B580 not being listed

[WizardlyBump17]

https://github.com/intel/xpumanager/releases/tag/v2.1.0 does not recognize my B580.

davi@davi:/tmp$ sudo dpkg --install --ignore-depends=libigsc1 xpu-smi_2.0.1-1.24.04_amd64.deb 
Selecting previously unselected package xpu-smi.
(Reading database ... 376905 files and directories currently installed.)
Preparing to unpack xpu-smi_2.0.1-1.24.04_amd64.deb ...
Unpacking xpu-smi (2.0.1-1~24.04) ...
Setting up xpu-smi (2.0.1-1~24.04) ...
davi@davi:/tmp$ xpu-smi
+----------------------------------------------+----------------------------+------------------------+
|            Intel XPU-SMI v2.0    Driver: 1CF0B7A60BF9DB210B10E48    Level Zero: 1.28.0             |
+----------------------------------------------+----------------------------+------------------------+
| GPU  Name                    Persistence-M   | Bus-Id            Disp.A   |   Volatile Uncorr. ECC |
| Fan              Temp  Pwr:Usage/Cap         | Memory-Usage               |   GPU-Util  Compute M. |
+----------------------------------------------+----------------------------+------------------------+
|   0                          Off             | 0000:09:00.0      Off      |               Disabled |
| 61%               N/A  34W / 190W            | 24MiB / 12216MiB           |      N/A       Default |
+----------------------------------------------+----------------------------+------------------------+

+-------+-------+--------+----------------+----------------------------------------------------------+
| Processes:                                                                                         |
|   GPU     PID    Type    Process Name                                             GPU Memory Usage |
+-------+-------+--------+----------------+----------------------------------------------------------+
|                                  No running GPU processes found.                                   |
+-------+-------+--------+----------------+----------------------------------------------------------+

davi@davi:/tmp$ xpu-smi discovery
[Error] Failed to get device properties: 0x78000020 (ZE_RESULT_ERROR_SURVIVABILITY_MODE_DETECTED)
[Error] Failed to get device properties: 0x78000020 (ZE_RESULT_ERROR_SURVIVABILITY_MODE_DETECTED)
[Error] Failed to get device properties: 0x78000020 (ZE_RESULT_ERROR_SURVIVABILITY_MODE_DETECTED)
[Error] Failed to get device properties: 0x78000020 (ZE_RESULT_ERROR_SURVIVABILITY_MODE_DETECTED)
                                                                    
  Device ID     Device Information                                  
  0             Device Name:                                        
                Device State: survivability mode (firmware upda...  
                Vendor Name: Intel(R) Corporation                   
                SOC UUID: 0000:09:00.0                              
                PCI BDF Address: 0000:09:00.0                       
                DRM Device: /dev/dri/card1                          
                Function Type: physical                             
                                                                    
davi@davi:/tmp$ sudo apt remove xpu-smi --yes
REMOVING:                       
  xpu-smi

Summary:
  Upgrading: 0, Installing: 0, Removing: 1, Not Upgrading: 1
  Freed space: 3,589 kB

(Reading database ... 376916 files and directories currently installed.)
Removing xpu-smi (2.0.1-1~24.04) ...
davi@davi:/tmp$ sudo dpkg --install --ignore-depends=libigsc1 xpu-smi_2.1.0+26.33.6468cec-1.24.04_amd64.deb
Selecting previously unselected package xpu-smi.
(Reading database ... 376905 files and directories currently installed.)
Preparing to unpack xpu-smi_2.1.0+26.33.6468cec-1.24.04_amd64.deb ...
Unpacking xpu-smi (2.1.0+26.33.6468cec-1~24.04) ...
Setting up xpu-smi (2.1.0+26.33.6468cec-1~24.04) ...
davi@davi:/tmp$ xpu-smi
No GPU devices found.
davi@davi:/tmp$ xpu-smi discovery
                                                                  
  Device ID     Device Information                                
  0             Device State: survivability mode                  
                Recovery Action: Firmware update and GPU reset    
                PCI BDF Address: 0000:09:00.0

--ignore-depends because I am using libigsc1 built from source and for some reason the package name is different.

