Package Details: decklink 12.4-1

Git Clone URL: (read-only, click to copy)
Package Base: decklink
Description: Drivers for Blackmagic Design DeckLink, Intensity or Multibridge video editing cards
Upstream URL:
Licenses: custom
Submitter: alub
Maintainer: goldensuneur
Last Packager: goldensuneur
Votes: 23
Popularity: 0.30
First Submitted: 2015-10-03 15:38 (UTC)
Last Updated: 2022-07-14 10:23 (UTC)

Pinned Comments

goldensuneur commented on 2020-03-02 19:00 (UTC)

I finally had time to investigate the issue. It looks like that as of 11.5, the DesktopVideoHelper.service has to be running for the capture card to work properly.

Latest Comments

goldensuneur commented on 2022-06-06 13:05 (UTC)

Thanks for the fix, I updated the package with a patch inspired from yours to make sure it's compatible with previous kernels.

tjb0607 commented on 2022-06-06 05:07 (UTC) (edited on 2022-06-06 08:34 (UTC) by tjb0607)

Martin8412's patches confirmed working for me on 5.18.1-arch1-1

Martin8412 commented on 2022-06-03 10:29 (UTC) (edited on 2022-06-03 10:30 (UTC) by Martin8412)

I spent a few minutes fixing it. These two patches fixes compile issue for me on 5.18.1-arch1-1

diff blackmagic-12.3a10/blackmagic_core.c /usr/src/blackmagic-12.3a10/blackmagic_core.c
<   if (pci_set_dma_mask(pdev, BMD_DMA_64_MASK) < 0)
>   if (dma_set_mask(&pdev->dev, BMD_DMA_64_MASK) < 0)
<       if (pci_set_dma_mask(pdev, BMD_DMA_32_MASK) < 0)
>       if (dma_set_mask(&pdev->dev, BMD_DMA_32_MASK) < 0)
diff blackmagic-12.3a10/blackmagic_dma.c /usr/src/blackmagic-12.3a10/blackmagic_dma.c
<       case DL_DMA_TO_DEVICE: return PCI_DMA_TODEVICE;
>       case DL_DMA_TO_DEVICE: return DMA_TO_DEVICE;
>       case DL_DMA_FROM_DEVICE: return DMA_FROM_DEVICE;
<   return PCI_DMA_NONE;
>   return DMA_NONE;
<       e->dma_addr = pci_map_page(pdev, pages[i], 0, PAGE_SIZE, direction);
>       e->dma_addr = dma_map_page(pdev, pages[i], 0, PAGE_SIZE, direction);
<           e->dma_addr = pci_map_page(pdev, page, 0, PAGE_SIZE, direction);
>           e->dma_addr = dma_map_page(pdev, page, 0, PAGE_SIZE, direction);
<       e->dma_addr = pci_map_single(pdev, address, size, direction);
>       e->dma_addr = dma_map_single(pdev, address, size, direction);
<           pci_unmap_page(sl->pdev, e->dma_addr, PAGE_SIZE, direction);
>           dma_unmap_page(&sl->pdev->dev, e->dma_addr, PAGE_SIZE, direction);
<       pci_unmap_single(sl->pdev, e->dma_addr, sl->size, direction);
>       dma_unmap_single(&sl->pdev->dev, e->dma_addr, sl->size, direction);


diff blackmagic-io-12.3a10/bm_mm.c /usr/src/blackmagic-io-12.3a10/bm_mm.c
<   #define bm_pci_dma_mapping_error(dev, addr) pci_dma_mapping_error(dev, addr)
>   #define bm_pci_dma_mapping_error(dev, addr) dma_mapping_error(dev, addr)
<   #define bm_pci_dma_mapping_error(dev, addr) pci_dma_mapping_error(addr)
>   #define bm_pci_dma_mapping_error(dev, addr) dma_mapping_error(addr)
<       dlist->addrs[i] = pci_map_page(pci->pdev, umem->pages[i], 0, PAGE_SIZE, dir);
<       if (bm_pci_dma_mapping_error(pci->pdev, dlist->addrs[i]))
>       dlist->addrs[i] = dma_map_page(&pci->pdev->dev, umem->pages[i], 0, PAGE_SIZE, dir);
>       if (bm_pci_dma_mapping_error(&pci->pdev->dev, dlist->addrs[i]))
<   dlist->addrs[0] = pci_map_single(pci->pdev, addr, size, dir);
<   if (bm_pci_dma_mapping_error(pci->pdev, dlist->addrs[0]))
>   dlist->addrs[0] = dma_map_single(&pci->pdev->dev, addr, size, dir);
>   if (bm_pci_dma_mapping_error(&pci->pdev->dev, dlist->addrs[0]))
<       dlist->addrs[i] = pci_map_page(pci->pdev, page, 0, PAGE_SIZE, dir);
<       if (bm_pci_dma_mapping_error(pci->pdev, dlist->addrs[i]))
>       dlist->addrs[i] = dma_map_page(&pci->pdev->dev, page, 0, PAGE_SIZE, dir);
>       if (bm_pci_dma_mapping_error(&pci->pdev->dev, dlist->addrs[i]))
<       pci_unmap_single(pci->pdev, dlist->addrs[0], dlist->length, dir);
>       dma_unmap_single(&pci->pdev->dev, dlist->addrs[0], dlist->length, dir);
<               pci_unmap_page(pci->pdev, dlist->addrs[i], PAGE_SIZE, dir);
>               dma_unmap_page(&pci->pdev->dev, dlist->addrs[i], PAGE_SIZE, dir);
diff blackmagic-io-12.3a10/bm_pci.c /usr/src/blackmagic-io-12.3a10/bm_pci.c
<   if (pci_set_dma_mask(pci->pdev, BM_DMA_64BIT_MASK) < 0)
>   if (dma_set_mask(&pci->pdev->dev, BM_DMA_64BIT_MASK) < 0)
<       if (pci_set_dma_mask(pci->pdev, BM_DMA_32BIT_MASK) < 0)
>       if (dma_set_mask(&pci->pdev->dev, BM_DMA_32BIT_MASK) < 0)

Serus commented on 2022-06-01 22:52 (UTC)

@CareAgain Send an email to developer'at'

CareAgain commented on 2022-06-01 20:34 (UTC)

It looks like this kernel patch is what broke it:

Basically just a deprecated API getting removed. It doesn't look too hard to fix. I actually messed around with a patch, but didn't quite get it working. If anyone knows how to contact blackmagic that might be the way to go.

daurnimator commented on 2022-05-30 16:15 (UTC)

DKMS compilation currently fails:

DKMS make.log for blackmagic-io-12.3a10 for kernel 5.18.0-arch1-1 (x86_64)
Tue 31 May 2022 02:15:07 AEST
touch .blackmagic.o.cmd
make -C /lib/modules/5.18.0-arch1-1/build M=/var/lib/dkms/blackmagic-io/12.3a10/build
make[1]: Entering directory '/usr/lib/modules/5.18.0-arch1-1/build'
  CC [M]  /var/lib/dkms/blackmagic-io/12.3a10/build/bmio_client.o
  CC [M]  /var/lib/dkms/blackmagic-io/12.3a10/build/bmio_dev.o
  CC [M]  /var/lib/dkms/blackmagic-io/12.3a10/build/bmio_device.o
  CC [M]  /var/lib/dkms/blackmagic-io/12.3a10/build/bmio_driver.o
  CC [M]  /var/lib/dkms/blackmagic-io/12.3a10/build/bmio_serial.o
  CC [M]  /var/lib/dkms/blackmagic-io/12.3a10/build/bmio_export.o
  CC [M]  /var/lib/dkms/blackmagic-io/12.3a10/build/bmio_pci_ids.o
  CC [M]  /var/lib/dkms/blackmagic-io/12.3a10/build/bm_locks.o
  CC [M]  /var/lib/dkms/blackmagic-io/12.3a10/build/bm_mm.o
  CC [M]  /var/lib/dkms/blackmagic-io/12.3a10/build/bm_pci.o
  CC [M]  /var/lib/dkms/blackmagic-io/12.3a10/build/bm_util.o
  COPY    /var/lib/dkms/blackmagic-io/12.3a10/build/blackmagic.o
  CC [M]  /var/lib/dkms/blackmagic-io/12.3a10/build/bmio_audio.o
/var/lib/dkms/blackmagic-io/12.3a10/build/bm_mm.c: In function ‘bm_dma_map_user_buffer’:
/var/lib/dkms/blackmagic-io/12.3a10/build/bm_mm.c:226:35: error: implicit declaration of function ‘pci_map_page’; did you mean ‘dma_map_page’? [-Werror=implicit-function-declaration]
  226 |                 dlist->addrs[i] = pci_map_page(pci->pdev, umem->pages[i], 0, PAGE_SIZE, dir);
      |                                   ^~~~~~~~~~~~
      |                                   dma_map_page
/var/lib/dkms/blackmagic-io/12.3a10/build/bm_pci.c: In function ‘bm_pci_start’:
/var/lib/dkms/blackmagic-io/12.3a10/build/bm_pci.c:79:13: error: implicit declaration of function ‘pci_set_dma_mask’ [-Werror=implicit-function-declaration]
   79 |         if (pci_set_dma_mask(pci->pdev, BM_DMA_64BIT_MASK) < 0)
      |             ^~~~~~~~~~~~~~~~
/var/lib/dkms/blackmagic-io/12.3a10/build/bm_mm.c:58:53: error: implicit declaration of function ‘pci_dma_mapping_error’; did you mean ‘bm_pci_dma_mapping_error’? [-Werror=implicit-function-declaration]
   58 |         #define bm_pci_dma_mapping_error(dev, addr) pci_dma_mapping_error(dev, addr)
      |                                                     ^~~~~~~~~~~~~~~~~~~~~
/var/lib/dkms/blackmagic-io/12.3a10/build/bm_mm.c:227:21: note: in expansion of macro ‘bm_pci_dma_mapping_error’
  227 |                 if (bm_pci_dma_mapping_error(pci->pdev, dlist->addrs[i]))
      |                     ^~~~~~~~~~~~~~~~~~~~~~~~
/var/lib/dkms/blackmagic-io/12.3a10/build/bm_mm.c: In function ‘bm_dma_map_kernel_buffer’:
/var/lib/dkms/blackmagic-io/12.3a10/build/bm_mm.c:249:27: error: implicit declaration of function ‘pci_map_single’; did you mean ‘dma_map_single’? [-Werror=implicit-function-declaration]
  249 |         dlist->addrs[0] = pci_map_single(pci->pdev, addr, size, dir);
      |                           ^~~~~~~~~~~~~~
      |                           dma_map_single
/var/lib/dkms/blackmagic-io/12.3a10/build/bm_mm.c: In function ‘bm_dma_unmap_buffer’:
/var/lib/dkms/blackmagic-io/12.3a10/build/bm_mm.c:295:17: error: implicit declaration of function ‘pci_unmap_single’; did you mean ‘dma_unmap_single’? [-Werror=implicit-function-declaration]
  295 |                 pci_unmap_single(pci->pdev, dlist->addrs[0], dlist->length, dir);
      |                 ^~~~~~~~~~~~~~~~
      |                 dma_unmap_single
cc1: some warnings being treated as errors
make[2]: *** [scripts/ /var/lib/dkms/blackmagic-io/12.3a10/build/bm_pci.o] Error 1
make[2]: *** Waiting for unfinished jobs....
/var/lib/dkms/blackmagic-io/12.3a10/build/bm_mm.c:304:33: error: implicit declaration of function ‘pci_unmap_page’; did you mean ‘dma_unmap_page’? [-Werror=implicit-function-declaration]
  304 |                                 pci_unmap_page(pci->pdev, dlist->addrs[i], PAGE_SIZE, dir);
      |                                 ^~~~~~~~~~~~~~
      |                                 dma_unmap_page
/var/lib/dkms/blackmagic-io/12.3a10/build/bm_mm.c: In function ‘bm_dma_get_physical_segment’:
/var/lib/dkms/blackmagic-io/12.3a10/build/bm_mm.c:331:9: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement]
  331 |         addr64_t startAddr = dlist->addrs[page_n] + page_offset;
      |         ^~~~~~~~
cc1: some warnings being treated as errors
make[2]: *** [scripts/ /var/lib/dkms/blackmagic-io/12.3a10/build/bm_mm.o] Error 1
make[1]: *** [Makefile:1834: /var/lib/dkms/blackmagic-io/12.3a10/build] Error 2
make[1]: Leaving directory '/usr/lib/modules/5.18.0-arch1-1/build'
make: *** [Makefile:47: all] Error 2

goldensuneur commented on 2022-03-22 08:44 (UTC)

Unfortunatly I don't have working blackmagic hardware anymore so I cannot help you. You should ask directly on their forum or open a ticket with their support.

lorddickfist commented on 2022-03-22 02:38 (UTC)

Blackmagic Intensity pro 4k is unable to be seen but is visible on windows and DesktopVideoHelper.service is enabled and started

idella.craddock commented on 2022-02-28 19:30 (UTC) (edited on 2022-02-28 19:31 (UTC) by idella.craddock)

if using thunderbolt blackmagic devices this appears to fail. (with recorder 3g for me, latest linux kernel 5.16.11-arch1-1). Kernel driver chokes and segfaults

Nember commented on 2022-02-10 12:14 (UTC) (edited on 2022-03-10 17:48 (UTC) by Nember)

I have installed my intensity pro 4K and this package and the desktop video setup cant find my card, despite it being visible in hardware settings. I'm on kernel 5.15.19, has an update broke this since it was last working?

goldensuneur commented on 2021-12-11 20:38 (UTC)

I also think that it is better to introduce as few patches as necessary. For now it is not an issue but I suggest you to inform blackmagic that they should update before it becomes one.

