uvcvideo doesn't work with USB webcam with linux 3.16

[Rocko]

uvcvideo doesn't work with linux 3.16 on at least two USB webcams.

guvcvideo shows a black screen and reports "Could not grab image" or "ignoring empty buffer".

The camera works fine with Motion or Skype in linux 3.16. guvcvideo works fine with linux 3.15.

There is more info at https://bugzilla.kernel.org/show_bug.cgi?id=81611.

ProblemType: Bug
DistroRelease: Ubuntu 14.10
Package: linux-image-3.16.0-10-generic 3.16.0-10.15
ProcVersionSignature: Ubuntu 3.16.0-10.15-generic 3.16.1
Uname: Linux 3.16.0-10-generic x86_64
ApportVersion: 2.14.6-0ubuntu2
Architecture: amd64
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: rocko 1894 F.... pulseaudio
CurrentDesktop: Unity
Date: Wed Aug 27 17:26:27 2014
HibernationDevice: RESUME=UUID=e586cc1f-5429-445e-b3fa-d15a1f5b54a0
InstallationDate: Installed on 2014-08-22 (5 days ago)
InstallationMedia: Ubuntu 14.10 "Utopic Unicorn" - Alpha amd64 (20140821)
IwConfig:
 eth0 no wireless extensions.

 lo no wireless extensions.
Lsusb:
 Bus 001 Device 002: ID 0408:2fb1 Quanta Computer, Inc.
 Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
 Bus 002 Device 002: ID 80ee:0021 VirtualBox USB Tablet
 Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
MachineType: innotek GmbH VirtualBox
ProcFB: 0 VESA VGA
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-3.16.0-10-generic root=UUID=fa1da5bc-d7d9-40bf-9d26-ad7bcf1ab6bc ro quiet splash vt.handoff=7
RelatedPackageVersions:
 linux-restricted-modules-3.16.0-10-generic N/A
 linux-backports-modules-3.16.0-10-generic N/A
 linux-firmware 1.132
RfKill:

SourcePackage: linux
UpgradeStatus: No upgrade log present (probably fresh install)
dmi.bios.date: 12/01/2006
dmi.bios.vendor: innotek GmbH
dmi.bios.version: VirtualBox
dmi.board.name: VirtualBox
dmi.board.vendor: Oracle Corporation
dmi.board.version: 1.2
dmi.chassis.type: 1
dmi.chassis.vendor: Oracle Corporation
dmi.modalias: dmi:bvninnotekGmbH:bvrVirtualBox:bd12/01/2006:svninnotekGmbH:pnVirtualBox:pvr1.2:rvnOracleCorporation:rnVirtualBox:rvr1.2:cvnOracleCorporation:ct1:cvr:
dmi.product.name: VirtualBox
dmi.product.version: 1.2
dmi.sys.vendor: innotek GmbH

[rockorequin (rockorequin-linux-kernel-bugs)]

Created attachment 145021
guvcview output, uvc trace output, lsusb output

guvcview used to work fine with kernel 3.15 and earlier with my laptop webcam, but with kernel 3.16 (amd64) it just shows (and captures) a black screen and reports "Could not grab image (select timeout): Resource temporarily unavailable" or "ignoring empty buffer".

With 3.15 it would complain that the resource was temporarily unavailable once or twice, but would then recover and display streaming video.

I built the kernel using the same config as the ubuntu mainline builds for 3.16.

'motion' and 'skype' are both able to capture video with 3.16, so the underlying v4l driver looks like it is working.

Is video (other) the right component to file this under?

[chrylis (chrylis-linux-kernel-bugs)]

I can confirm identical behavior on a Dell M6600 integrated webcam. The device has USB ID 1bcf:280b and is listed as "Sunplus Innovation Technology Inc.".

[rockorequin (rockorequin-linux-kernel-bugs)]

I tested and it occurs with 3.16-rc2, so it regressed fairly early on. I couldn't test with 3.16-rc1 in the VM I set up because it failed to boot.