As you can see, the previous release works fine. I have compute-runtime installed from https://github.com/intel/compute-runtime/releases/tag/26.27.39122.11

[savery42]

Weird.. seemingly related to #151 r.e. false survivability mode which likely causes further issues within xpu-smi

Will triage

[savery42]

@WizardlyBump17 would you mind sharing the output of

1. sudo dmesg | grep -iE "xe|gsc|guc|huc|surviv|firmware"
2. xpu-smi discovery -j -d 0 (w/ 2.1.0)
3. uname -r
4. dpkg -l libze-intel-gpu1 libze1 intel-opencl-icd

[savery42]

Nevermind for now. I am able to reproduce on a fresh system with the following steps:

sudo add-apt-repository -y ppa:kobuk-team/intel-graphics
sudo apt-get update

sudo apt-get install -y \
  libze1=1.28.6-1~24.04~ppa1 \
  libigsc1

BASE_CR=https://github.com/intel/compute-runtime/releases/download/26.27.39122.11
BASE_IGC=https://github.com/intel/intel-graphics-compiler/releases/download/v2.38.2

wget $BASE_CR/libigdgmm12_22.10.0_amd64.deb
wget $BASE_CR/libze-intel-gpu1_26.27.39122.11-0_amd64.deb
wget $BASE_IGC/intel-igc-core-2_2.38.2+22051_amd64.deb

sudo dpkg -i \
  libigdgmm12_22.10.0_amd64.deb \
  intel-igc-core-2_2.38.2+22051_amd64.deb \
  libze-intel-gpu1_26.27.39122.11-0_amd64.deb

wget https://github.com/intel/xpumanager/releases/download/v2.1.0/xpu-smi_2.1.0+26.33.6468cec-1.24.04_amd64.deb
sudo dpkg -i xpu-smi_2.1.0+26.33.6468cec-1.24.04_amd64.deb

sudo modprobe xe
xpu-smi
xpu-smi discovery

I will continue to triage independently.

This is on the default kernel 6.8.0-137-generic

[WizardlyBump17]

@WizardlyBump17 would you mind sharing the output of

1. sudo dmesg | grep -iE "xe|gsc|guc|huc|surviv|firmware"
2. xpu-smi discovery -j -d 0 (w/ 2.1.0)
3. uname -r
4. dpkg -l libze-intel-gpu1 libze1 intel-opencl-icd