ToadKing commented on 2021-12-11 20:26 (UTC)

@Ammako dkms.conf is provided by the package, so as long as it still builds it seems like overkill to introduce a patch just to remove it.

Ammako commented on 2021-12-11 19:28 (UTC) (edited on 2021-12-11 19:29 (UTC) by Ammako)

REMAKE_INITRD is deprecated and should probably be removed from dkms.conf

It doesn't really cause problems, but it would probably be best practice that way, to get rid of warnings.

goldensuneur commented on 2021-11-10 13:02 (UTC)

Thanks, I pushed a new version with a patch inspired by yours. Let me know if it does not work on kernel 5.15 and newer.

Mischa85 commented on 2021-11-04 16:35 (UTC)

Patch for kernel 5.15, which deprecated alloc_tty_driver and put_tty_driver:

--- package.orig/Blackmagic_Desktop_Video_Linux_12.2/other/x86_64/desktopvideo-12.2a12-x86_64/usr/src/blackmagic-12.2a12/blackmagic_serial.c    2021-11-04 17:18:37.053086911 +0100
+++ 2021-11-04 17:30:49.791835250 +0100
@@ -613,7 +613,7 @@
    int ret;
    struct tty_driver *driver;

-   driver = alloc_tty_driver(BLACKMAGIC_SERIAL_MINORS);
+   driver = tty_alloc_driver(BLACKMAGIC_SERIAL_MINORS, 0);
    if (!driver)
        return -ENOMEM;

@@ -639,7 +639,7 @@

    printk(KERN_ERR "failed to register blackmagic serial driver");
-   put_tty_driver(driver);
+   tty_driver_kref_put(driver);
    return ret;

@@ -648,6 +648,6 @@
    if (blackmagic_tty_driver)
-       put_tty_driver(blackmagic_tty_driver);
+       tty_driver_kref_put(blackmagic_tty_driver);
diff --unified --recursive --text --color package.orig/Blackmagic_Desktop_Video_Linux_12.2/other/x86_64/desktopvideo-12.2a12-x86_64/usr/src/blackmagic-io-12.2a12/bmio_serial.c
--- package.orig/Blackmagic_Desktop_Video_Linux_12.2/other/x86_64/desktopvideo-12.2a12-x86_64/usr/src/blackmagic-io-12.2a12/bmio_serial.c   2021-11-04 17:18:37.053086911 +0100
+++    2021-11-04 17:29:38.749443820 +0100
@@ -304,7 +304,7 @@
    int ret;

-   struct tty_driver* serial = alloc_tty_driver(MAX_SERIAL_LINES);
+   struct tty_driver* serial = tty_alloc_driver(MAX_SERIAL_LINES, 0);
    if (!serial)
        return -ENOMEM;

@@ -324,7 +324,7 @@
    ret = tty_register_driver(serial);
    if (ret)
-       put_tty_driver(serial);
+       tty_driver_kref_put(serial);
        return ret;

@@ -340,7 +340,7 @@
    if (serial_driver)
-       put_tty_driver(serial_driver);
+       tty_driver_kref_put(serial_driver);
        serial_driver = NULL;

rama_g commented on 2021-09-14 15:30 (UTC)

awesome, also working here on 5.14.3-arch1-1 -- happy to be able to update the system, was holding off until now since I rely on the decklink drivers, using it everyday -- thanks for the package update! FWIW it did require a reboot for the device to be found by BlackmagicDesktopVideoSetup.

YAOMTC commented on 2021-09-13 04:19 (UTC) (edited on 2021-09-13 08:27 (UTC) by YAOMTC)

Can confirm it works for me too with 5.14.2.arch1-2. Thanks for updating the PKGBUILD.

EDIT: Though, the audio stutters quite badly. I'll have to see if I can get that fixed...

EDIT2: happens on Windows too, I think it may have to do with PCI bandwidth available. I get a warning about dropped frames when I record with MediaExpress. But there's not as many dropped frames when I record a 720p60 signal vs 1080p60.

EDIT3: yeah that was it. Unplugging the TV tuner card I never use gave me the PCI bandwidth the card needs to work properly.

frostwork commented on 2021-09-10 13:22 (UTC) (edited on 2021-09-10 13:37 (UTC) by frostwork)

edit - sorry, I don't know the half markdown syntax here. deleted my PKGBUILD patch.

edit2: sorry for the lame workaround - uploaded a PKGBUILD gist here:

edit3: I can confirm that the decklink works fine with 5.14.2-xanmod1-1 with the patched drivers

frostwork commented on 2021-09-10 13:09 (UTC)

thx for the quick reply! just gave it a quick try. The patch applies fine, but when manually building the build fails with "bmio_serial.c:174:28: Fehler: Initialisierung von »unsigned int ()(struct tty_struct )« von inkompatiblem Zeigertyp »int ()(struct tty_struct )« [-Werror=incompatible-pointer-types]" (the other module blackmagic-12.1a9 has the same error). the Werror can be disabled by adding "-Wno-error=incompatible-pointer-types" to the EXTRA_CFLAGS in the Makefiles. Both modules build fine then. Haven't tested functionality, as I'm almost off for a short trip \o/ :) Good luck and thanks a lot for maintaining the PKGBUILD! :)

goldensuneur commented on 2021-09-10 13:04 (UTC)

thx, I'll try the patch later today after work. If it builds on my machine I'll push a new revision to the decklink package.

frostwork commented on 2021-09-10 12:56 (UTC)

fyi: a blackmagic dev posted a patch in the thread mentioned before here: haven't tested it yet

frostwork commented on 2021-09-07 03:27 (UTC)

fyi, upstream forum issue thread:

YAOMTC commented on 2021-08-26 01:23 (UTC)

About a month ago I had sent in my card to their technical support team, thinking it might have been a hardware issue. The other day I finally heard back from them, and they said it was working on Windows. They said they did not have a Linux machine to test with (despite advertising Linux support on the box) and were not able (allowed?) to run a live DVD. They are now shipping it back to me.

frostwork commented on 2021-08-19 14:42 (UTC)

@scaramanga I just coincidentally found out that the drivers are currently broken (I barely use the card) and therefore found your comment. Thanks for the heads-up & contacting upstream! :)

scaramanga commented on 2021-08-14 10:32 (UTC)

Short update: After almost a week Customer Support told me they won't help with Linux "due to the high level of user customisation and configurability", but referred me to the Developer Support team.

There they actually helped and provided a fix. It sounded like there should be an official update to the driver some time soon, but they didn't say a date.

goldensuneur commented on 2021-07-29 13:20 (UTC)

Yes, the module doesn't build on 5.13.5-arch1-1 and unfortunately I'm not sure I can patch this error myself.

I suggest you to stay on LTS and report the error to blackmagic support.

tjb0607 commented on 2021-07-28 20:18 (UTC)

the dkms modules don't build on 5.13.5-arch1-1, either. Running off of LTS for now.

  MODPOST /var/lib/dkms/blackmagic-io/12.1a9/build/Module.symvers
ERROR: modpost: "X86_FEATURE_LA57" [/var/lib/dkms/blackmagic-io/12.1a9/build/blackmagic-io.ko] undefined!

scaramanga commented on 2021-07-23 14:22 (UTC) (edited on 2021-07-23 14:22 (UTC) by scaramanga)

No device is recognized for me after upgrading the kernel from 5.12 to 5.13.4-arch1-1. Might be a regression in the kernel.

For now I have to fall back to LTS. Did anyone else experience that?

YAOMTC commented on 2021-07-16 20:36 (UTC) (edited on 2021-07-17 00:27 (UTC) by YAOMTC)

Oh yes, I tested it with that, too. I also tried the card in Linux Mint, and on a different PC running Ubuntu 21.04, and then the first PC again but I had replaced Linux Mint with Ubuntu 20.04 LTS. Nothing but black capture with all of them, with a docked Nintendo Switch attached to it via HDMI. This may be a faulty unit, though the seller said it was never used. I'm corresponding with BMD now.

goldensuneur commented on 2021-07-16 07:47 (UTC)

Is the DesktopVideoHelper.service enabled and runnning ? I remember having weird issue like that when it is not.

Otherwise I won't be able to help you much. I don't own a blackmagic capture card anymore, my old one broke.

YAOMTC commented on 2021-07-15 21:38 (UTC) (edited on 2021-07-16 04:08 (UTC) by YAOMTC)

I can't get my Intensity Pro 4K working. Desktop Video Setup detects it just fine (I wish this included some troubleshooting utility) but OBS doesn't see it. Here's where blackmagic shows up in dmesg:

blackmagic_io: loading out-of-tree module taints kernel.
blackmagic_io: module license 'Proprietary' taints kernel.
blackmagic_io: module verification failed: signature and/or required key missing - tainting kernel
blackmagic-io 0000:05:00.0: enabling device (0000 -> 0002)
BlackmagicIO: Intensity Pro 4K as blackmagic!io0 [0000:05:00.0]
BlackmagicIO: Initialised serial driver for blackmagic!io0

I could try the LTS kernel, but the drivers are being initialized, so maybe that's not the issue? What else should I try? I've tested /dev/video0 through /dev/video3 and found they were my other video devices (Index cameras & webcam). You'd think this would be /dev/video4 but it doesn't exist. EDIT: Using linux-lts didn't help. EDIT2: Neither did using Linux Mint. Using Media Express there just shows black. I'll try another PCIe slot. EDIT3: That didn't work either. I've emailed BMD support.

ericfont commented on 2021-04-06 00:52 (UTC)

Waking from suspend results in threads for "DesktopVideoUpdater" and two threads for "DesktopVideoHelper" consuming an entire 100% of a CPU each, for a total of 400% of a CPU being consumed. This performance bug has been reported to DesktopVideoUpdater and DesktopVideoHelper and comments say it happens on Manjaro, Debian, Mint, and Arch.

I am also noting that video capture in OBS doesn't work on the realtime kernel (linux-rt)...simply produces a black image in OBS and freezes OBS.

lorddickfist commented on 2021-01-23 21:05 (UTC) (edited on 2021-01-24 06:59 (UTC) by lorddickfist)

thanks for the info, will try that as soon as I figure out how :)

MAJOR EDIT: Was able to fix by switching to a different obs package. I ended up switching to obs-studio-browser. It doesn't seem to have anything missing from the original package, but it is updated to 26.1.1

daren commented on 2021-01-23 20:39 (UTC) (edited on 2021-01-23 20:39 (UTC) by daren)

Wrote this a couple posts back before it got buried.

You need to update OBS to 26.1.1 by adapting the PKGBUILD.

For some reason the Arch repo version is still stuck at 26.1.0 containing a bug for Intensity cards (I had the all green thing as well).

lorddickfist commented on 2021-01-23 20:32 (UTC)

@daren can you share all of your obs settings for the blackmagic because I get the same issue after updating to 12.0

YUV doesn't show anything but green

daren commented on 2021-01-23 18:17 (UTC)

Yes, no issues using my Blackmagic Intensity Pro with OBS 26.1.1 feeding a 720p60 HDMI input.

Sebiann commented on 2021-01-23 17:53 (UTC)

@daren So you don't get the half purple half black screen?

daren commented on 2021-01-23 09:22 (UTC)

It works for me post 12.0 update on latest 5.10.9 kernel. Use other cables and other input sources as well, clean HDMI slot(s) maybe. Just to rule these things out.

lorddickfist commented on 2021-01-23 07:49 (UTC)


Well The box was open and the seal in the static bag was broken so I am trying a replacement first because I only get those half purple half black input if the resolution is manually matched to the ps5

If issue persists I will leave a thread on their forum, thanks!

goldensuneur commented on 2021-01-22 10:11 (UTC)

Blackmagic just released new drivers however they do not compile on new kernel so I kept the patches I made for older release. As usual, I'm not a kernel dev so my patches may introduce bugs. I encourage you to complain to Blackmagic if you want proper driver for linux 5.10+ (they have been aware of this issue for months and haven't done anything yet) or stick to lts releases.

