[Support] ich777 - AMD Vendor Reset, CoralTPU, hpsahba,...


Recommended Posts

My current problem is that the Intel GPU Top plugin is not creating the /dev/dri direrectory.

 

I'm new to Unraid and just built my first server. Loving it so far but can't get hardware transcoding in Plex. Last night I installed the Intel GPU Top plugin and I swear it created the /dev/dri directory. I remember being surprised it was that easy. But I still couldn't get hardware transcoding going. Today I realized I had missed the instructions for adding the /dev/dri path to the Plex container. After I added it and tried to start the container it failed and the logs showed that there is no /dev/dri path. Sure enough it's gone. I did reboot this morning, before doing this, so I'm assuming that's when I lost it.

 

I've tried uninstalling/reinstalling the plugin, but still no /dev/dri. I'm way out of my depth here so would appreciate any thoughts on what to try next.

Link to comment
1 minute ago, naxos said:

My current problem is that the Intel GPU Top plugin is not creating the /dev/dri direrectory.

 

I'm new to Unraid and just built my first server. Loving it so far but can't get hardware transcoding in Plex. Last night I installed the Intel GPU Top plugin and I swear it created the /dev/dri directory. I remember being surprised it was that easy. But I still couldn't get hardware transcoding going. Today I realized I had missed the instructions for adding the /dev/dri path to the Plex container. After I added it and tried to start the container it failed and the logs showed that there is no /dev/dri path. Sure enough it's gone. I did reboot this morning, before doing this, so I'm assuming that's when I lost it.

 

I've tried uninstalling/reinstalling the plugin, but still no /dev/dri. I'm way out of my depth here so would appreciate any thoughts on what to try next.

Are you running 6.10rc2? if so you may need to add the black list back in then intel top should load correctly. 

 

if you run echo "blacklist i915" > /boot/config/modprobe.d/i915.conf and then reboot.

Link to comment

Sorry. Of course I forget to include any details on my very first forum post.

 

Unraid v6.9.2

cpu: i5-11400

mobo: MSI b650 mortar wifi

 

I have read through so many posts on this that my brain is soup ... and I may well have missed a post that already has the solution I need.

Link to comment
4 minutes ago, naxos said:

Sorry. Of course I forget to include any details on my very first forum post.

 

Unraid v6.9.2

cpu: i5-11400

mobo: MSI b650 mortar wifi

 

I have read through so many posts on this that my brain is soup ... and I may well have missed a post that already has the solution I need.

Have you enabled the igpu in the bios. If you have physical gpu in the system also it will default to the physical 

 

have a look at this thread 

 

Link to comment
4 hours ago, SimonF said:

Have you enabled the igpu in the bios. If you have physical gpu in the system also it will default to the physical

 

Yes, the iGPU is enabled in the BIOS, set to GPU (and not PEG). I had gone through the thread you linked yesterday, but went through it again today. I did these steps:

  1. Removed all text from from i915.conf and removed all gpu-related lines from syslinux.conf
  2. uninstalled intel gpu top plugin
  3. rebooted
  4. reinstalled intel gpu top plugin
  5. rebooted - still no /dev/dri folder
  6. added "options i915 force_probe=4c8a" to i915.conf
  7. rebooted - still no /dev/dri folder
  8. added "i915.force_probe=4c8B" to syslinux.conf
  9. rebooted - now there was a /dev/dri folder, but ls -la /dev/dri/ gave me the message about no i915 being present (can't remember the exact message)

So, this is where I am currently. Any other thoughts? It may be that I just need to sit on my hands until Unraid and/or Plex have releases that make this work out of the box.

Link to comment
9 hours ago, Nobi said:

ich hab eine RX570 und Unraid version: Version: 6.10.0-rc2 

somit habe ich den AMD-Vendor-Reset plugin installiert.

 

Mein Problem ist das sich Unraid aufhängt wenn ich meine VM Macinabox  Herunterfahre.

Nach der Installation des Plugins kann ich zwar die VM neustarten aber beim Herunterfahren hängt sich Unraid immer noch auf.

Bitte ab in den Deutschen Bereich damit und mich markieren wie im folgenden Beispiel und nochmal auf meinen Nicknamen klicken, hier gibt es nur Support auf Englisch:

grafik.png.701b61aab2a1e0a71b899a41ddedd281.png

 

Link to comment
8 hours ago, naxos said:

My current problem is that the Intel GPU Top plugin is not creating the /dev/dri direrectory.

Please share your Diagnostics.

 

8 hours ago, naxos said:

no /dev/dri path

Have you mounted it as Path or as Device, these are two different things and it has to be mounted as Path.

 

Do you have a dedicated GPU installed, the Diagnostics would also show that...

 

EDIT:

3 hours ago, naxos said:

Any other thoughts?

