Support for Intel Arc? #7026

uhthomas · 2023-03-30T13:41:28Z

uhthomas
Mar 30, 2023

Hi!

Does Talos Linux support Intel Arc at all? Intel has official GPU support for Kubernetes, but I couldn't see any documentation or system extensions for Talos which explicitly support Intel Arc.

https://github.com/siderolabs/extensions

I do see i915, though my impression is that it's for integrated graphics?

These articles suggest there should be native support in the Linux kernel from 6.2, but I believe Talos 1.4 alpha currently targets 6.1?

https://news.itsfoss.com/linux-kernel-6-2-release/

https://www.phoronix.com/review/linux-62-features

smira · 2023-03-30T16:29:21Z

smira
Mar 30, 2023
Maintainer

Talos will stay on Linux 6.1 for the year of 2023, as it's the LTS release.

We're happy to accept any system extensions which enable support for GPUs.

8 replies

smira Mar 31, 2023
Maintainer

Talos ships with the latest LTS release of Linux.

Linux 6.2 is never going to be an LTS release. Next LTS after 6.1 will come late 2023 or early 2024.

uhthomas Mar 31, 2023
Author

So Talos targets LTS not stable. Makes sense - thanks!

uhthomas Apr 2, 2023
Author

@smira I thought it would be helpful to update you on how I've been getting on.

It hasn't been as easy as I'd hoped. I upgraded the Linux kernel and it would get stuck on

10.0.0.104: kern:    info: [2023-04-02T18:25:22.172270158Z]: pci 0000:06:00.0: vgaarb: VGA device added: decodes=io+mem,owns=none,locks=none
 SUBSYSTEM=pci

Then nothing. I don't know if it got stuck there and then hung, or whether it just kept going and stopped updating the graphics output.

I messed around a bit and eventually made some changes to the i915 configuration and like before, the VGA output stopped at that message. However talosctl dmesg shows that it continued to boot and try to join the cluster. So there are two things for me now:

Why didn't the machine join the cluster automatically? It says

10.0.0.104: user: warning: [2023-04-02T18:26:54.990833158Z]: [talos] etcd is waiting to join the cluster, if this node is the first node in the cluster, please run `talosctl bootstrap` against one of the following IPs:

I don't think Intel Arc works properly with Talos and Linux 6.2. Please see the full i915 dmesg log:

10.0.0.104: kern:    info: [2023-04-02T18:26:23.423385158Z]: i915 0000:06:00.0: enabling device (0000 -> 0002)
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:23.493013158Z]: i915 0000:06:00.0: BAR 0: releasing [mem 0xd8000000-0xd8ffffff 64bit]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:23.582741158Z]: i915 0000:06:00.0: BAR 2: releasing [mem 0x3ffe0000000-0x3ffefffffff 64bit pref]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:23.683912158Z]: pcieport 0000:05:01.0: BAR 9: releasing [mem 0x3ffe0000000-0x3ffff7fffff 64bit pref]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:05:01.0
10.0.0.104: kern:    info: [2023-04-02T18:26:23.789237158Z]: pcieport 0000:04:00.0: BAR 9: releasing [mem 0x3ffe0000000-0x3ffff7fffff 64bit pref]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:04:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:23.894576158Z]: pcieport 0000:04:00.0: BAR 9: no space for [mem size 0x200000000 64bit pref]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:04:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:23.991567158Z]: pcieport 0000:04:00.0: BAR 9: failed to assign [mem size 0x200000000 64bit pref]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:04:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:24.092741158Z]: pcieport 0000:05:01.0: BAR 9: no space for [mem size 0x200000000 64bit pref]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:05:01.0
10.0.0.104: kern:    info: [2023-04-02T18:26:24.189736158Z]: pcieport 0000:05:01.0: BAR 9: failed to assign [mem size 0x200000000 64bit pref]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:05:01.0
10.0.0.104: kern:    info: [2023-04-02T18:26:24.290895158Z]: i915 0000:06:00.0: BAR 2: no space for [mem size 0x200000000 64bit pref]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:24.383739158Z]: i915 0000:06:00.0: BAR 2: failed to assign [mem size 0x200000000 64bit pref]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:24.480731158Z]: i915 0000:06:00.0: BAR 0: assigned [mem 0xd8000000-0xd8ffffff 64bit]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:24.569428158Z]: pcieport 0000:00:03.0: PCI bridge to [bus 04-07]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:00:03.0
10.0.0.104: kern:    info: [2023-04-02T18:26:24.637285158Z]: pcieport 0000:00:03.0:   bridge window [mem 0xd8000000-0xdaffffff]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:00:03.0
10.0.0.104: kern:    info: [2023-04-02T18:26:24.735017158Z]: pcieport 0000:00:03.0:   bridge window [mem 0x3ffe0000000-0x3ffffffffff 64bit pref]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:00:03.0
10.0.0.104: kern:    info: [2023-04-02T18:26:24.839289158Z]: pcieport 0000:04:00.0: PCI bridge to [bus 05-07]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:04:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:24.907136158Z]: pcieport 0000:04:00.0:   bridge window [mem 0xd8000000-0xdaffffff]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:04:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:24.993722158Z]: pcieport 0000:04:00.0:   bridge window [mem 0x3ffe0000000-0x3ffff7fffff 64bit pref]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:04:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:25.098007158Z]: pcieport 0000:05:01.0: PCI bridge to [bus 06]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:05:01.0
10.0.0.104: kern:    info: [2023-04-02T18:26:25.162725158Z]: pcieport 0000:05:01.0:   bridge window [mem 0xd8000000-0xd9ffffff]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:05:01.0
10.0.0.104: kern:    info: [2023-04-02T18:26:25.249307158Z]: pcieport 0000:05:01.0:   bridge window [mem 0x3ffe0000000-0x3ffff7fffff 64bit pref]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:05:01.0
10.0.0.104: kern:    info: [2023-04-02T18:26:25.353584158Z]: i915 0000:06:00.0: [drm] Failed to resize BAR2 to 8192M (-ENOSPC)
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:25.439121158Z]: i915 0000:06:00.0: BAR 2: assigned [mem 0x3ffe0000000-0x3ffefffffff 64bit pref]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:25.539265158Z]: i915 0000:06:00.0: [drm] Local memory IO size: 0x0000000010000000
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:25.624779158Z]: i915 0000:06:00.0: [drm] Local memory available: 0x000000017c800000
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:25.712410158Z]: i915 0000:06:00.0: [drm] Using a reduced BAR size of 256MiB. Consider enabling 'Resizable BAR' or similar, if available in the BIOS.
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:25.882651158Z]: i915 0000:06:00.0: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=none:owns=none
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:     err: [2023-04-02T18:26:25.993677158Z]: i915 0000:06:00.0: GuC firmware i915/dg2_guc_70.bin: fetch failed with error -2
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:  notice: [2023-04-02T18:26:26.093711158Z]: i915 0000:06:00.0: Please file a bug on drm/i915; see https://gitlab.freedesktop.org/drm/intel/-/wikis/How-to-file-i915-bugs for details.
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:26.254080158Z]: i915 0000:06:00.0: [drm] GuC firmware(s) can be downloaded from https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/i915