@lorddickfist I used to get this kind of artifacts when I selected the wrong framerate / resolution.

lorddickfist commented on 2021-01-22 01:49 (UTC)

Getting black screen on auto, and if I set it to 1080p 59.94 I get a purple screen with half black like this:

Mediaexpress same thing DesktopVideoHelper.service starts correctly and recognizes device.

rama_g commented on 2021-01-20 09:29 (UTC)

and... it works!

I first tried LTS but it doesn't seem to be compatible with the newer drivers I need for the RTX 3070.

Then finally, back in an updated kernel 5.10.8, I was digging around in the BlackmagicDesktopVideoSetup and found the connector settings, it defaulted to paired connectors. After configuring the connector to single ports for my setup, I can now see the video feed in CapturePreview! yay!

thanks again for your help with getting it working!

goldensuneur commented on 2021-01-19 17:19 (UTC)

Great !

I checked and there is no BlackmagicFirmwareUpdater in the tar file I can download from Blackmagic website so I don't know why they still have the corresponding man page. I think you could have used DesktopVideoUpdateTool to update your card firmware.

rama_g commented on 2021-01-19 16:01 (UTC) (edited on 2021-01-20 09:21 (UTC) by rama_g)

I will try -lts eventually if I can't get the video feed to show up in CapturePreview

rama_g commented on 2021-01-19 15:54 (UTC)

it was the firmware! I wasn't able to locate the BlackmagicFirmwareUpdater application from the module install (but I did see the man page), so I installed the 11.7 driver on MacOS and updated the card firmware there, and then moved the card back to the new computer, and it now BlackmagicDesktopVideoSetup sees the board! great! no video coming through yet in CapturePreview, but the board info is there, so this seems promising.

thanks for all your help!

rama_g commented on 2021-01-19 15:16 (UTC)

ok sure, I can try -lts also.

lspci -k reports:

24:00.0 Multimedia video controller: Blackmagic Design DeckLink Duo 2
    Subsystem: Blackmagic Design DeckLink Duo 2
    Kernel driver in use: blackmagic-io
    Kernel modules: blackmagic_io

looking at dmesg I see a warning about the card's firmware version being older than the driver, so maybe I will try updating the firmware first -- and then try -lts if that doesn't work. will report back -- cheers! rama

goldensuneur commented on 2021-01-19 14:54 (UTC)

Could you try the module with -lts kernel ?

You can also look for errors related to blackmagic and blackmagic-io in dmesg. Also make sure the card uses the correct driver with lspci -k.

rama_g commented on 2021-01-19 14:31 (UTC)

I just also tested the card on MacOS and it is working there, so the hardware seems fine.

rama_g commented on 2021-01-19 14:17 (UTC)

ok great thank you, so the module seems installed, and the DesktopVideoHelper.service is starting on load. I just did PCI refresh, with a power cycle in between to see if that might force it to load something new, but no luck.

And now just tried loading BlackmagicDesktopVideoSetup then rebooting and loading BlackmagicDesktopVideoSetup again, but also no improvement yet.

when I try to use it with the BlackmagicDesktopVideoSetup app or SDK examples, it says No Desktop Video Devices were found.

CapturePreview loads, but the input ports are not listed.

I have an active SDI signal attached to the card as well.

Any other ideas that I might try?

goldensuneur commented on 2021-01-19 13:56 (UTC)

modprobe returning 0 and displaying nothing is the sign it loaded the module correctly.

If modules are correctly loaded then launch BlackmagicDesktopVideoSetup and should list your card. Otherwise reboot to make sure everything get loaded correctly and try BlackmagicDesktopVideoSetup again.

rama_g commented on 2021-01-19 10:00 (UTC)

sorry, you're right I had mis-typed it, I was just editing my last comment when you replied.

sudo modprobe blackmagic and sudo modprobe blackmagic-io both return empty

but modinfo sees both blackmagic and blackmagic-io as installed, not sure what to try next

have to run to a meeting, but hoping to get this working! maybe I will try reseating the pci board.

goldensuneur commented on 2021-01-19 09:52 (UTC)

Hmm, if dkms reports blackmagic and blackmagic-io as installed then you should be able to load them (and they should load automatically at boot). You don't have anything to copy, dkms handles that for you.

Just to be sure, are you typing the exact command in your previous message or are you typing sudo modprobe blackmagic and sudo modprobe blackmagic-io ? Modprobe expects a single module name so you cannot load both modules like that in a single command.

rama_g commented on 2021-01-19 09:48 (UTC) (edited on 2021-01-19 09:57 (UTC) by rama_g)

ok, I see that it's there in /lib/modules/5.10.7-arch1-1/kernel/drivers hmm

sudo modprobe blackmagic and sudo modprobe blackmagic-io both return empty

but modinfo sees both blackmagic and blackmagic-io as installed, not sure what to try next

rama_g commented on 2021-01-19 09:30 (UTC)

@goldensuneur, thanks! that got me a bit further, after installing the headers sudo dkms status is the same as yours (minus the -lts)

but still the having same issue with the device not being found.

following the Blackmagic Desktop Video Linux tests in the readme, I'm seeing:

$ sudo modprobe blackmagic[-io]
modprobe: FATAL: Module blackmagic[-io] not found in directory /lib/modules/5.10.7-arch1-1

do I need to install the module into this folder this somehow?

I'm new to dealing with installing driver modules on linux, sorry if this is basic stuff, I'll start reading about modprobe --

thanks for your help!

goldensuneur commented on 2021-01-18 18:31 (UTC)

Hi, I suspect that you don't have kernel headers installed.

Does the output of sudo dkms status looks similar than this :

$ sudo dkms status
blackmagic, 11.7a12, 5.10.7-arch1-1, x86_64: installed
blackmagic, 11.7a12, 5.4.89-1-lts, x86_64: installed
blackmagic-io, 11.7a12, 5.10.7-arch1-1, x86_64: installed
blackmagic-io, 11.7a12, 5.4.89-1-lts, x86_64: installed

It will list all dkms kernel modules and their state. Here you can see that blackmagic and blackmagic-io modules are correctly installed for kernels 5.4 and 5.10.

If you don't have kernel headers installed dkms will not compile nor install the module. The package is named after the kernel you use with -headers at the end (linux-headers, linux-lts-headers, ...). Try installing it and check if the modules are correctly built/installed (you may need to reinstall your kernel package to trigger dkms rebuild).

rama_g commented on 2021-01-18 17:00 (UTC)

hello! new arch user here, thanks for your work on this. I'm trying to get the DeckLink Duo 2 driver working on kernal 5.10.7, but so far I'm getting No Blackmagic Design devices were found. when I try to run any of the test programs.

I've enabled the DesktopVideoHelper.service, which is listed active and running.

lspci | grep Blackmagic outputs: 24:00.0 Multimedia video controller: Blackmagic Design DeckLink Duo 2

but lsmod | grep blackmagic comes back empty

is there a step that I'm missing?

daren commented on 2021-01-12 12:33 (UTC) (edited on 2021-01-12 12:34 (UTC) by daren)

In case someone is wondering why their decklink device doesn't work (green screen) with the current obs-studio 26.1.0 from the repo, there is some bug that is now fixed in 26.1.1:

For some reason this isn't updated for days in the repo. Building obs from PKGBUILD with adapted version string and md5sums worked for me.

goldensuneur commented on 2020-11-13 10:07 (UTC)

My intensity card died and I replaced it with a Magewell one. So as of this release I cannot test if new version work on actual hardware.

goldensuneur commented on 2020-10-22 14:01 (UTC)

My intensity pro card is dying. I won't replace it with another Blackmagic product. So I won't actively use the package anymore.

I can continue to update this package. However, if newer kernels break the driver I won't be able to test the patches I provide on actual hardware.

So, if anyone is interested in becoming maintainer for this package let me know.

zangoku commented on 2020-10-19 00:47 (UTC)

Just tested a little bit and it seems to work well with the update on 5.9.1

goldensuneur commented on 2020-10-18 20:31 (UTC)

I just pushed an update with patches for kernels 5.9.1. As usual, I'm not a kernel module developer so I hope my patches do not introduce bugs. Otherwise I recommend you to use the kernel-lts and let blackmagic that their modules to be patched for newer kernels.

mSSM commented on 2020-08-19 13:01 (UTC)

Thank you, ToadKing.

Setting mode to 1080p59.94 and pixel format to 8-bit BGRA on my DeckLink Mini Recorder 4k did it.

ToadKing commented on 2020-08-17 15:52 (UTC)

Same, new version builds and works on my Intensity Pro 4K. Thanks for the quick fix!

Also for people that are running into an issue with the Intensity Pro 4K in OBS, there's currently a bug where you have to set the mode and pixel format manually for it to work:

daren commented on 2020-08-17 14:01 (UTC)

Works for me on kernel 5.8.1 with my Blackmagic Intensity Pro.

goldensuneur commented on 2020-08-17 13:43 (UTC)

I just pushed a new revision with fixes for kernel 5.8.1. On top of the Makefile error I had to patch blackmagic_lib.c to reflect changes in kernel 5.8. Modules build and load correctly however I cannot test them on my card right now so let me know if my second fix is not working.

mSSM commented on 2020-08-17 07:46 (UTC) (edited on 2020-08-17 08:00 (UTC) by mSSM)

Just to add evidence: I am experiencing exactly the same issue as ToadKing, same kernel 5.8.1-arch1-1, and latest decklink 11.6-1.

It looks like the user nettings over at the blackmagic forums figured out a fix:

ToadKing commented on 2020-08-15 04:43 (UTC)

Looks like the new 5.8 kernel broke the DKMS install.

KMS make.log for blackmagic-io-11.6a26 for kernel 5.8.1-arch1-1 (x86_64)
Fri 14 Aug 2020 11:41:37 PM CDT
make -C /lib/modules/5.8.1-arch1-1/build M=/var/lib/dkms/blackmagic-io/11.6a26/build
make[1]: Entering directory '/usr/lib/modules/5.8.1-arch1-1/build'
AR      /var/lib/dkms/blackmagic-io/11.6a26/build/built-in.a
CC [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/bmio_client.o
CC [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/bmio_dev.o
CC [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/bmio_device.o
CC [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/bmio_driver.o
CC [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/bmio_export.o
CC [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/bmio_serial.o
SHIPPED /var/lib/dkms/blackmagic-io/11.6a26/build/blackmagic.o
CC [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/bmio_pci_ids.o
CC [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/bm_locks.o
CC [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/bm_mm.o
CC [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/bm_pci.o
CC [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/bm_util.o
CC [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/bmio_audio.o
/var/lib/dkms/blackmagic-io/11.6a26/build/bmio_serial.c: In function ‘bmio_serial_remove’:
/var/lib/dkms/blackmagic-io/11.6a26/build/bmio_serial.c:267:2: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement]
267 |  struct tty_struct *tty = tty_port_tty_get(&serial->port);
    |  ^~~~~~
LD [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/blackmagic-io.o
LD [M]  /var/lib/dkms/blackmagic-io/11.6a26/build/snd_blackmagic-io.o
MODPOST /var/lib/dkms/blackmagic-io/11.6a26/build/Module.symvers
/var/lib/dkms/blackmagic-io/11.6a26/build/.blackmagic.o.cmd: No such file or directory
make[2]: *** [scripts/Makefile.modpost:111: /var/lib/dkms/blackmagic-io/11.6a26/build/Module.symvers] Error 1
make[1]: *** [Makefile:1669: modules] Error 2
make[1]: Leaving directory '/usr/lib/modules/5.8.1-arch1-1/build'
make: *** [Makefile:46: all] Error 2

goldensuneur commented on 2020-08-10 17:44 (UTC)

I get the same error on my intensity pro. However obs and MediaExpress are both working fine on my machine. So I suppose this is a bug in gstreamer.

ToadKing commented on 2020-08-10 16:30 (UTC)

Don't know if this is a bug with decklink itself or the AUR but after upgrading from 11.5 to 11.6 recording no longer works for my Intensity Pro 4k.

Output from gst-launch-1.0 decklinkvideosrc:

Setting pipeline to PAUSED ...
Pipeline is live and does not need PREROLL ...
Setting pipeline to PLAYING ...
New clock: GstSystemClock
ERROR: from element /GstPipeline:pipeline0/GstDecklinkVideoSrc:decklinkvideosrc0: Internal data stream error.
Additional debug info:
../gstreamer/libs/gst/base/gstbasesrc.c(3072): gst_base_src_loop (): /GstPipeline:pipeline0/GstDecklinkVideoSrc:decklinkvideosrc0:
streaming stopped, reason not-negotiated (-4)
Execution ended after 0:00:00.000274115
Setting pipeline to PAUSED ...
Setting pipeline to READY ...
Setting pipeline to NULL ...
Freeing pipeline ...