Please start over, delete the file i915.conf in your /boot/config/modprobe.d/ directory, uninstall the Intel GPU TOP plugin, undo the changes to your syslinux.conf, restart your server, after that install the Intel GPU TOP plugin again and then pull your Diagnostics please.

  • Like 1
Link to comment
6 hours ago, Dazog said:

Mellanox 4LX and I run no vm's

I will try to reflash my card (Mellanox Connect X3) and will see if the same happens to me and report back.

 

Where you able to flash this exact card in the past or did you flashed other cards?

 

Do you use this command: 'mstflint -d 01:00.0 -i *.bin burn' with the asterisk or do you specify the full filename?

 

EDIT: Flash went fine without any error:

mstflint -d 01:00.0 -i fw-ConnectX3-rel-2_42_5000-MCX311A-XCA_Ax-FlexBoot-3.4.752.bin burn

    Current FW version on flash:  2.42.5000
    New FW version:               2.42.5000

    Note: The new FW version is the same as the current FW version on flash.

 Do you want to continue ? (y/n) [n] : y

Burn process will not be failsafe. No checks will be performed.
ALL flash, including the Invariant Sector will be overwritten.
If this process fails, computer may remain in an inoperable state.

 Do you want to continue ? (y/n) [n] : y
Burning FS2 FW image without signatures - OK  
Restoring signature                     - OK

 

Also removed the boot ROM after that:

mstflint -d 01:00.0 --allow_rom_change drom

-I- Preparing to remove ROM ...
Removing ROM image    - OK  
Restoring signature  - OK

 

Link to comment
14 hours ago, ich777 said:

Please start over, delete the file i915.conf in your /boot/config/modprobe.d/ directory, uninstall the Intel GPU TOP plugin, undo the changes to your syslinux.conf, restart your server, after that install the Intel GPU TOP plugin again and then pull your Diagnostics please.

 

I completed all the above steps and attached are my diagnostics after a reboot. Thank you for trying to help me.

 

I don't have a dedicated GPU installed. The one brief time I had a /dev/dri folder I tried mounting it to the Plex container as a path.

galaga-diagnostics-20211208-1456.zip

Link to comment
20 hours ago, ich777 said:

I will try to reflash my card (Mellanox Connect X3) and will see if the same happens to me and report back.

 

Where you able to flash this exact card in the past or did you flashed other cards?

 

Do you use this command: 'mstflint -d 01:00.0 -i *.bin burn' with the asterisk or do you specify the full filename?

 

EDIT: Flash went fine without any error:

mstflint -d 01:00.0 -i fw-ConnectX3-rel-2_42_5000-MCX311A-XCA_Ax-FlexBoot-3.4.752.bin burn

    Current FW version on flash:  2.42.5000
    New FW version:               2.42.5000

    Note: The new FW version is the same as the current FW version on flash.

 Do you want to continue ? (y/n) [n] : y

Burn process will not be failsafe. No checks will be performed.
ALL flash, including the Invariant Sector will be overwritten.
If this process fails, computer may remain in an inoperable state.

 Do you want to continue ? (y/n) [n] : y
Burning FS2 FW image without signatures - OK  
Restoring signature                     - OK

 

Also removed the boot ROM after that:

mstflint -d 01:00.0 --allow_rom_change drom

-I- Preparing to remove ROM ...
Removing ROM image    - OK  
Restoring signature  - OK

 

I use the asterisk

 

I will try using the full name.

  • Like 1
Link to comment
8 hours ago, naxos said:

I completed all the above steps and attached are my diagnostics after a reboot. Thank you for trying to help me.

I think you don't installed the Intel GPU TOP plugin.

But from what I see you don't installed the Intel GPU TOP plugin, at least it's not listed in the Diagnostics and the Kernel module is also not loaded.

 

The Intel GPU TOP plugin should probe against the device:

4c8b

and should activate your iGPU without any file or modification.

 

Please install the Intel GPU TOP plugin and then do:

ls -la /dev/dri

If you get an error from that command please post your Diagnostics again.

  • Like 1
Link to comment
10 hours ago, ich777 said:

Please install the Intel GPU TOP plugin and then do:

ls -la /dev/dri

If you get an error from that command please post your Diagnostics again.

 

I installed the Intel GPU TOP plugin. ls -la /dev/dri returns "cannot access '/dev/dri': No such file or directory." Diagnostics attached.

 

galaga-diagnostics-20211209-1629.zip

 

EDIT: Attached a picture with confirmation that the iGPU is enabled in the BIOS:

image.thumb.png.44471ecf8b34927ad6c36fac5de4e439.png

 

Edited by naxos
Link to comment
Just now, naxos said:

I installed the Intel GPU TOP plugin. ls -la /dev/dri returns "cannot access '/dev/dri': No such file or directory." Diagnostics attached.