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:     err: [2023-04-02T18:26:26.527557158Z]: i915 0000:06:00.0: [drm] *ERROR* render: timed out waiting for forcewake ack to clear.
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:  notice: [2023-04-02T18:26:26.634858158Z]: i915 0000:06:00.0: [drm:add_taint_for_CI [i915]] CI tainted:0x9 by fw_domains_get_with_fallback+0x222/0x240 [i915]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:     err: [2023-04-02T18:26:26.872087158Z]: i915 0000:06:00.0: [drm] *ERROR* gt: timed out waiting for forcewake ack to clear.
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:  notice: [2023-04-02T18:26:26.975227158Z]: i915 0000:06:00.0: [drm:add_taint_for_CI [i915]] CI tainted:0x9 by fw_domains_get_with_fallback+0x222/0x240 [i915]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:     err: [2023-04-02T18:26:27.212462158Z]: i915 0000:06:00.0: [drm] *ERROR* vdbox0: timed out waiting for forcewake ack to clear.
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:  notice: [2023-04-02T18:26:27.319764158Z]: i915 0000:06:00.0: [drm:add_taint_for_CI [i915]] CI tainted:0x9 by fw_domains_get_with_fallback+0x222/0x240 [i915]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:     err: [2023-04-02T18:26:27.556994158Z]: i915 0000:06:00.0: [drm] *ERROR* vdbox2: timed out waiting for forcewake ack to clear.
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:  notice: [2023-04-02T18:26:27.664292158Z]: i915 0000:06:00.0: [drm:add_taint_for_CI [i915]] CI tainted:0x9 by fw_domains_get_with_fallback+0x222/0x240 [i915]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:     err: [2023-04-02T18:26:27.901563158Z]: i915 0000:06:00.0: [drm] *ERROR* vebox0: timed out waiting for forcewake ack to clear.
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:  notice: [2023-04-02T18:26:28.008859158Z]: i915 0000:06:00.0: [drm:add_taint_for_CI [i915]] CI tainted:0x9 by fw_domains_get_with_fallback+0x222/0x240 [i915]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:     err: [2023-04-02T18:26:28.246092158Z]: i915 0000:06:00.0: [drm] *ERROR* vebox1: timed out waiting for forcewake ack to clear.
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:  notice: [2023-04-02T18:26:28.353396158Z]: i915 0000:06:00.0: [drm:add_taint_for_CI [i915]] CI tainted:0x9 by fw_domains_get_with_fallback+0x222/0x240 [i915]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:28.491614158Z]: i915 0000:06:00.0: [drm] GuC firmware i915/dg2_guc_70.bin version 0.0.0
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:     err: [2023-04-02T18:26:28.583414158Z]: i915 0000:06:00.0: GuC initialization failed -2
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:     err: [2023-04-02T18:26:28.650150158Z]: i915 0000:06:00.0: Enabling uc failed (-5)
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:     err: [2023-04-02T18:26:28.711683158Z]: i915 0000:06:00.0: Failed to initialize GPU, declaring it wedged!
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern: warning: [2023-04-02T18:26:28.797679158Z]: i915 0000:06:00.0: Direct firmware load for i915/dg2_dmc_ver2_08.bin failed with error -2
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:  notice: [2023-04-02T18:26:28.908121158Z]: i915 0000:06:00.0: [drm] Failed to load DMC firmware i915/dg2_dmc_ver2_08.bin. Disabling runtime power management.
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:  notice: [2023-04-02T18:26:29.044572158Z]: i915 0000:06:00.0: [drm] DMC firmware homepage: https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/i915
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:  notice: [2023-04-02T18:26:29.201341158Z]: i915 0000:06:00.0: [drm:add_taint_for_CI [i915]] CI tainted:0x9 by intel_gt_init+0xb3/0x330 [i915]
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:29.366523158Z]: [drm] Initialized i915 1.6.0 20201103 for 0000:06:00.0 on minor 0
10.0.0.104: kern:    info: [2023-04-02T18:26:29.453443158Z]: i915 0000:06:00.0: [drm] Cannot find any crtc or sizes
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0
10.0.0.104: kern:    info: [2023-04-02T18:26:29.527610158Z]: i915 0000:06:00.0: [drm] Cannot find any crtc or sizes
 SUBSYSTEM=pci
 DEVICE=+pci:0000:06:00.0

uhthomas Apr 2, 2023
Author

fwiw opened https://gitlab.freedesktop.org/drm/intel/-/issues/8328

uhthomas Apr 2, 2023
Author

Okay! So I eventually just replaced the node and made sure the i915-code extension installed and it seems to be okay now.

"intel.feature.node.kubernetes.io/gpu": "true",

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Support for Intel Arc? #7026

{{title}}

{{editor}}'s edit

{{editor}}'s edit

Replies: 1 comment 8 replies

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

Select a reply

Support for Intel Arc? #7026

uhthomas Mar 30, 2023

Replies: 1 comment · 8 replies

smira Mar 30, 2023 Maintainer

smira Mar 31, 2023 Maintainer

uhthomas Mar 31, 2023 Author

uhthomas Apr 2, 2023 Author

uhthomas Apr 2, 2023 Author

uhthomas Apr 2, 2023 Author

uhthomas
Mar 30, 2023

Replies: 1 comment 8 replies

smira
Mar 30, 2023
Maintainer

smira Mar 31, 2023
Maintainer

uhthomas Mar 31, 2023
Author

uhthomas Apr 2, 2023
Author

uhthomas Apr 2, 2023
Author

uhthomas Apr 2, 2023
Author