[Brad Figg (brad-figg)]

This change was made by a bot.

[Joseph Salisbury (jsalisbury)]

Would it be possible for you to test the latest upstream kernel? Refer to https://wiki.ubuntu.com/KernelMainlineBuilds . Please test the latest v3.17 kernel[0].

If this bug is fixed in the mainline kernel, please add the following tag 'kernel-fixed-upstream'.

If the mainline kernel does not fix this bug, please add the tag: 'kernel-bug-exists-upstream'.

If you are unable to test the mainline kernel, for example it will not boot, please add the tag: 'kernel-unable-to-test-upstream'.
Once testing of the upstream kernel is complete, please mark this bug as "Confirmed".

Thanks in advance.

[0] http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.17-rc2-utopic/

[Rocko (rockorequin)]

Same problem in 3.17-rc2.

[Joseph Salisbury (jsalisbury)]

Just to confirm, booting an earlier kernel version will make this bug go away?

[Rocko (rockorequin)]

Yes, it works fine with 3.15.10-031510-generic.

[Joseph Salisbury (jsalisbury)]

This issue appears to be an upstream bug, since you tested the latest upstream kernel. Would it be possible for you to open an upstream bug report[0]? That will allow the upstream Developers to examine the issue, and may provide a quicker resolution to the bug.

Please follow the instructions on the wiki page[0]. The first step is to email the appropriate mailing list. If no response is received, then a bug may be opened on bugzilla.kernel.org.

Once this bug is reported upstream, please add the tag: 'kernel-bug-reported-upstream'.

[0] https://wiki.ubuntu.com/Bugs/Upstream/kernel

[Joseph Salisbury (jsalisbury)]

If this issue is not already known upstream, we can perform a kernel bisect to identify the commit that introduced the regression.

[Joseph Salisbury (jsalisbury)]

Sorry, there is already an upstream bug report :-) We can perform a bisect then

[Joseph Salisbury (jsalisbury)]

So, 3.16-rc2 is the first kernel that exhibts this bug, and you cannot test -rc1? If so, we can bisect between 3.15 final and 3.16-rc2. However, it may be time consuming until we find out what is preventing -rc1 from booting for you.

[Rocko (rockorequin)]

3.16-rc1 doesn't boot on my main laptop, either, even in recovery mode. I can see a bunch of "[169198.816426] init: plymouth-upstart-bridge respawning too fast, stopped" messages in the log.

There doesn't seem to be any interest in the upstream bug, which I reported four weeks ago - is it not the right place to report it?

[Joseph Salisbury (jsalisbury)]

Did you not receive a response from upstream or was the response that there was no interest?

The -rc4 kernel is now out. Can you also see if this version is affected by the bug?

http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.17-rc4-utopic/

[Rocko (rockorequin)]

I haven't received any response from the bugzilla report (see the link I posted, https://bugzilla.kernel.org/show_bug.cgi?id=81611).

I tried 3.17-rc5 but it failed to create the /dev/videoX file, so guvcview just reported it couldn't find a video device.

[Rocko (rockorequin)]

With 3.17-rc6, the VM recognises the video device again, but guvcview segfaults.

[Rocko (rockorequin)]

I tried 3.17-rc6 on the laptop instead of in a VM, and guvcview doesn't segfault, but exhibits the same error as originally reported for 3.16 in this bug.

[rockorequin (rockorequin-linux-kernel-bugs)]

The bug is still present in both 3.16.3 and 3.17-rc6.

I started attempting to bisect, but 3.16-rc1 doesn't get past the very early boot stages on the laptop I first saw the problem on (I can't even get a tty prompt in recovery mode).

[Paulo Assis (pj-assis)]

Hi,
after increasing verbosity on the driver and get:
EOF on empty payload
this is consistent with the zero sized frames returned by the driver.
I'll try to install the latest media tree:
http://git.linuxtv.org/cgit.cgi/media_build.git/about/
and check if this issue is fixed upstream

I've also tested with gspca driver (eye toy camera) and it worked fine.
I'll report this at the linux media mailing list.