davi@davi:/tmp$ sudo dmesg | grep -iE "xe|gsc|guc|huc|surviv|firmware"
[    0.152524] Spectre V2 : Enabling Restricted Speculation for firmware calls
[    0.286279] ... fixed-purpose counters:    0
[    0.286279] ... fixed-purpose bitmap:      0000000000000000
[    0.373044] ACPI: [Firmware Bug]: BIOS _OSI(Linux) query ignored
[    0.419519] acpi PNP0A08:00: [Firmware Info]: ECAM [mem 0xf0000000-0xf7ffffff] for domain 0000 [bus 00-7f] only partially covers this bridge
[    0.420765] pci 0000:00:01.2:   bridge window [io  0xe000-0xffff]
[    0.424425] pci 0000:01:00.2:   bridge window [io  0xe000-0xffff]
[    0.426076] pci 0000:02:09.0:   bridge window [io  0xe000-0xefff]
[    0.428913] pci 0000:06:00.0: BAR 0 [io  0xe000-0xe0ff]
[    0.459756] pci 0000:02:09.0:   bridge window [io  0xe000-0xefff]
[    0.459778] pci 0000:01:00.2:   bridge window [io  0xe000-0xffff]
[    0.459800] pci 0000:00:01.2:   bridge window [io  0xe000-0xffff]
[    0.459929] pci_bus 0000:01: resource 0 [io  0xe000-0xffff]
[    0.459935] pci_bus 0000:02: resource 0 [io  0xe000-0xffff]
[    0.459949] pci_bus 0000:06: resource 0 [io  0xe000-0xefff]
[    0.838093] evm: security.SMACK64EXEC
[    4.884229] xe 0000:09:00.0: [drm] Found battlemage (device ID e20b) discrete display version 14.01 stepping B0
[    4.885523] xe 0000:09:00.0: [drm] VISIBLE VRAM: 0x0000007800000000, 0x0000000400000000
[    4.885546] xe 0000:09:00.0: [drm] VRAM[0]: Actual physical size 0x0000000300000000, usable size exclude stolen 0x00000002fb800000, CPU accessible size 0x00000002fb800000
[    4.885548] xe 0000:09:00.0: [drm] VRAM[0]: DPA range: [0x0000000000000000-300000000], io range: [0x0000007800000000-7afb800000]
[    4.885551] xe 0000:09:00.0: [drm] VRAM[0]: Actual physical size 0x0000000300000000, usable size exclude stolen 0x00000002fb800000, CPU accessible size 0x00000002fb800000
[    4.885553] xe 0000:09:00.0: [drm] VRAM[0]: DPA range: [0x0000000000000000-300000000], io range: [0x0000007800000000-7afb800000]
[    4.911143] xe 0000:09:00.0: [drm] Finished loading DMC firmware i915/bmg_dmc.bin (v2.6)
[    4.913292] xe 0000:09:00.0: [drm] Tile0: GT0: Using GuC firmware from xe/bmg_guc_70.bin version 70.65.0
[    5.204551] xe 0000:09:00.0: [drm] Tile0: GT0: ccs2 fused off
[    5.204554] xe 0000:09:00.0: [drm] Tile0: GT0: ccs3 fused off
[    5.231371] xe 0000:09:00.0: [drm] Tile0: GT1: Using GuC firmware from xe/bmg_guc_70.bin version 70.65.0
[    5.233305] xe 0000:09:00.0: [drm] Tile0: GT1: Using HuC firmware from xe/bmg_huc.bin version 8.2.10
[    5.245090] xe 0000:09:00.0: [drm] Tile0: GT1: vcs1 fused off
[    5.245093] xe 0000:09:00.0: [drm] Tile0: GT1: vcs3 fused off
[    5.245096] xe 0000:09:00.0: [drm] Tile0: GT1: vcs4 fused off
[    5.245098] xe 0000:09:00.0: [drm] Tile0: GT1: vcs5 fused off
[    5.245100] xe 0000:09:00.0: [drm] Tile0: GT1: vcs6 fused off
[    5.245102] xe 0000:09:00.0: [drm] Tile0: GT1: vcs7 fused off
[    5.245104] xe 0000:09:00.0: [drm] Tile0: GT1: vecs2 fused off
[    5.245106] xe 0000:09:00.0: [drm] Tile0: GT1: vecs3 fused off
[    5.275467] xe 0000:09:00.0: [drm] Registered 4 planes with drm panic
[    5.275470] [drm] Initialized xe 1.1.0 for 0000:09:00.0 on minor 1
[    5.303810] xe 0000:09:00.0: [drm] Cannot find any crtc or sizes
[    5.304613] xe 0000:09:00.0: [drm] Using mailbox commands for power limits
[    5.304937] xe 0000:09:00.0: [drm] Number of fans detected: 3
[    5.306366] xe 0000:09:00.0: [drm] PL2 is supported on channel 0
[    5.361062] xe 0000:09:00.0: [drm] Cannot find any crtc or sizes
[    5.389306] xe 0000:09:00.0: [drm] Cannot find any crtc or sizes
[    8.191483] systemd[1]: Set up automount proc-sys-fs-binfmt_misc.automount - Arbitrary Executable File Formats File System Automount Point.
[    8.207019] systemd[1]: systemd-hibernate-clear.service - Clear Stale Hibernate Storage Info was skipped because of an unmet condition check (ConditionPathExists=/sys/firmware/efi/efivars/HibernateLocation-8cf2644b-4b0b-428f-9387-6d876050dc67).
[    8.593067] Creating 4 MTD partitions on "xe.nvm.2304":
[    8.593073] 0x000000000000-0x000000001000 : "xe.nvm.2304.DESCRIPTOR"
[    8.602707] 0x000000001000-0x00000054e000 : "xe.nvm.2304.GSC"
[    8.610556] 0x00000054e000-0x00000074e000 : "xe.nvm.2304.OptionROM"
[    8.617216] 0x00000074e000-0x00000075e000 : "xe.nvm.2304.DAM"
[    8.631444] snd_hda_intel 0000:0a:00.0: bound 0000:09:00.0 (ops intel_audio_component_bind_ops [xe])
[    8.677163] RAPL PMU: API unit is 2^-32 Joules, 2 fixed counters, 163840 ms ovfl timer
[    8.786447] mei_lb xe.mei-gscfi.2304-e2c2afa2-3817-4d19-9d95-06b16b588a5d: bound 0000:09:00.0 (ops xe_late_bind_component_ops [xe])
[ 7961.105946] audit: type=1400 audit(1786644000.783:297): apparmor="DENIED" operation="open" class="file" profile="snap.firmware-updater.firmware-notifier" name="/proc/sys/vm/max_map_count" pid=143967 comm="firmware-notifi" requested_mask="r" denied_mask="r" fsuid=1000 ouid=0
[ 7986.794847] audit: type=1400 audit(1786644026.471:298): apparmor="DENIED" operation="open" class="file" profile="snap.firmware-updater.firmware-notifier" name="/proc/sys/vm/max_map_count" pid=144975 comm="firmware-notifi" requested_mask="r" denied_mask="r" fsuid=1000 ouid=0
----- MORE OF THAT FOR HUNDRED OF LINES ------
davi@davi:/tmp$ xpu-smi discovery -j -d 0
[Error] Device not found: '0'
{
    "device_list": [
        {
            "device_id": 0,
            "device_state": "survivability mode",
            "recovery_action": "Firmware update and GPU reset",
            "pci_bdf_address": "0000:09:00.0"
        }
    ]
}
davi@davi:/tmp$ uname -r 
7.2.0-rc4
davi@davi:/tmp$ dpkg -l libze-intel-gpu1 libze1 intel-opencl-icd
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend
|/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad)
||/ Name             Version          Architecture Description
+++-================-================-============-========================================================
ii  intel-opencl-icd 26.27.39122.11-0 amd64        Intel graphics compute runtime for OpenCL
ii  libze-intel-gpu1 26.27.39122.11-0 amd64        Intel(R) Graphics Compute Runtime for oneAPI Level Zero.
ii  libze1           1.32.0           amd64        oneAPI Level Zero

