwc) webcams to work. gspca
> based webcams don't experience the problem, or at least I was unable to
> reproduce it (The gspa driver does not trigger sending USB messages in the
> open function, and otherwise uses the locking mechanism provided by the
> v4l2/vb2 core).
>
> I don't presume to claim that I found every issue, but this patch series
> should fix at least the major problems.
>
> The patch series was tested exensively on a Chromebook running chromeos-4.19
> and on a Linux system running a v5.8.y based kernel.
>
> v3:
> - In patch 5/5, add missing calls to usb_autopm_put_interface() and kfree()
> to failure code path
>
> v2:
> - Added details about problem frequency and testing with non-uvc webcams
> to summary
> - In patch 4/5, return EPOLLERR instead of -ENODEV on poll errors
> - Fix description in patch 5/5
>
> ----------------------------------------------------------------
> Guenter Roeck (5):
> media: uvcvideo: Cancel async worker earlier
> media: uvcvideo: Lock video streams and queues while unregistering
> media: uvcvideo: Release stream queue when unregistering video device
> media: uvcvideo: Protect uvc queue file operations against disconnect
> media: uvcvideo: Abort uvc_v4l2_open if video device is unregistered
>
> drivers/media/usb/uvc/uvc_ctrl.c | 11 ++++++----
> drivers/media/usb/uvc/uvc_driver.c | 12 ++++++++++
> drivers/media/usb/uvc/uvc_queue.c | 32 +++++++++++++++++++++++++--
> drivers/media/usb/uvc/uvc_v4l2.c | 45 ++++++++++++++++++++++++++++++++++++--
> drivers/media/usb/uvc/uvcvideo.h | 1 +
> 5 files changed, 93 insertions(+), 8 deletions(-)

--
Regards,

Laurent Pinchart