Got your issue, please remove those two lines from your go file:

modprobe i915
chmod -R 777 /dev/dri

after that reboot your system with the Intel GPU TOP plugin installed and we can further troubleshoot if it's not working after the reboot.

 

Also do you have bluetooth enabled on your system and do you actively using it? If not, please disable it in your BIOS.

Your syslog is full with this messages:

Dec  9 04:40:02 Galaga kernel: usb 1-2: USB disconnect, device number 91
Dec  9 04:40:02 Galaga kernel: usb 1-2: new full-speed USB device number 92 using xhci_hcd
Dec  9 04:40:02 Galaga kernel: Bluetooth: hci0: Reading Intel version information failed (-22)
Dec  9 04:40:02 Galaga kernel: Bluetooth: hci0: Intel Read version failed (-22)
Dec  9 04:40:02 Galaga kernel: Bluetooth: hci0: Intel reset sent to retry FW download

 

Link to comment
1 hour ago, ich777 said:

I see that the iGPU is enabled from your Diagnostics, but with that many customization to the system it's really difficult to find what causes this that Intel GPU TOP isn't working.

 

Apparently I am not able to follow simple instructions very well this week. I removed the lines from my go file and rebooted. ls -la /dev/dri now looks good. Earlier you said to add /dev/dri as a path to the Plex container but I haven't been able to find any specific details. I assume /dev/dri is the host path, but what should the container path be?

 

I tried adding it as as device, as many other posts suggest, and it seems like I am getting hardware transcode on my Shield, but on my phone app and laptop browser, when I force transcode I just get a spinning wheel after clicking play.

 

P.S. Thanks for the reminder about the Bluetooth issue. I had it on my list of things to look at after I get this transcoding issue fixed, but I disabled it in the BIOS when I rebooted.

  • Like 1
Link to comment
3 minutes ago, naxos said:

I assume /dev/dri is the host path, but what should the container path be?

 

the same /dev/dri on host and docker, but adding as device should be fine in plex

 

when transcoding is working on 1 device but not on the other(s) i would start checking the plex logs (webui, buttom left, console ...) and check the plex forums with the errors you get.

 

btw.

 

6 minutes ago, naxos said:

I am getting hardware transcode on my Shield

if your shield is local in your lan there shouldnt be any transcode running, you should always direct play instead ... just in case you are not testing only now to see if transcode is working.

 

browser, which one are you using for "testing" ?

phone, which one ?

 

may all ios devices ?

  • Like 2
Link to comment
1 hour ago, alturismo said:

if your shield is local in your lan there shouldnt be any transcode running, you should always direct play instead ... just in case you are not testing only now to see if transcode is working.

 

browser, which one are you using for "testing" ?

phone, which one ?

 

may all ios devices ?

 

I forced transcoding on the shield for testing. I definitely direct play on the LAN. I was testing with Edge Chromium on a Windows 10 laptop and an Android 11 phone with the Plex app. Both can direct play with no problem. I had started digging into the Plex logs, but wanted to make sure the problem wasn't related to adding /dev/dri as a device and not a path.

 

Thank you very much for all your help! I'll work my remaining transcoding issue over on the Plex forums.

Link to comment
1 hour ago, alturismo said:

the same /dev/dri on host and docker, but adding as device should be fine in plex

 

Interestingly, when I removed /dev/dri as a device and added it as a path I wasn't getting any hardware transcoding. Changing it back to a device re-enabled hardware transcoding.

Link to comment
22 minutes ago, naxos said:

Interestingly, when I removed /dev/dri as a device and added it as a path I wasn't getting any hardware transcoding. Changing it back to a device re-enabled hardware transcoding.

 

Just an update in case it helps anyone else, unchecking "Enable HDR tone mapping" in Settings > Transcoder appears to have fixed the problem for me. I tested with three simultaneous 1080p transcodes and they all did hardware transcoding with no problem. This is with /dev/dri added to the Plex container as a device.

  • Like 1
Link to comment
7 hours ago, alturismo said:

what cpu do you have ?

 

I've got an i5-11400. At this point it seems like my one remaining problem is that all remote streaming, using clients capable of direct streaming, is forced to transcode at 2 Mbps which takes it down to SD, I assume because they're being forced through the Plex relay, which caps at 2 Mbps. This did not happen on my HP Mini with an I5-8500T with a bare metal install of Plex. Looks like I'm going to have to give myself a crash course on docker and Unraid networking to figure this out. I'm also going to get my HP server back online and confirm that it will still let remote clients direct stream.

Link to comment
  • ich777 changed the title to [Support] ich777 - AMD Vendor Reset, CoralTPU, hpsahba,...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.