Forgot to mention, but I am using 7.2.0 rc4 with https://patchwork.freedesktop.org/series/168027/ rev 1, but since you were able to reproduce it on another kernel, the kernel version seems to be not important

[savery42]

Yeah, it is purely an xpu-smi issue introduced in e40b180 (with some of the other odd results being caused by the naive survivability mode detection)

Working on a fix and better regression testing to run on the repro env I listed above (albeit on 7.0.0-28-generic / linux-image-generic-hwe-24.04 as that has full BMG support). Should have better testing in place prior to the next large release, but should be able to get a patch release out today or Monday Aug 17 (pending reviews etc)

I'll provide an rc build here once I finish up with a fix + validation on a few hardware configs.

A basic fix looks to be working on a dGPU+iGPU machine with xpu-smi [ | discovery ], so I'm hopeful that I didn't miss anything else.

user@server:~$ sudo dpkg --install --ignore-depends=libigsc1 xpu-smi_2.1.0+26.33.6468cec-1.24.04_amd64.deb
(Reading database ... 148079 files and directories currently installed.)
Preparing to unpack xpu-smi_2.1.0+26.33.6468cec-1.24.04_amd64.deb ...
Unpacking xpu-smi (2.1.0+26.33.6468cec-1~24.04) over (2.0.1-1~24.04) ...
Setting up xpu-smi (2.1.0+26.33.6468cec-1~24.04) ...
user@server:~$ xpu-smi
No GPU devices found.
user@server:~$ xpu-smi discovery

  Device ID     Device Information
  0             Device State: survivability mode
                Recovery Action: Firmware update and GPU reset
                PCI BDF Address: 0000:03:00.0

  1             Device State: survivability mode
                Recovery Action: Firmware update and GPU reset
                PCI BDF Address: 0000:00:02.0

user@server:~$ ./savery/builddir-158/ial/cli/xpu-smi
+----------------------------------------------+----------------------------+------------------------+
|            Intel XPU-SMI v2.1    Driver: 85B7CA089405934276CBAD3    Level Zero: 1.28.6             |
+----------------------------------------------+----------------------------+------------------------+
| GPU  Name                    Persistence-M   | Bus-Id            Disp.A   |   Volatile Uncorr. ECC |
| Fan              Temp  Pwr:Usage/Cap         | Memory-Usage               |   GPU-Util  Compute M. |
+----------------------------------------------+----------------------------+------------------------+
|   0  Intel(R) Arc(TM) B570   Off             | 0000:03:00.0      Off      |               Disabled |
|      Graphics                                |                            |                        |
| 0%                N/A  41W / 190W            | 32MiB / 10172MiB           |      N/A       Default |
+----------------------------------------------+----------------------------+------------------------+
|   1  Intel(R) UHD Graphics   Off             | 0000:00:02.0      Off      |               Disabled |
|      730                                     |                            |                        |
| N/A               N/A  N/A                   | 2975MiB / 30078MiB         |      N/A       Default |
+----------------------------------------------+----------------------------+------------------------+

+-------+-------+--------+----------------+----------------------------------------------------------+
| Processes:                                                                                         |
|   GPU     PID    Type    Process Name                                             GPU Memory Usage |
+-------+-------+--------+----------------+----------------------------------------------------------+
|                                  No running GPU processes found.                                   |
+-------+-------+--------+----------------+----------------------------------------------------------+

user@server:~$ ./savery/builddir-158/ial/cli/xpu-smi discovery

  Device ID     Device Information
  0             Device Name: Intel(R) Arc(TM) B570 Graphics
                Device State: normal
                Vendor Name: Intel(R) Corporation
                SOC UUID: 00000000-0000-0003-0000-0000e20c8086
                PCI BDF Address: 0000:03:00.0
                DRM Device: /dev/dri/card0
                Function Type: physical