Regards,
Paulo

[suse (suse-linux-kernel-bugs)]

I was able to bisect and it looks like commit 61bd8fb37db948154ae01c8ed43a86633ae1d2be
introduces the black screen.

Using v3.16 and reverting 61bd8fb37db948154ae01c8ed43a86633ae1d2be works in virtualbox.

I will try to test the fix on real hardware soon.

[suse (suse-linux-kernel-bugs)]

OK, laptop test successfully finished.
with kernel v3.17-rc6 I need to revert two commits:
8a75ffb81b1c1b6949d191fbef3eaa03fd648852 [media] vb2: fix bytesused == 0 handling
and again
61bd8fb37db948154ae01c8ed43a86633ae1d2be [media] vb2: if bytesused is 0, then fill with output buffer length

Reverting both commits makes my two webcams working again.
(integrated webcam 04f2:b1b4 and external 046d:09a4)

[Paulo Assis (pj-assis)]

I've set a warkaround in guvcview for this buggy behaviour, if bytesused is zero guvcview will now use the length field.
It serves as a workaround but this clearly violates the v4l2 api:
http://linuxtv.org/downloads/v4l-dvb-apis/buffer.html

bytesused - ... Drivers must set this field when type refers to an input stream, ...

you can try my testing ppa:
https://code.launchpad.net/~pj-assis/+archive/ubuntu/testing/+packages

build is pending, but it should be available in an hour or so.

[hverkuil (hverkuil-linux-kernel-bugs)]

I think you hit this guvcview bug:

http://<email address hidden>/msg80144.html

So it's not a driver bug, it's a bug in the application which was exposed by the driver change.

[Oleksij Rempel (olerem)]

seems to be fixed:
http://git.linuxtv.org/cgit.cgi/pinchartl/media.git/commit/?h=uvcvideo/fixes&id=a7f053c67c357c4b68c1be21976a1d464f97916b

but currently not yet upstream.

[Pilot6 (hanipouspilot)]

Now it is fixed upstream. It needs to be backported
https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/?id=c601f53f8fe5aab4d8b506104d0fd0a7b6a19922

[Andrew Pogrebennyk (andrew-nau-ua)]

I'm not sure if this is of any help, but I've had trouble with Suyin webcam on my Asus laptop running the 3.16.0-31 kernel. Kernel 3.13.0-37.64 worked well...

Here is the hw id:
Bus 001 Device 003: ID 064e:9700 Suyin Corp.
in skype the video screen from me and remote was black and I noticed error
libv4l2: error dequeuing buf: Invalid argument
and guvcview didn't show any video either. It was just flooding the console with "Ignoring empty buffer ..." messages.

I've tried testing ppa from Paulo and after installing the following packages:
ii guvcview 2.0.1+ubuntu2~ppa1+1278-0ubuntu0~201501091246~ubuntu14.04.1 amd64 GTK+ base UVC Viewer
ii libguvcview-1.0-1:amd64 2.0.1+ubuntu2~ppa1+1278-0ubuntu0~201501091246~ubuntu14.04.1 amd64 GTK+ base UVC Viewer - runtime
the problem magically went away - in skype too..

[Matthias Homann (palto)]

The solution in #20 worked for me as well with kernel 3.16.0-38 on Ubuntu 14.04 .2
With the older 14.04 kernel 3.13.0-53 it also worked with the version guvcview 2.0.0 from the stable ppa.

[Michel-Ekimia (michel.ekimia)]

That's a real serious problem

Here on 14.04.3 fresh kernel 3.19 , no application can access the UVC webcam.

SOlution on #20 does not work on 14.04.3 , don't understand why

[Rolf Leggewie (r0lf)]

utopic has seen the end of its life and is no longer receiving any updates. Marking the utopic task for this ticket as "Won't Fix".

[suse (suse-linux-kernel-bugs)]

bug report should be closed.