(gst-launch-1.0:3622): GStreamer-CRITICAL **: 11:27:44.290: 
Trying to dispose element pipeline0, but it is in PLAYING instead of the NULL state.
You need to explicitly set elements to the NULL state before
dropping the final reference, to allow them to clean up.
This problem may also be caused by a refcounting bug in the
application or some element.

arch_fram commented on 2020-07-02 14:43 (UTC) (edited on 2020-07-02 14:43 (UTC) by arch_fram)

I resolved the problem and I got which my card worked again. First at all I have to stop the DesktopVideoHelper service and rebuilt the package, then, I did run the service and used the DesktopVideoUpdater to update the driver, after the reboot the card was able to work again. Maybe is a issue with my board.

zangoku commented on 2020-06-24 18:38 (UTC)

I just rebuilt the package and the dkms installed.

arch_fram commented on 2020-06-21 17:06 (UTC)

@goldensuneur ● DesktopVideoHelper.service - DesktopVideo helper daemon Loaded: loaded (/usr/lib/systemd/system/DesktopVideoHelper.service; enabled; vendor > Active: active (running) since Sun 2020-06-21 18:56:01 CEST; 6min ago Main PID: 483 (DesktopVideoHel) Tasks: 3 (limit: 9438) Memory: 1.9M CGroup: /system.slice/DesktopVideoHelper.service └─483 /usr/lib/blackmagic/DesktopVideo/DesktopVideoHelper -n Yes and still is not detected by the driver.

goldensuneur commented on 2020-06-21 09:23 (UTC)

@arch_fram is the DesktopVideoHelper.service running ? Since 11.5 the service need run otherwise the card will not work.

arch_fram commented on 2020-06-20 07:51 (UTC) (edited on 2020-06-20 08:19 (UTC) by arch_fram)

I have 5.7.4-arch1-1 kernel, and It seem like my Decklink mini recorder is not detected :/ ERR The driver was not able to be unloaded because it is in use ERR You will need to reboot to use the new driver

I'll reboot the system but nothing happen, is still not working

daren commented on 2020-06-18 09:37 (UTC)

Works for me as well.

@zangoku IIRC one time it didn't build for me years back was when there were leftovers of the dkms stuff in /var/lib/dkms/blackmagic* that was set up for a previous gcc version and had to be removed. Maybe it helps.

goldensuneur commented on 2020-06-18 08:46 (UTC)

Decklink 11.5.1 builds and works fine on my computer with kernels 5.7.x and the patches I added in the latest commit.

zangoku commented on 2020-06-17 21:18 (UTC)

A little heads up for those on kernel 5.7, dkms modules don't build until blackmagic releases an update as usual.

Andy2 commented on 2020-06-16 16:43 (UTC)

Anybody else having issues with 100% CPU load on the DesktopVideoHelper and the DesktopVideoUpdater executables? I didn't find out yet what exactly triggers it, but I can see both of them spinning at 100% load sooner or later on my machine until I stop the systemd unit and kill the processes.

Houtworm commented on 2020-06-15 19:10 (UTC)

@filmtools, Install linux-headers with the following command and try again

sudo pacman -S linux-headers

filmtools commented on 2020-06-15 18:16 (UTC)

I get this error on Manjaro

(3/6) Install DKMS modules ==> Unable to install module blackmagic/11.5.1a14 for kernel 5.6.16-1-MANJARO: Missing kernel headers. ==> Unable to install module blackmagic-io/11.5.1a14 for kernel 5.6.16-1-MANJARO: Missing kernel headers. (4/6) Updating icon theme caches... (5/6) Fixing hardcoded icons... (6/6) Updating the desktop file MIME type cache...

TexZeTech commented on 2020-05-06 17:57 (UTC)

Seems to be working on my end as well.

bigtimmyc commented on 2020-05-05 13:58 (UTC)

Your update is working perfectly for me!

Polarity commented on 2020-05-05 09:51 (UTC)

That was quick! Thank you! I just built it with 5.6.10 without issues. After a brief look at Resolve and Nuke with Decklink Studio 4K it seems to be working as expected.

goldensuneur commented on 2020-05-05 09:33 (UTC)

Version 11.5.1 has just been released unfortunately it does not build with kernels 5.6.0+ so I patched the driver to make it compatible.

However I'm not familiar with kernel module development so consider support for newer kernel experimental. Changes are minimal so hopefully I did not introduce any bug.

goldensuneur commented on 2020-05-04 08:25 (UTC)

Yes, you can clone the git repository (, go to the commit corresponding to the version you want and build the package with makepkg -i.

TexZeTech commented on 2020-05-03 15:42 (UTC) (edited on 2020-05-04 00:05 (UTC) by TexZeTech)

Is there a way to build the previous version? Sorry still new to building files I'm also running manjaro.


goldensuneur commented on 2020-05-02 14:38 (UTC)


No update yet and support could not give me estimate release date last time I asked. You can ask them directly, maybe it will make the issue more visible.

bigtimmyc commented on 2020-05-02 13:15 (UTC)


I'm guessing still no luck with the BM update yet?

Here's my failed install log for dkms for kernel 5.6.8

DKMS make.log for blackmagic-io-11.5a34 for kernel 5.6.8-arch1-1 (x86_64) Sat 02 May 2020 23:07:25 AEST make -C /lib/modules/5.6.8-arch1-1/build M=/var/lib/dkms/blackmagic-io/11.5a34/build make[1]: Entering directory '/usr/lib/modules/5.6.8-arch1-1/build' AR /var/lib/dkms/blackmagic-io/11.5a34/build/built-in.a CC [M] /var/lib/dkms/blackmagic-io/11.5a34/build/bmio_device.o CC [M] /var/lib/dkms/blackmagic-io/11.5a34/build/bmio_dev.o CC [M] /var/lib/dkms/blackmagic-io/11.5a34/build/bmio_client.o CC [M] /var/lib/dkms/blackmagic-io/11.5a34/build/bmio_driver.o CC [M] /var/lib/dkms/blackmagic-io/11.5a34/build/bmio_export.o CC [M] /var/lib/dkms/blackmagic-io/11.5a34/build/bmio_serial.o CC [M] /var/lib/dkms/blackmagic-io/11.5a34/build/bmio_pci_ids.o CC [M] /var/lib/dkms/blackmagic-io/11.5a34/build/bm_locks.o SHIPPED /var/lib/dkms/blackmagic-io/11.5a34/build/blackmagic.o CC [M] /var/lib/dkms/blackmagic-io/11.5a34/build/bm_mm.o CC [M] /var/lib/dkms/blackmagic-io/11.5a34/build/bm_pci.o CC [M] /var/lib/dkms/blackmagic-io/11.5a34/build/bm_util.o CC [M] /var/lib/dkms/blackmagic-io/11.5a34/build/bmio_audio.o /var/lib/dkms/blackmagic-io/11.5a34/build/bm_util.c: In function ‘bm_uptime’: /var/lib/dkms/blackmagic-io/11.5a34/build/bm_util.c:920:18: error: storage size of ‘t’ isn’t known 920 | struct timespec t; | ^ /var/lib/dkms/blackmagic-io/11.5a34/build/bm_util.c:921:2: error: implicit declaration of function ‘getrawmonotonic’ [-Werror=implicit-function-declaration] 921 | getrawmonotonic(&t); | ^~~~~~~~~~~~~~~ /var/lib/dkms/blackmagic-io/11.5a34/build/bm_util.c:920:18: warning: unused variable ‘t’ [-Wunused-variable] 920 | struct timespec t; | ^ /var/lib/dkms/blackmagic-io/11.5a34/build/bm_util.c:926:1: warning: control reaches end of non-void function [-Wreturn-type] 926 | } | ^ cc1: some warnings being treated as errors make[2]: [scripts/ /var/lib/dkms/blackmagic-io/11.5a34/build/bm_util.o] Error 1 make[2]: Waiting for unfinished jobs.... /var/lib/dkms/blackmagic-io/11.5a34/build/bmio_serial.c: In function ‘bmio_serial_remove’: /var/lib/dkms/blackmagic-io/11.5a34/build/bmio_serial.c:267:2: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement] 267 | struct tty_struct *tty = tty_port_tty_get(&serial->port); | ^~~~~~ make[1]: [Makefile:1683: /var/lib/dkms/blackmagic-io/11.5a34/build] Error 2 make[1]: Leaving directory '/usr/lib/modules/5.6.8-arch1-1/build' make: [Makefile:46: all] Error 2

goldensuneur commented on 2020-04-19 10:27 (UTC)

Yes, and it won't work until Blackmagic releases an update. I'll open another support ticket asking if they know when the next version will be released.

zangoku commented on 2020-04-19 00:32 (UTC)

DKMS modules do not build on 5.6.5 either.

goldensuneur commented on 2020-04-06 19:29 (UTC) (edited on 2020-04-19 10:15 (UTC) by goldensuneur)

DKMS modules do not build with linux 5.6+, stay on lts for now. Blackmagic support is aware of the issue we have to wait until they release an update.

guitaristtom commented on 2020-03-19 22:55 (UTC)

@goldensuneur Weird. It wasn't showing up until I reinstalled, rebooted, and reinstalled again.

I always get stuck with weird issues like that.

But I am still having issues with audio on this new driver, but I'm guessing that it's a Blackmagic issue with their driver, since it's also happening on my Windows install.

goldensuneur commented on 2020-03-18 20:13 (UTC) (edited on 2020-03-19 07:44 (UTC) by goldensuneur)

You have it if you installed the decklink package. It's located under /usr/lib/systemd/system/DesktopVideoHelper.service.

But this is a .service file for systemd, you have to start/enable the service.

sudo systemctl start DesktopVideoHelper.service
sudo systemctl enable DesktopVideoHelper.service

guitaristtom commented on 2020-03-18 19:49 (UTC)

Since updating to 11.5 my capture card isn't showing up in either the Blackmagic Video Setup Utility or OBS.

I don't seem to have DesktopVideoHelper.service on my system, and I've even tried to reinstall this package just in case I was missing something.

troyBORG commented on 2020-03-04 17:51 (UTC)

@goldensuneur Yep that fixed it, as soon as I picked my 5.4 headers, it installed and added the dkms module.

goldensuneur commented on 2020-03-04 07:10 (UTC) (edited on 2020-03-04 07:10 (UTC) by goldensuneur)

@troyBORG You must install headers for your kernel otherwise dkms cannot build the kernel module. I don't know the mane of the package for Manjaro but on Arch it would be linux-headers or linux-lts-headers.

troyBORG commented on 2020-03-04 06:37 (UTC) (edited on 2020-03-04 06:37 (UTC) by troyBORG)

Is this anything I need to worry about?

(3/6) Install DKMS modules
==> Unable to install module blackmagic-io/11.5a34 for kernel 5.4.23-1-MANJARO: Missing kernel headers.
==> Unable to install module blackmagic/11.5a34 for kernel 5.4.23-1-MANJARO: Missing kernel headers.

goldensuneur commented on 2020-03-02 19:00 (UTC)

I finally had time to investigate the issue. It looks like that as of 11.5, the DesktopVideoHelper.service has to be running for the capture card to work properly.

goldensuneur commented on 2020-02-20 15:53 (UTC)

You can use the PKGBUILD in this commit (, it will build the package for version 1.14.1.

About the error you got earlier, it looks like Blackmagicdesign updated the upstream file without notice. I tried this new file but it does not solve our problem of undetected card.

Sorry I don't have much time to investigate now.

troyBORG commented on 2020-02-20 15:02 (UTC) (edited on 2020-02-20 15:06 (UTC) by troyBORG)

How do I get the older version? I'm planning on getting the card soon. Or should I wait for new version before I order the card?

goldensuneur commented on 2020-02-19 07:00 (UTC) (edited on 2020-02-20 15:53 (UTC) by goldensuneur)

You are right. Mine also shows on lspci with the correct driver but nothing in obs or in the setup.

I'm currently away from my computer so I cannot easily do proper that involve display. I'll do my best but for the moment I recommend not to update and stay on the 1.14.1 version (

spark9ng commented on 2020-02-19 04:51 (UTC)

Did 11.5 break for anyone else? Intensity Pro 4k not showing up in the setup anymore ("No Desktop Video Device Detected"). Device still shows up in lspci and driver still shows up in lsmod. Tested with linux and linux-lts kernels.

goldensuneur commented on 2019-12-26 10:44 (UTC)

I changed the PKGBUILD to split the package into two. decklink containing the driver and utilities (firmware update and configuration). And mediaexpress that contains the MediaExpress software.

Raketenjoint commented on 2019-12-16 21:46 (UTC)

Please add the missing dependency: dkms

goldensuneur commented on 2019-08-27 14:37 (UTC)

Yes, that is a good idea. However I don't have time right now to split the package.

Do not expect an update soon but it's definitely on my todo list.