  1             Device Name: Intel(R) UHD Graphics 730
                Device State: normal
                Vendor Name: Intel(R) Corporation
                SOC UUID: 00000000-0000-0200-0000-000c46828086
                PCI BDF Address: 0000:00:02.0
                DRM Device: /dev/dri/card1
                Function Type: physical

[savery42]

Ok, with rc patch:

user@server:~$  sudo dpkg --install --ignore-depends=libigsc1 xpu-smi_2.1.1~rc1-1~24.04_amd64.deb
dpkg: warning: downgrading xpu-smi from 2.1.1-1~24.04 to 2.1.1~rc1-1~24.04
(Reading database ... 148099 files and directories currently installed.)
Preparing to unpack xpu-smi_2.1.1~rc1-1~24.04_amd64.deb ...
Unpacking xpu-smi (2.1.1~rc1-1~24.04) over (2.1.1-1~24.04) ...
Setting up xpu-smi (2.1.1~rc1-1~24.04) ...
user@server:~$ xpu-smi
+----------------------------------------------+----------------------------+------------------------+
|            Intel XPU-SMI v2.1    Driver: 85B7CA089405934276CBAD3    Level Zero: 1.28.6             |
+----------------------------------------------+----------------------------+------------------------+
| GPU  Name                    Persistence-M   | Bus-Id            Disp.A   |   Volatile Uncorr. ECC |
| Fan              Temp  Pwr:Usage/Cap         | Memory-Usage               |   GPU-Util  Compute M. |
+----------------------------------------------+----------------------------+------------------------+
|   0  Intel(R) Arc(TM) B570   Off             | 0000:03:00.0      Off      |               Disabled |
|      Graphics                                |                            |                        |
| 0%                N/A  41W / 190W            | 32MiB / 10172MiB           |      N/A       Default |
+----------------------------------------------+----------------------------+------------------------+
|   1  Intel(R) UHD Graphics   Off             | 0000:00:02.0      Off      |               Disabled |
|      730                                     |                            |                        |
| N/A               N/A  N/A                   | 3473MiB / 30077MiB         |      N/A       Default |
+----------------------------------------------+----------------------------+------------------------+

+-------+-------+--------+----------------+----------------------------------------------------------+
| Processes:                                                                                         |
|   GPU     PID    Type    Process Name                                             GPU Memory Usage |
+-------+-------+--------+----------------+----------------------------------------------------------+
|                                  No running GPU processes found.                                   |
+-------+-------+--------+----------------+----------------------------------------------------------+

user@server:~$ xpu-smi discovery

  Device ID     Device Information
  0             Device Name: Intel(R) Arc(TM) B570 Graphics
                Device State: normal
                Vendor Name: Intel(R) Corporation
                SOC UUID: 00000000-0000-0003-0000-0000e20c8086
                PCI BDF Address: 0000:03:00.0
                DRM Device: /dev/dri/card0
                Function Type: physical

  1             Device Name: Intel(R) UHD Graphics 730
                Device State: normal
                Vendor Name: Intel(R) Corporation
                SOC UUID: 00000000-0000-0200-0000-000c46828086
                PCI BDF Address: 0000:00:02.0
                DRM Device: /dev/dri/card1
                Function Type: physical

user@server:~$ xpu-smi -v
CLI:
  Version: 2.1.1.20250225
  Build ID: 8389eee7

Service:
  Version: 2.1.1.20250225
  Build ID: 8389eee7
  Level Zero Version: 1.28.6

Pre-release: xpu-smi_2.1.1~rc1-1~24.04_amd64.deb.zip

> sha256sum xpu-smi*
5736285dfd226a3912e60015bd901fc9aa21a42d4cddb20e74b9d9ae60765d59  xpu-smi_2.1.1~rc1-1~24.04_amd64.deb
2a286164432dcc43197940297ffb307652a367f057ce2dedf5f8fefb8bb6d375  xpu-smi_2.1.1~rc1-1~24.04_amd64.deb.zip

Should land officially next week

[savery42]

FYI I did add an env var XPU_SMI_LOG_LEVEL in that prerelease to enable logging for any remaining issues

Allows for:

• `debug'/'dbg'
• 'error'/'err'
• 'info'
• 'trace'

They're fairly unreadable but helpful. Might change to a CLI flag in the official release, not sure.