just did the may update , it failed with : Error installing: vero564-image-4.9.269-93-osmc:arm64
uploaded the logs right after :
https://paste.osmc.tv/geyiziwalu
nothing HDR plays right now , the framerate is very low , the MAX blue beggining color looks purple .
Reboot and update again. This should solve the problem.
Sam
i tried , but the system said that new updates were available ,
made a backup of setting reinstalled system to the august update and then updated to may and it went trough no issue . just lost one series time slot with the missus , but all is good now .
never had an update fail before .
I’ve identified the issue, it’s been around for a while.
Rebooting before updating should fix it but this shouldn’t be necessary in the near future.
Sam,
I’m having the same problem that the update failed. tried to run update again and no updates available. rebooted and tried running updates and updates aren’t available. in my case audio plays when I play a dvd/bluray or uhd. but video is just stuttering black with the menu in the background. no image. blue circle in the middle. I can post logs but I don’t have logs after the failed update. is there anything I can do to get this to see the proper updates?
here are logs (edited) when trying to play first a dvd, then a blu ray, and then a UHD. first attempt at playing a dvd failed b/c the hard drive wasn’t loaded yet. I waited for a bit and then it was loaded.
Try the following:
sudo apt updatesudo apt install --reinstall vero564-image-4.9.269-93-osmc vero564-kernel-osmc vero5-device-osmcrebootThanks Jim, but unfortunately I get the following error:
E: Internal Error, No file name for vero564-image-4.9.269-93-osmc:arm64
It looks like your system already had strange problems during the last update in December 2025:
Log started: 2025-12-15 18:04:01
...
Setting up vero564-image-4.9.269-87-osmc:arm64 (4.9.269-87-osmc) ...
10+1 records in
10+1 records out
11109120 bytes (11 MB, 11 MiB) copied, 0.178899 s, 62.1 MB/s
/bin/dd: error writing '/dev/dtb': Cannot allocate memory
0+1 records in
0+0 records out
0 bytes copied, 0.0348605 s, 0.0 kB/s
/bin/dd: error writing '/dev/dtb': Cannot allocate memory
0+1 records in
0+0 records out
0 bytes copied, 0.000668867 s, 0.0 kB/s
Could not upload DTB to eMMC
run-parts: /etc/kernel/postinst.d/upload-to-emmc exited with return code 1
dpkg: error processing package vero564-image-4.9.269-87-osmc:arm64 (--configure):
installed vero564-image-4.9.269-87-osmc:arm64 package post-installation script subprocess returned error exit status 1
And now
Log started: 2026-05-04 21:04:48
...
Setting up vero564-image-4.9.269-93-osmc:arm64 (4.9.269-93-osmc) ...
10+1 records in
10+1 records out
11107072 bytes (11 MB, 11 MiB) copied, 0.106631 s, 104 MB/s
/bin/dd: error writing '/dev/dtb': Cannot allocate memory
0+1 records in
0+0 records out
0 bytes copied, 0.00103978 s, 0.0 kB/s
/bin/dd: error writing '/dev/dtb': Cannot allocate memory
0+1 records in
0+0 records out
0 bytes copied, 0.000744618 s, 0.0 kB/s
Could not upload DTB to eMMC
run-parts: /etc/kernel/postinst.d/upload-to-emmc exited with return code 1
dpkg: error processing package vero564-image-4.9.269-93-osmc:arm64 (--configure):
installed vero564-image-4.9.269-93-osmc:arm64 package post-installation script subprocess returned error exit status 1
Setting up libpam-systemd:armhf (247.3-7+deb11u8) ...
Hey Jim - what would you suggest that I do then? Thanks.
You can reboot and try:
sudo apt-get update
sudo apt-get -f install
I had a look at this problem, as it’s lingered for a while and affects users with high playback uptime when they update. It should be fixed in the next update, but obviously users will be affected until this update is propagated. In short:
mmc_dtb_write() was calling kmalloc(256KB) on every write. On high-uptime
systems with active video playback, Amlogic CMA compaction hooks block the
buddy allocator from coalescing the required order-6 pages, causing the
allocation to fail with ENOMEM and the kernel postinst DTB upload to abort.
Pre-allocate the write buffer once at amlmmc_dtb_init() time before CMA
users fragment memory. The buffer must remain physically contiguous because
mmc_transfer() passes it directly to sg_init_one() for DMA.
thanks Sam. That change did the trick. I don’t understand the intricacies of what you’ve mentioned, but would it make sense for me, before the next update, to turn off the system for a while?
Next time you see an update, click No.
Reboot.
Go to My OSMC → Updates → Manual Controls → Check for Updates Now for it to prompt you.
And then it’s safe. But you already know how to safely update if worst comes to worst.
I need to do a bit more testing on my proposed fix but if it’s good I’ll get it out shortly so this is entirely avoidable.
Cheers
Sam
I had the same error and same symptom, even for files I know played this week, so guess I’ll follow the same advice as you did.
Yay, fixed for me too thanks.
Glad to hear this.