Ashark commented on 2019-08-26 07:22 (UTC)

Maybe it is a good idea to split this package to several packages, like its done in debian variant?

goldensuneur commented on 2019-08-21 12:24 (UTC)

I forgot to regen the .SRCINFO file last time so the source refers to a previous version. However, this file only contains metadata for the AUR website and helpers so I think it's ok.

Ashark commented on 2019-08-21 11:49 (UTC)

Is it ok that current version is 11.3-1, but filename in sources is decklink-11.2.tar.gz?

inspector77 commented on 2019-08-20 09:20 (UTC)

Thanks for this package, working perfect with Decklink mini monitor and Davinci resolve studio 16

goldensuneur commented on 2019-05-29 12:36 (UTC)

You are right, I removed the dependency on Qt4 and added qt5.

I also noticed with ldd that all GUIs link to system Qt5 instead of the provided one. So let me know if anything breaks on your side.

daurnimator commented on 2019-05-25 02:10 (UTC)

The GUIs seem to work with Qt5? At least, I see e.g. /usr/lib/blackmagic/DesktopVideo/libQt5* in my local package.

goldensuneur commented on 2019-04-01 11:59 (UTC)

It looks like they fixed the problems with kernel 5.0. It is know building and working normally.

Polarity commented on 2019-04-01 11:08 (UTC)

Looks like the next release is out.

Serus commented on 2019-03-15 20:06 (UTC)

I got an email back from BM stating they're aware of the problem and will be addressed in the next release.

guitaristtom commented on 2019-03-13 03:26 (UTC)

Well at least I'm not crazy about it not working with Kernel 5.0.0

I guess I'll have to wait to use my capture card again sigh. Hopefully, BM fixes it on their end soon.

lukas2511 commented on 2019-03-06 23:19 (UTC)

Mh. It seems that the sound module actually might be negatively affected by leaving the FPU disabled. I'm not sure. I'll use my patch for now as it's working for me and I don't need sound, but I'm hoping to see a new decklink release soon, also adressing the last build issue...

lukas2511 commented on 2019-03-06 23:05 (UTC)

@Serus Well, we can also just leave the FPU disabled. That might result in some slowdown, but would be completely legal. See the updated patch

I'm not sure if this might have some other unforeseen side-effects, but at least in my tests I don't really see any difference. Maybe some scaling setups or the alpha blends on the decklink duo would be a bit more problematic? No idea. Also hoping for a vendor-side fix.

Serus commented on 2019-03-06 22:41 (UTC)

@lukas2511 I did something similar you did.

But the license rewrite is illegal and not something we can put up here on the AUR.

I also confirmed it works with my Intensity Pro 4K, I also sent an email to BlackMagic explaining the situation, hoping they'll come up with a fix soon.

At this point in time the approved workaround is my earlier mentioned workaround.

lukas2511 commented on 2019-03-06 22:17 (UTC)

Got my other modules fixed. Decklink works fine with 5.0, the only real issue are those GPL-only symbols.

lukas2511 commented on 2019-03-06 21:42 (UTC)

Well... actually... maybe somebody could test this patch:

The affected symbols weren't mentioned in the pre-built module, I took the wrong name, it actually calls a function in blackmagic_lib.c, that way we actually can wrap every affected call.

The patch is done directly against /usr/src directory, didn't prepare anything for the PKGBUILD as I don't like rewriting the modules license for a release.

Other than loading the modules I can't test it myself as my decklink system has a few other modules with the same issue and I can't easily boot into 5.0 because of this.

For a release I'd suggest waiting for ZFSonLinux to write their workaround for the issue, maybe it's reusable for the decklink module and that way we wouldn't need to rewrite the license.

lukas2511 commented on 2019-03-06 20:33 (UTC)

Unfortunately the affected symbols are also used in the pre-compiled blackmagic library so we can't cleanly replace the affected code...

I have an idea on mind on how to fix this, but I'm hoping for a fix from Blackmagic as my fix would be very very ugly involving a binary patch to the shipped library... :-/

Serus commented on 2019-03-06 18:26 (UTC) (edited on 2019-03-06 18:27 (UTC) by Serus)

As of kernel 5.0 this requires a GPL only symbol.

Either downgrade your kernel to 4.20, or install linux-lts and boot into that when you need to use your capture card.

ggrub commented on 2019-02-24 10:47 (UTC)

@lukas2511 Thank you very much.

guitaristtom commented on 2019-02-21 20:14 (UTC)

@lukas2511 Thank you so much!! It's working for me again!

goldensuneur commented on 2019-02-21 10:34 (UTC) (edited on 2019-02-21 10:35 (UTC) by goldensuneur)

@lukas2511 I tested your fix, it works fine on my computer. Thank you for the DLAGENT.

I modified it a bit:

  • removed the -s flag to see progress.

  • added a small hack to hide the html sources shown by curl.

lukas2511 commented on 2019-02-21 00:16 (UTC)

I'm not entirely sure what happened, but it seems that the kernel makefiles somehow started ignoring the .a file...

I think the .a file is actually just a .o file with a wrong filename, it's definitively not an archive, it's just an object file. Maybe that's why the kernel is ignoring it... or it may be a bug... I think there should be an error about a missing library or something, it shouldn't just be silently ignored...

Well anyway...

Manually linking or renaming the .a files to .o_shipped and modifying the Makefile allows building working kernel modules tested on 4.20.10.

You can find my modified PKGBUILD here:

Next to the actual build-fix (the sed and two mv in the package step) it also optimizes the download-behavior by replacing the whole prepare step with a DLAGENT, allowing the use of all makepkg caching and validation methods.

guitaristtom commented on 2019-02-14 18:51 (UTC)

Some of the other packages I have installed refuse to let me switch to the LTS kernel... so I really hope that Blackmagic deals with this issue and fixes it.

ggrub commented on 2019-01-07 18:10 (UTC) (edited on 2019-01-07 22:46 (UTC) by ggrub)

Damn it. My asus Z390 board works only with kernel 4.20. My intensity pro 4k seems to work only with kernel 4.19. Such a crap.

fthiery commented on 2019-01-07 10:37 (UTC)

Done, merci :)

goldensuneur commented on 2019-01-07 10:31 (UTC)

@fthiery sure, I'm not an expert but if updates stay as simple as this one I can help

fthiery commented on 2019-01-07 10:29 (UTC)

i updated the AUR to use your own PKGBUILD, thanks !

fthiery commented on 2019-01-07 10:23 (UTC)

@goldensuneur would you like to co-maintain this AUR ?

goldensuneur commented on 2019-01-03 18:06 (UTC) (edited on 2019-01-03 18:07 (UTC) by goldensuneur)

I updated the PKGBUILD for the latest release (10.11.4) but it doesn't work any better with 4.20. For now just stick with the LTS Kernel.

Here is the updated PKGBUILD I used to get the 10.11.4:

There is no need for the decklink.install file anymore since DKMS handles the building and loading of the module.

laemplein commented on 2019-01-03 07:56 (UTC)

Something broke with Kernel 4.20 and DKMS, it doesn't work after the update even with lots of 'dkms autoinstall'. I just installed the LTS Kernel (4.19.13) -> DKMS and my card are up and running again.

OniLink commented on 2019-01-02 04:29 (UTC)

This is super out of date. Any plans of getting it caught up with Decklink 10.11?

fthiery commented on 2018-07-13 13:38 (UTC)

BTW, is the install script still necessary? DKMS has an hook to rebuild modules automatically on install/upgrade...

I think it does not work properly. If you manage to make it work i'll happily let you comaintain

Marte2007 commented on 2018-06-10 22:28 (UTC)

I solved the stdarg.h problem by removing /usr/src/blackmagic-*/ then rebuilding the module with sudo dkms autoinstall.

Apparently the old GCC's include path was still in the cache.

BTW, is the install script still necessary? DKMS has an hook to rebuild modules automatically on install/upgrade...

ggrub commented on 2018-06-09 18:19 (UTC)

Thank you very much for this package. Installed with pamac. Works very well. Only the settings in "desktop video setup" have to be reconfigured after each reboot.

daren commented on 2018-06-07 07:10 (UTC) (edited on 2018-06-09 21:53 (UTC) by daren)

It fails to compile the DKMS module with GCC 8.1.1+20180531 and/or kernel 4.16.13-2, not sure which one is the culprit.

Apparently it can't find the stdarg.h header file, did something get botched with the GCC update?

Edit: It still works with GCC 8.1.0

Edit2: Nevermind, removing and reinstalling decklink fixed the issue somehow.

fthiery commented on 2018-05-28 12:57 (UTC)

Updated to 10.10, it did compile for 4.16 (and device gets listed)

Marte2007 commented on 2018-05-21 07:33 (UTC) (edited on 2018-05-21 07:33 (UTC) by Marte2007)

I had the same problem, but I just had to rebuild the module to make it work again.

DKMS had built the module for the wrong kernel version.

I am currently on kernel 4.16.9-1-ARCH.

Houtworm commented on 2018-05-10 13:57 (UTC)

I rebooted my pc yesterday, and sadly my decklink device is once again not found.

It was working great on the normal kernel for quite a long time.

I also see it is flagged out of date and there is indeed a newer version.

Sure hope this update fixes the problem,

for now i will try the LTS kernel, i will post back here if that does not work.


Kiniamaro commented on 2018-03-21 02:18 (UTC) (edited on 2018-03-21 02:19 (UTC) by Kiniamaro)

Just grabbed this, works fine for me, I'm using a PCI Intensity Pro card.

did a little

$ modprobe blackmagic

booted up OBS and the card showed up, plugged in my switch, everything worked, no input delay.

Thanks for this package!

ps: I'm on kernel 4.14.15, hopefully it doesn't stop working after I update, You'll hear from me if it stops working.

Kakashi-sensei commented on 2018-02-12 09:34 (UTC)

This is pretty funny to watch, tbh.

fthiery commented on 2018-02-12 09:32 (UTC)

Okay, so right now linux-lts (which just switched to 4.14) is the only working version.

OniLink commented on 2018-02-09 01:07 (UTC) (edited on 2018-02-09 02:48 (UTC) by OniLink)

TT;DR (too technical, didn't read): Linux 4.15 doesn't work with Blackmagic anymore because it removed some ancient code that Blackmagic still uses.


Uninstalled my previous hacky driver install as the Blackmagic was no longer recognized after a reboot. Tried to reinstall with the current version and ran into an error - as of 4.15.1, an old timer interface is removed.

If Blackmagic wants to get their code working on new Linux kernels, they need to replace the end of bm_timer_alloc at bm_util.c line 420 with

timer_setup(timer, (void(*)(bm_timer_t*))callback, 0);

return timer;

instead of the init_timer line and followup. Unfortunately this will break any code using the driver as it breaks expected behavior (data becomes unused, the callback gets sent a pointer to the timer instead), but that's on them. Trying to patch it like this and install leads to a crash as soon as the module blackmagic-io is inserted into the kernel, telling me I missed something probably. I couldn't find any use of this function in the driver code, though. It's likely a problem in Blackmagic's other software expecting certain behavior from the driver which necessarily changes due to this fix.

fthiery commented on 2018-02-01 12:09 (UTC)

I tried to patch dkms.conf to fix the path, but it's still not being called. It looks to be doing depmod && modprobe anyway, which we are also doing.

We need a dkms guru to sed/patch it so that dkms install and dkms remove do work. Currently i'm unsure whether the dkms stuff even has any effect at all (and the install script is doing everything by hand).

If anyone is interested, the files are in /usr/src/blackmagic-10.9.10a2/

I reworked the install script to cleanup better (e.g. remove /usr/src/blackmagic-*); please feedback it the thing also works for you. Seems to work for me. But this package is looking more and more like frankenstein.

fthiery commented on 2018-02-01 11:43 (UTC)

Apparently the dmks script is trying to load the modules

POST_INSTALL="../../lib/blackmagic/blackmagic-loader $PACKAGE_NAME $PACKAGE_VERSION"

The path is wrong, because it's here: /usr/lib/blackmagic/DesktopVideo/blackmagic-loader

So i'm guessing it's not just not executable, but not found at all. We are loading the modules manually anyway and it seems to be working...

Kakashi-sensei commented on 2018-02-01 11:14 (UTC)

To be honest, I haven't had the chance to test it personally. I trust your confirmation.

fthiery commented on 2018-02-01 11:10 (UTC)

I also see the warning, but it does seem to work anyway "Warning: The post_install script is not executable."

@Kakashi-sensei doesn't the board work at all ?

Kakashi-sensei commented on 2018-01-30 08:56 (UTC)

I'm getting the following error after each install of a dkms module: Warning: The post_install script is not executable.

fthiery commented on 2018-01-29 17:07 (UTC)

@OniLink thanks for testing, i updated the package and it does compile and insert against 4.14.15-1-ARCH.

Kakashi-sensei commented on 2018-01-28 21:31 (UTC)

Well, at least the package maintainer knows what to do?

OniLink commented on 2018-01-28 21:22 (UTC) (edited on 2018-01-28 21:37 (UTC) by OniLink)

@storrgie I got it built, installed, and functioning just before I reported that it works. Kernel version is 4.14.15.

I couldn't find a permanent download link on Blackmagic's website that could be inserted into the PKGBUILD for download so instead I had to do something a little hacky. It took a while to get it working but I think I can boil it down to the steps that worked iirc. I downloaded 10.9.10's tar.gz from Blackmagic's website and renamed it to "decklink-10.9.10.tar.gz". I also updated the pkgver and sha256 in the PKGBUILD. At this point my method is... probably not recommended by any means, but it worked for me.

Comment out these lines of the PKGBUILD (42-43) as the newer downloads renamed /usr/sbin to /usr/bin so it's not necessary anymore:

mv $pkgdir/usr/sbin/* $pkgdir/usr/bin

rm -rf $pkgdir/usr/sbin

I ran makepkg once to generate the package source directory (./src), then I removed everything from that source directory and copied in decklink-10.9.10.tar.gz. I then ran makepkg again which created the proper package for installation.

Apologies if this method infuriates anybody for being improper or incorrect. If I could've found a permanent download link I could've updated the PKGBUILD's pkgsrc_url to that instead in addition to updating the package version and sha256 and the PKGBUILD would've worked just fine. Instead I had to do a manual replacement of files mid-process. :/

The sha256 for 10.9.10 is "1b66bb89aaef77dbfdd6913eb7067f0511866070ff8916a5e3af1d68e3fcfe84"

storrgie commented on 2018-01-28 15:51 (UTC) (edited on 2018-01-28 15:52 (UTC) by storrgie)

@OniLink, after trying to use this PKGBUILD with 10.9.10 on linux and linux-lts I'm still unable to install properly. Would you mind sharing your approach?

storrgie commented on 2018-01-27 00:56 (UTC)

@OniLink positive it builds on modern kernels?

OniLink commented on 2018-01-26 23:55 (UTC)

Upstream is fixed in 10.9.10.

gibbz commented on 2017-12-20 03:48 (UTC)

Any idea when this will work on 4.14 LTS?

fthiery commented on 2017-11-08 10:20 (UTC)

Well i could include the patch, but then it would break building on lts. The proper fix would be that the blackmagic upstream code would use the proper function name depending on the target kernel. @Houtworm just rgrep for wait_queue_t and replace by wait_queue_entry_t ; a recursive sed might also work

Houtworm commented on 2017-11-05 09:19 (UTC)

@rskennymore Very nice, You are a smart man :) Any chance you could share these edited .c files? I am struggling to get it working on arch, I am currently using an Ubuntu Dual boot just for the intensity 4K. Also do these edits make it possible to use it on the latest kernel? or do you use the LTS kernel?

rskennymore commented on 2017-11-04 14:45 (UTC)

It's only broken because of this: I hand-jammed in a fix by just manually editing blackmagic_gate.c and bm_util.c in the source tarball and rebuilding. Seems to work fine Changing some names around in the kernel... I emailed support through the upstream link

fthiery commented on 2017-09-22 14:35 (UTC)

Should i make linux-lts and linux-lts-headers a hard dependency ?

fthiery commented on 2017-09-22 14:33 (UTC)

Btw LTS is 4.9.50-1-lts

fthiery commented on 2017-09-22 14:33 (UTC)

As specified, this driver only works on the lts kernel. Other flavors might work, but if not you may report it to

gibbz commented on 2017-09-22 13:20 (UTC)

Latest version doesn't seem to work. Im getting the unable to detect device in decklink. The card shows fine under lspci. And firmware says its up to date. Does this require a newer kernel? Im on 4.12

LIONHEART commented on 2017-06-23 08:31 (UTC)

@fthiery Thanks, got video capture working. Still working on audio.

fthiery commented on 2017-06-14 07:28 (UTC)

@LIONHEAR be sure to flash the board with Blackmagic Firmware Updater and reboot; had the same issue too, but after a reboot the device is detected in Blackmagic Desktop Video setup and gst-launch works again.

LIONHEART commented on 2017-06-13 19:39 (UTC)

@fthiery No I don't. It says No Desktop Video Device Detected.

fthiery commented on 2017-06-13 16:40 (UTC)

@Alexandrer-Prime my bad, it should work now @LIONHEART do you see the device in the blackmagic software ?

fthiery commented on 2017-06-13 15:46 (UTC)

@Alexander-Prime is the path to your build folder containing spaces by any chance ?

Alexander-Prime commented on 2017-06-13 12:42 (UTC)

The build script seems to be looking in the wrong place for the sources after downloading them. Changing the pkgver in PKGBUILD to "10.9.3a2" allows it to continue installing.

LIONHEART commented on 2017-06-12 01:06 (UTC)

Having trouble with the pipeline in gstreamer. Any tips getting this working would be greatly appreciated, $ gst-launch-1.0 decklinkvideosrc device-number=1 connection=1 mode=16 ! xvimagesink Setting pipeline to PAUSED ... ERROR: Pipeline doesn't want to pause. Setting pipeline to NULL ... Freeing pipeline ... $ gst-launch-1.0 decklinkvideosrc ! videoconvert ! xvimagesink Setting pipeline to PAUSED ... ERROR: Pipeline doesn't want to pause. Setting pipeline to NULL ... Freeing pipeline ...

Houtworm commented on 2017-06-06 11:06 (UTC) (edited on 2017-06-06 11:36 (UTC) by Houtworm)

Just installed 10.9.3-1 and for me it does not work on the latest kernel. It says no desktop video device connected (PCIE Device) I am going to install linux-lts linux-lts-headers and nvidia-lts remove nvidia and try again :) Edit: It also does not work on LTS Kernel any more, Software is installed and runs but i only get a black screen in media express and OBS I will spend the day puzzling hope i can find the problem, How is this AUR Package working for others?

daren commented on 2017-06-05 19:28 (UTC) (edited on 2017-06-05 19:44 (UTC) by daren)

Version 10.9.3a2 got released, got it working on the latest kernel.

fthiery commented on 2017-05-12 14:45 (UTC)

@Houtworm nicely spotted, i trusted Firefox's "copy as curl" (and i shouldnt have). Now the AUR should download back automatically (but not if the file is already downloaded). They have added user agent and content-type checking curl '' -H 'User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:53.0) Gecko/20100101 Firefox/53.0' -H 'Content-Type: application/json;charset=utf-8' --data '{"country":"us","platform":"Linux"}'

Houtworm commented on 2017-05-10 21:58 (UTC)

I am Using the LTS Kernel, If you want i could try the Non LTS version. I am upgrading to a m.2 ssd anyway :)

Alexander-Prime commented on 2017-05-10 09:44 (UTC)

@fthiery, your request JSON is malformed. Fixing it: {country":"us","platform":"Linux"} <- before {"country":"us","platform":"Linux"} <- after Seems to give the desired response.

fthiery commented on 2017-05-10 07:46 (UTC)

@Houtworm are you using the lts kernel or the latest one ?

Houtworm commented on 2017-05-09 23:47 (UTC)

Thanks fthiery, It works perfectly now. Capturing with my Blackmagic Intensity Pro 4K on Arch :)

fthiery commented on 2017-05-09 14:50 (UTC)

I replaced the broken download method by a static path (redefine pkgsrc_file after download, by default /tmp/Blackmagic_Desktop_Video_Linux_10.9.tar). Tried to fake the POST request to (supposed to return the actual download url) but it always ends up with a 504 GATEWAY_TIMEOUT error. curl '' -H 'Host:' -H 'User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:53.0) Gecko/20100101 Firefox/53.0' -H 'Accept: application/json, text/plain, */*' -H 'Accept-Language: en-US,en;q=0.5' --compressed -H 'Content-Type: application/json;charset=utf-8' -H 'Referer:' -H 'Content-Length: 35' -H 'DNT: 1' -H 'Connection: keep-alive' --data '{country":"us","platform":"Linux"}' -v The request is supposed to return a url like: If anyone succeeds in programmatically getting this url i'll happily integrate it instead.

fthiery commented on 2017-05-09 14:19 (UTC)

Indeed, apparently bmd has improved the download protection, working on it

Houtworm commented on 2017-05-05 03:55 (UTC)

Sadly it does not build, Failed to download. It gives an error unable to resolve host: invalid request to download. Also manually downloading the file and putting it in the makepkg folder does not work. It would be a nice add to the PKGBUILD to search for local files so that the curl download is not necessary whenever blackmagic desides to change urls. Thanks And Greets

fthiery commented on 2017-04-25 07:57 (UTC)

I know, said it below: you need linux-lts to have the driver working, not the latest linux package.

Joir2000 commented on 2017-04-23 09:34 (UTC)

Over here compilation still fails : DKMS make.log for blackmagic-io-10.8.6a2 for kernel 4.10.10-1-ARCH (x86_64) Sun Apr 23 11:26:12 CEST 2017 make -C /lib/modules/4.10.10-1-ARCH/build M=/var/lib/dkms/blackmagic-io/10.8.6a2/build make[1]: Entering directory '/usr/lib/modules/4.10.10-1-ARCH/build' LD /var/lib/dkms/blackmagic-io/10.8.6a2/build/built-in.o CC [M] /var/lib/dkms/blackmagic-io/10.8.6a2/build/bmio_client.o CC [M] /var/lib/dkms/blackmagic-io/10.8.6a2/build/bmio_dev.o CC [M] /var/lib/dkms/blackmagic-io/10.8.6a2/build/bmio_device.o CC [M] /var/lib/dkms/blackmagic-io/10.8.6a2/build/bmio_driver.o CC [M] /var/lib/dkms/blackmagic-io/10.8.6a2/build/bmio_serial.o CC [M] /var/lib/dkms/blackmagic-io/10.8.6a2/build/bmio_export.o CC [M] /var/lib/dkms/blackmagic-io/10.8.6a2/build/bmio_pci_ids.o CC [M] /var/lib/dkms/blackmagic-io/10.8.6a2/build/bm_locks.o CC [M] /var/lib/dkms/blackmagic-io/10.8.6a2/build/bm_mm.o CC [M] /var/lib/dkms/blackmagic-io/10.8.6a2/build/bm_pci.o /var/lib/dkms/blackmagic-io/10.8.6a2/build/bm_mm.c: In function ‘bm_get_user_pages’: /var/lib/dkms/blackmagic-io/10.8.6a2/build/bm_mm.c:110:10: error: too few arguments to function ‘get_user_pages_remote’ ret = get_user_pages_remote(task, task->mm, (unsigned long)address & PAGE_MASK, umem->length, write ? FOLL_WRITE : 0, umem->pages, NULL); ^~~~~~~~~~~~~~~~~~~~~ In file included from /var/lib/dkms/blackmagic-io/10.8.6a2/build/bm_mm.c:27:0: ./include/linux/mm.h:1267:6: note: declared here long get_user_pages_remote(struct task_struct *tsk, struct mm_struct *mm, ^~~~~~~~~~~~~~~~~~~~~ make[2]: *** [scripts/ /var/lib/dkms/blackmagic-io/10.8.6a2/build/bm_mm.o] Error 1 make[2]: *** Waiting for unfinished jobs.... make[1]: *** [Makefile:1490: _module_/var/lib/dkms/blackmagic-io/10.8.6a2/build] Error 2 make[1]: Leaving directory '/usr/lib/modules/4.10.10-1-ARCH/build' make: *** [Makefile:47: all] Error 2

fthiery commented on 2017-04-13 16:59 (UTC)

I confirm that it finally works ! gst-launch-1.0 decklinkvideosrc device-number=1 connection=1 mode=16 ! xvimagesink

fthiery commented on 2017-04-13 16:57 (UTC)

After a reboot, the device firmware updater reports the device to be up to date.

fthiery commented on 2017-04-13 16:55 (UTC)

So it still fails on the current linux, but does compile on linux-lts (4.9.22-1-lts). So you need the linux-lts and linux-lts-headers packages. Device flashing did reportedly fail however, even if the dmesg doesn't report failure [ 89.800430] blackmagic: loading out-of-tree module taints kernel. [ 89.800434] blackmagic: module license 'Proprietary' taints kernel. [ 89.800434] Disabling lock debugging due to kernel taint [ 89.802668] blackmagic: Loading driver (version: 10.8.6a2) [ 93.169164] BlackmagicIO: Driver version 10.8.6a2 loading [ 93.169746] BlackmagicIO: WARNING: The device "DeckLink Duo 2" has firmware that is older than the version shipped with the driver (device: 0x14 driver: 0xcc) [ 93.169753] BlackmagicIO: Enabled device "DeckLink Duo 2 (1)" x4/5 Gbps (38254df0,510100) FW Date: 03-15 08:29 [ 93.170277] BlackmagicIO: DeckLink Duo 2 (1) as blackmagic!io0 [0000:01:00.0] [ 93.170278] Probe subdriver ffffffffa0a57320 (serial): ffff8804031d9b40 [ 93.170279] BlackmagicIO: Initialised serial driver for blackmagic!io0 [ 93.170352] BlackmagicIO: WARNING: The device "DeckLink Duo 2" has firmware that is older than the version shipped with the driver (device: 0x14 driver: 0xcc) [ 93.170357] BlackmagicIO: Enabled device "DeckLink Duo 2 (3)" x4/5 Gbps (38254df2,510101) FW Date: 03-15 08:29 [ 93.170893] BlackmagicIO: DeckLink Duo 2 (3) as blackmagic!io1 [0000:01:00.0] [ 93.170894] Probe subdriver ffffffffa0a57320 (serial): ffff8803b57843c0 [ 93.170895] BlackmagicIO: Initialised serial driver for blackmagic!io1 [ 93.170984] BlackmagicIO: WARNING: The device "DeckLink Duo 2" has firmware that is older than the version shipped with the driver (device: 0x14 driver: 0xcc) [ 93.170990] BlackmagicIO: Enabled device "DeckLink Duo 2 (2)" x4/5 Gbps (38254df1,510102) FW Date: 03-15 08:29 [ 93.171565] BlackmagicIO: DeckLink Duo 2 (2) as blackmagic!io2 [0000:01:00.0] [ 93.171566] Probe subdriver ffffffffa0a57320 (serial): ffff88037c6279c0 [ 93.171566] BlackmagicIO: Initialised serial driver for blackmagic!io2 [ 93.171657] BlackmagicIO: WARNING: The device "DeckLink Duo 2" has firmware that is older than the version shipped with the driver (device: 0x14 driver: 0xcc) [ 93.171663] BlackmagicIO: Enabled device "DeckLink Duo 2 (4)" x4/5 Gbps (38254df3,510103) FW Date: 03-15 08:29 [ 93.172266] BlackmagicIO: DeckLink Duo 2 (4) as blackmagic!io3 [0000:01:00.0] [ 93.172267] Probe subdriver ffffffffa0a57320 (serial): ffff8803b5755000 [ 93.172267] BlackmagicIO: Initialised serial driver for blackmagic!io3 [ 97.703260] BlackmagicIO: Start programming firmware (SPI) [ 116.623785] device-mapper: uevent: version 1.0.3 [ 116.623843] device-mapper: ioctl: 4.35.0-ioctl (2016-06-23) initialised: [ 129.644274] BlackmagicIO: Completed programming firmware

fthiery commented on 2017-04-13 16:41 (UTC)

Got news: "I am pleased to advise that the recently released Desktop Video 10.8.6 includes the correction for the compilation issue on kernel 4.9+." Testing now

daren commented on 2017-03-03 20:21 (UTC)

Damn, they should hopefully fix it when Debian/Ubuntu relase the new versions with kernels >= 4.8.

fthiery commented on 2017-02-27 16:42 (UTC)

I'm afraid it's even worse, it doesn't compile against 4.9.11-1; note that the driver was not updated, just the desktop app. make -C /lib/modules/`uname -r`/build M=/usr/src/blackmagic-io-10.8.5a4 make[1]: Entering directory '/usr/lib/modules/4.9.11-1-ARCH/build' CC [M] /usr/src/blackmagic-io-10.8.5a4/bm_mm.o /usr/src/blackmagic-io-10.8.5a4/bm_mm.c: In function 'bm_get_user_pages': /usr/src/blackmagic-io-10.8.5a4/bm_mm.c:108:85: error: passing argument 5 of 'get_user_pages' from incompatible pointer type [-Werror=incompatible-pointer-types] ret = get_user_pages((unsigned long)address & PAGE_MASK, umem->length, write, 0, umem->pages, NULL); ^~~~ In file included from /usr/src/blackmagic-io-10.8.5a4/bm_mm.c:27:0: ./include/linux/mm.h:1280:6: note: expected 'struct vm_area_struct **' but argument is of type 'struct page **' long get_user_pages(unsigned long start, unsigned long nr_pages, ^~~~~~~~~~~~~~ /usr/src/blackmagic-io-10.8.5a4/bm_mm.c:108:10: error: too many arguments to function 'get_user_pages' ret = get_user_pages((unsigned long)address & PAGE_MASK, umem->length, write, 0, umem->pages, NULL); ^~~~~~~~~~~~~~ In file included from /usr/src/blackmagic-io-10.8.5a4/bm_mm.c:27:0: ./include/linux/mm.h:1280:6: note: declared here long get_user_pages(unsigned long start, unsigned long nr_pages, ^~~~~~~~~~~~~~ /usr/src/blackmagic-io-10.8.5a4/bm_mm.c:110:108: error: passing argument 7 of 'get_user_pages_remote' from incompatible pointer type [-Werror=incompatible-pointer-types] ret = get_user_pages_remote(task, task->mm, (unsigned long)address & PAGE_MASK, umem->length, write, 0, umem->pages, NULL); ^~~~ In file included from /usr/src/blackmagic-io-10.8.5a4/bm_mm.c:27:0: ./include/linux/mm.h:1276:6: note: expected 'struct vm_area_struct **' but argument is of type 'struct page **' long get_user_pages_remote(struct task_struct *tsk, struct mm_struct *mm, ^~~~~~~~~~~~~~~~~~~~~ /usr/src/blackmagic-io-10.8.5a4/bm_mm.c:110:10: error: too many arguments to function 'get_user_pages_remote' ret = get_user_pages_remote(task, task->mm, (unsigned long)address & PAGE_MASK, umem->length, write, 0, umem->pages, NULL); ^~~~~~~~~~~~~~~~~~~~~ In file included from /usr/src/blackmagic-io-10.8.5a4/bm_mm.c:27:0: ./include/linux/mm.h:1276:6: note: declared here long get_user_pages_remote(struct task_struct *tsk, struct mm_struct *mm, ^~~~~~~~~~~~~~~~~~~~~ cc1: some warnings being treated as errors make[2]: *** [scripts/ /usr/src/blackmagic-io-10.8.5a4/bm_mm.o] Error 1 make[1]: *** [Makefile:1490: _module_/usr/src/blackmagic-io-10.8.5a4] Error 2 make[1]: Leaving directory '/usr/lib/modules/4.9.11-1-ARCH/build' make: *** [Makefile:47: all] Error 2

daren commented on 2017-02-27 07:17 (UTC)

10.8.5 is out, I want to believe.

isopede commented on 2016-12-23 02:08 (UTC)

This is very likely a bug in blackmagic code. The 4.8 kernel introduced CONFIG_HARDENED_USERCOPY, which adds protections to the copying of data to/from userspace. The blackmagic kernel module is out-of-tree and is probably doing something wrong, which the protections are catching. I am not at all interested in fixing their kernel code, though this is possible if somebody has the gumption. You will have to wait until BM releases new modules, or downgrade, or perhaps try recompiling your kernel with CONFIG_HARDENED_USERCOPY unset.

fthiery commented on 2016-12-10 14:54 (UTC)


Excrement commented on 2016-12-10 14:51 (UTC)

Yea, that seems like a silly solution downgrade; im using a gtx1080 and im not sure if that solution will support my video card.. ill have to check.. incidentally is this your forum post?

fthiery commented on 2016-12-09 13:48 (UTC)

Yes, 4.4.36-1

Excrement commented on 2016-12-08 12:52 (UTC) (edited on 2016-12-10 14:52 (UTC) by Excrement)

Why do i have the sneaky suspicion they (blackmagic) dont really care and are just wanting you (the consumer) to buy their 4k card or something newer.... Using the LTS kernel means reverting from 4.8 to 4.4? edit; guess its happening with the 4k cards as well..

fthiery commented on 2016-12-07 22:23 (UTC)

I think that reverting to an LTS kernel is probably your best bet.

fthiery commented on 2016-12-07 22:22 (UTC)

I am in contact with them, but i don't get much feedback :( All i know is that they reproduced the issue, which is a start.

Excrement commented on 2016-12-07 21:46 (UTC)

checking in as well; module pukes (crashes) when loaded. any idea when this will be addressed as i use this app to do some video work.

fthiery commented on 2016-12-07 10:47 (UTC)

Tried, pushed, still crashing.

daren commented on 2016-12-06 09:32 (UTC) (edited on 2016-12-06 09:32 (UTC) by daren)

They released 10.8.3 and 10.8.4, had any luck with those?

fthiery commented on 2016-11-11 17:50 (UTC)

Good news, they reproduced the issue. There's hope ;)

fthiery commented on 2016-11-10 09:53 (UTC)

Got a response, they can't reproduce on Ubuntu 16.04 with v4.8.2 Mainline PPA Build. I'll keep you posted

shiggitay commented on 2016-10-23 12:18 (UTC)

DarenK, yeah. I also upgraded the kernel to 4.8. I'm just about to rollback to 4.7 if I can.

fthiery commented on 2016-10-21 11:14 (UTC)

I sent an email to, still no response

daren commented on 2016-10-20 20:42 (UTC)

Damn, I updated the kernel and didn't think the drivers would malfunction. Any updates?

fthiery commented on 2016-10-19 08:36 (UTC)

Apparently even the latest version crashes on 4.8.2-1-ARCH at boot oct. 19 10:32:18 kernel: usercopy: kernel memory overwrite attempt detected to ffff8803ccabfe10 (<process stack>) (16 bytes) oct. 19 10:32:18 kernel: ------------[ cut here ]------------ oct. 19 10:32:18 kernel: kernel BUG at mm/usercopy.c:75! oct. 19 10:32:18 kernel: invalid opcode: 0000 [#5] PREEMPT SMP oct. 19 10:32:18 kernel: Modules linked in: fuse nfsv3 nfs_acl rpcsec_gss_krb5 auth_rpcgss oid_registry nfsv4 dns_resolver nfs lockd grace sunrpc fscache snd_hda_codec_hdmi snd_hda_codec_re oct. 19 10:32:18 kernel: evdev tpm_tis acpi_pad mac_hid tpm_tis_core tpm sch_fq_codel vboxnetflt(O) vboxnetadp(O) pci_stub vboxpci(O) vboxdrv(O) ip_tables x_tables ext4 crc16 jbd2 fscrypto oct. 19 10:32:18 kernel: CPU: 3 PID: 958 Comm: BlackmagicFirmw Tainted: P D O 4.8.2-1-ARCH #1 oct. 19 10:32:18 kernel: Hardware name: ASUS All Series/H97M-PLUS, BIOS 2603 02/22/2016 oct. 19 10:32:18 kernel: task: ffff8803ccac2ac0 task.stack: ffff8803ccabc000 oct. 19 10:32:18 kernel: RIP: 0010:[<ffffffff81205e3f>] [<ffffffff81205e3f>] __check_object_size+0x13f/0x1d6 oct. 19 10:32:18 kernel: RSP: 0018:ffff8803ccabfd98 EFLAGS: 00010282 oct. 19 10:32:18 kernel: RAX: 0000000000000063 RBX: ffff8803ccabfe10 RCX: 0000000000000000 oct. 19 10:32:18 kernel: RDX: 0000000000000000 RSI: ffff88041facdba8 RDI: ffff88041facdba8 oct. 19 10:32:18 kernel: RBP: ffff8803ccabfdb8 R08: 000000000009dd9f R09: 0000000000000005 oct. 19 10:32:18 kernel: R10: 0000000000000541 R11: 0000000000000381 R12: 0000000000000010 oct. 19 10:32:18 kernel: R13: 0000000000000000 R14: ffff8803ccabfe20 R15: 00007fff8fa12320 oct. 19 10:32:18 kernel: FS: 00007f401049f4c0(0000) GS:ffff88041fac0000(0000) knlGS:0000000000000000 oct. 19 10:32:18 kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 oct. 19 10:32:18 kernel: CR2: 00000000016c7028 CR3: 00000003cc856000 CR4: 00000000001406e0 oct. 19 10:32:18 kernel: Stack: oct. 19 10:32:18 kernel: ffff8803ccabfe10 0000000000000010 00007fff8fa12320 ffff8803dfb99e00 oct. 19 10:32:18 kernel: ffff8803ccabfde0 ffffffffa0680b90 ffff8803ccabfe10 00007fff8fa12320 oct. 19 10:32:18 kernel: ffff88040c66d628 ffff8803d47dc800 ffffffffa067c83d 0000000000000001 oct. 19 10:32:18 kernel: Call Trace: oct. 19 10:32:18 kernel: [<ffffffffa0680b90>] bm_copy_from_user+0x20/0x40 [blackmagic_io] oct. 19 10:32:18 kernel: [<ffffffffa067c83d>] _ZN18IoctlMessageKernel6unpackEv+0x3d/0x140 [blackmagic_io] oct. 19 10:32:18 kernel: [<ffffffffa064566a>] ? _ZN20UserClientClassLinux5ioctlEjm+0x6a/0xa0 [blackmagic_io] oct. 19 10:32:18 kernel: [<ffffffff811b3cfe>] ? handle_mm_fault+0x3ee/0x14a0 oct. 19 10:32:18 kernel: [<ffffffffa067cc8a>] ? bmio_ioctl+0x1a/0x30 [blackmagic_io] oct. 19 10:32:18 kernel: [<ffffffff8121c313>] ? do_vfs_ioctl+0xa3/0x5f0 oct. 19 10:32:18 kernel: [<ffffffff81227547>] ? __fget+0x77/0xb0 oct. 19 10:32:18 kernel: [<ffffffff8121c8d9>] ? SyS_ioctl+0x79/0x90 oct. 19 10:32:18 kernel: [<ffffffff815f7cf2>] ? entry_SYSCALL_64_fastpath+0x1a/0xa4 oct. 19 10:32:18 kernel: Code: 87 71 81 48 0f 45 d0 48 c7 c6 60 a5 72 81 48 c7 c0 db 43 73 81 48 0f 45 f0 4d 89 e1 48 89 d9 48 c7 c7 18 0d 73 81 e8 67 02 f7 ff <0f> 0b 48 89 df e8 77 76 e6 oct. 19 10:32:18 kernel: RIP [<ffffffff81205e3f>] __check_object_size+0x13f/0x1d6 oct. 19 10:32:18 kernel: RSP <ffff8803ccabfd98> oct. 19 10:32:18 kernel: ---[ end trace 0139f39ddae78683 ]---

fthiery commented on 2016-09-20 15:00 (UTC)

I think that pulseaudio is using the snd_blackmagic_io module, therefore preventing it from being unloaded.

Kakashi-sensei commented on 2016-09-18 09:23 (UTC)

:: Running post-transaction hooks... (1/3) Install DKMS modules ==> dkms install -m blackmagic -v 10.8a2 -k 4.7.4-1-ARCH ==> dkms install -m blackmagic -v 10.8a2 -k 4.7.4-1-ck Good news! Module version 10.8a2 for blackmagic.ko exactly matches what is already found in kernel 4.7.4-1-ck. DKMS will not replace this module. You may override by specifying --force. ==> dkms install -m blackmagic-io -v 10.8a2 -k 4.7.4-1-ARCH ==> dkms install -m blackmagic-io -v 10.8a2 -k 4.7.4-1-ck Good news! Module version 10.8a2 for blackmagic-io.ko exactly matches what is already found in kernel 4.7.4-1-ck. DKMS will not replace this module. You may override by specifying --force. Good news! Module version 10.8a2 for snd_blackmagic-io.ko exactly matches what is already found in kernel 4.7.4-1-ck. DKMS will not replace this module. You may override by specifying --force. modprobe: FATAL: Module blackmagic_io is in use. ERR The existing blackmagic-io driver was not able to be unloaded ERR Please check 'dmesg' for more information

isopede commented on 2016-09-16 18:59 (UTC)

Updated to 10.8a2.

fthiery commented on 2016-09-16 10:15 (UTC)

I sent him an email

Excrement commented on 2016-09-15 01:07 (UTC)

fthiery, that worked like a charm.. question is; where is the maintainer..

fthiery commented on 2016-09-13 10:44 (UTC)

@Maintainer, compilation works if updated to the 10.8, here is a working PKGFILE # Maintainer: Perry Hung <> pkgname=decklink _dvver=10.8a2 # DesktopVideo _mever=3.5.2a2 # MediaExpress pkgver=${_dvver} pkgrel=1 pkgdesc="Drivers for Blackmagic Design DeckLink, Intensity or Multibridge video editing cards" arch=('i686' 'x86_64') url="" license=('custom') makedepends=('curl') depends=('linux-headers' 'libxml2' 'libpng12' 'glu' 'qt4') options=('!strip' 'staticlibs') install='decklink.install' [ "$CARCH" = "i686" ] && _arch='i386' [ "$CARCH" = "x86_64" ] && _arch='x86_64' pkgsrc_url="" pkgsrc_file=$pkgname-${_dvver}.tar.gz pkgsrc_sha256sum="09d3593b2fd499bdeb794e1cea3972c0dcd997d08eedbe959ea5965d18983268" prepare() { temp_url=`curl --data '{country":"us","platform":"Linux"}' $pkgsrc_url` curl -o $pkgsrc_file $temp_url shasum=`sha256sum $pkgsrc_file | cut -d " " -f1` [ "${shasum}" != "${pkgsrc_sha256sum}" ] && ( echo "Integrity check failed."; exit 1 ) tar xf ${pkgsrc_file} } package() { mkdir -p "$pkgdir/usr/share/licenses/$pkgname" ln -s /usr/share/doc/desktopvideo/License.txt "$pkgdir/usr/share/licenses/$pkgname/COPYING" cd "$srcdir/Blackmagic_Desktop_Video_Linux_${pkgver%a*}/other/${_arch}" tar xf "desktopvideo-${_dvver}-${_arch}.tar.gz" cp -a "desktopvideo-${_dvver}-${_arch}/"* "$pkgdir" mv "$pkgdir/usr/sbin/"* "$pkgdir/usr/bin" rm -rf "$pkgdir/usr/sbin" tar xf "mediaexpress-${_mever}-${_arch}.tar.gz" cp -a "mediaexpress-${_mever}-${_arch}/"* "$pkgdir" }

Deepcool commented on 2016-09-10 20:07 (UTC)

It appears I was able to install using linux kernel 4.4-lts. During the install it complained about linux Kernel 4.7 Here is some additional information. ==> Continue installing decklink ? [Y/n] ==> [v]iew package contents [c]heck package with namcap ==> --------------------------------------------------- ==> y loading packages... resolving dependencies... looking for conflicting packages... Packages (1) decklink-10.6.7a1-1 Total Installed Size: 155.28 MiB :: Proceed with installation? [Y/n] y (1/1) checking keys in keyring [#########################] 100% (1/1) checking package integrity [#########################] 100% (1/1) loading package files [#########################] 100% (1/1) checking for file conflicts [#########################] 100% (1/1) checking available disk space [#########################] 100% :: Processing package changes... (1/1) installing decklink [#########################] 100% warning: directory permissions differ on /usr/ filesystem: 755 package: 775 warning: directory permissions differ on /usr/share/ filesystem: 755 package: 775 warning: directory permissions differ on /usr/share/doc/ filesystem: 755 package: 775 make -C /lib/modules/`uname -r`/build M=/usr/src/blackmagic-10.6.7a1 make[1]: Entering directory '/usr/lib/modules/4.4.19-1-lts/build' LD /usr/src/blackmagic-10.6.7a1/built-in.o CC [M] /usr/src/blackmagic-10.6.7a1/blackmagic_serial.o CC [M] /usr/src/blackmagic-10.6.7a1/blackmagic_core.o CC [M] /usr/src/blackmagic-10.6.7a1/blackmagic_lib.o CC [M] /usr/src/blackmagic-10.6.7a1/blackmagic_dma.o CC [M] /usr/src/blackmagic-10.6.7a1/blackmagic_gate.o LD [M] /usr/src/blackmagic-10.6.7a1/blackmagic.o Building modules, stage 2. MODPOST 1 modules CC /usr/src/blackmagic-10.6.7a1/blackmagic.mod.o LD [M] /usr/src/blackmagic-10.6.7a1/blackmagic.ko make[1]: Leaving directory '/usr/lib/modules/4.4.19-1-lts/build' rm -rf *.o *~ core .depend .*.cmd *.ko *.mod.c .tmp_versions built-in.o Module.symvers Module.markers modules.order make -C /lib/modules/`uname -r`/build M=/usr/src/blackmagic-io-10.6.7a1 make[1]: Entering directory '/usr/lib/modules/4.4.19-1-lts/build' LD /usr/src/blackmagic-io-10.6.7a1/built-in.o CC [M] /usr/src/blackmagic-io-10.6.7a1/bmio_client.o CC [M] /usr/src/blackmagic-io-10.6.7a1/bmio_dev.o CC [M] /usr/src/blackmagic-io-10.6.7a1/bmio_device.o CC [M] /usr/src/blackmagic-io-10.6.7a1/bmio_driver.o CC [M] /usr/src/blackmagic-io-10.6.7a1/bmio_serial.o CC [M] /usr/src/blackmagic-io-10.6.7a1/bmio_export.o CC [M] /usr/src/blackmagic-io-10.6.7a1/bmio_pci_ids.o CC [M] /usr/src/blackmagic-io-10.6.7a1/bm_locks.o CC [M] /usr/src/blackmagic-io-10.6.7a1/bm_mm.o CC [M] /usr/src/blackmagic-io-10.6.7a1/bm_pci.o CC [M] /usr/src/blackmagic-io-10.6.7a1/bm_util.o LD [M] /usr/src/blackmagic-io-10.6.7a1/blackmagic-io.o CC [M] /usr/src/blackmagic-io-10.6.7a1/bmio_audio.o LD [M] /usr/src/blackmagic-io-10.6.7a1/snd_blackmagic-io.o Building modules, stage 2. MODPOST 2 modules CC /usr/src/blackmagic-io-10.6.7a1/blackmagic-io.mod.o LD [M] /usr/src/blackmagic-io-10.6.7a1/blackmagic-io.ko CC /usr/src/blackmagic-io-10.6.7a1/snd_blackmagic-io.mod.o LD [M] /usr/src/blackmagic-io-10.6.7a1/snd_blackmagic-io.ko make[1]: Leaving directory '/usr/lib/modules/4.4.19-1-lts/build' >>> Updating module dependencies. Please wait... >>> Loading modules... :: Running post-transaction hooks... (1/3) Install DKMS modules ==> dkms install -m blackmagic -v 10.6.7a1 -k 4.4.19-1-lts Good news! Module version 10.6.7a1 for blackmagic.ko exactly matches what is already found in kernel 4.4.19-1-lts. DKMS will not replace this module. You may override by specifying --force. ==> No kernel 4.7.1-1-ARCH headers. You must install them to use DKMS! ==> No kernel 4.7.2-1-ARCH modules. You must install them to use DKMS! ==> dkms install -m blackmagic-io -v 10.6.7a1 -k 4.4.19-1-lts Good news! Module version 10.6.7a1 for blackmagic-io.ko exactly matches what is already found in kernel 4.4.19-1-lts. DKMS will not replace this module. You may override by specifying --force. Good news! Module version 10.6.7a1 for snd_blackmagic-io.ko exactly matches what is already found in kernel 4.4.19-1-lts. DKMS will not replace this module. You may override by specifying --force. modprobe: FATAL: Module blackmagic_io is in use. ERR The existing blackmagic-io driver was not able to be unloaded ERR Please check 'dmesg' for more information ==> No kernel 4.7.1-1-ARCH headers. You must install them to use DKMS! ==> No kernel 4.7.2-1-ARCH modules. You must install them to use DKMS! (2/3) Updating icon theme caches... (3/3) Updating the desktop file MIME type cache...  dmbomer  ~  

Excrement commented on 2016-09-08 16:23 (UTC)

yea, im curious whats the status as im getting similar errors

Kakashi-sensei commented on 2016-09-03 11:58 (UTC)

The software was updated on August 1st. Has the issue been solved?

jankoppe commented on 2016-06-24 13:16 (UTC)

@rstoye This is a problem with the decklink driver itself, and can only be fixed by blackmagic. I've already contacted the blackmagic development departement, so that they look into it.

rstoye commented on 2016-06-24 00:50 (UTC)

doesn't compile with linux-4.6.2-1-x86_64.pkg.tar.xz had to downgrade to linux-4.5.4-1-x86_64.pkg.tar.xz errors encountered: pacman -U decklink-10.6.7a1-1-x86_64.pkg.tar.xz ... /usr/src/blackmagic-10.6.7a1/blackmagic_lib.c/usr/src/blackmagic-10.6.7a1/blackmagic_lib.c:674:8: Fehler: zu viele Argumente für Funktion »get_user_pages« ret = get_user_pages(current_task, ... /usr/src/blackmagic-10.6.7a1/blackmagic_lib.c:704:3: Fehler: Implizite Deklaration der Funktion »page_cache_release« both are related to kernel changes: get_user_pages gets 6 Parameters now (the first two are defaults now, so maybe it should be changed to: ret = get_user_pages((unsigned long)ptr & PAGE_MASK, *nr_pages, write, 0, pages, NULL); 2nd problem: /usr/src/blackmagic-10.6.7a1/blackmagic_lib.c:704:3: Fehler: Implizite Deklaration der Funktion »page_cache_release« thats because page_cache_release was renamed to put_page sadly i am a newbie regarding ARCH, so maybe someone else can fix it?...