[SOLVED] kvm_nx_huge_page_recovery_worker message in log...

[ianmbetts]

me too, clean install of pve 8 today, and seeing the same error.

I am on non-subscription repo, not too keen on building kernel, and reading this thread it seems noexec=off is not a viable
interim workaround ?

FYI I have hyper converged setup 3 identical nodes (dl380p gen8), all with GPU passthrough, but only seen this error on one of them so far.

 

[AncientBriton]

Me too. Anyone know when we might get a fix?

 

[AddiCgn]

apt showed me a new kernel today and I think the above mentioned fix was backported. But I'm not sure if this is the right kernel.

 

[AncientBriton]

I'm running 6.2.16-4-pve which is the latest I can see in pve-no-subscription and I'm still having the issue:

Jul 19 09:40:05 pve1 kernel: ------------[ cut here ]------------
Jul 19 09:40:05 pve1 kernel: WARNING: CPU: 5 PID: 53705 at arch/x86/kvm/mmu/mmu.c:6949 kvm_nx_huge_page_recovery_worker+0x3c4/0x410 [kvm]
Jul 19 09:40:05 pve1 kernel: Modules linked in: tcp_diag inet_diag veth ebtable_filter ebtables ip_set ip6table_raw iptable_raw ip6table_filter ip6_tables iptable_filter bpfilter scsi_transport_iscsi nf_tables softdog sunrpc binfmt_misc bonding tls nfnetlink_log nfnetlink snd_hda_codec_hdmi snd_ctl_led snd_hda_codec_realtek snd_hda_codec_generic intel_rapl_msr intel_rapl_common intel_tcc_cooling snd_soc_avs x86_pkg_temp_thermal intel_powerclamp snd_soc_hda_codec coretemp snd_hda_ext_core snd_soc_core kvm_intel snd_compress mei_hdcp mei_pxp ac97_bus snd_pcm_dmaengine kvm i915 crct10dif_pclmul polyval_clmulni polyval_generic ghash_clmulni_intel sha512_ssse3 aesni_intel crypto_simd snd_hda_intel snd_intel_dspcfg snd_intel_sdw_acpi drm_buddy snd_hda_codec ttm dell_wmi drm_display_helper ledtrig_audio cryptd snd_hda_core cec rc_core dell_smbios snd_hwdep dell_wmi_aio dcdbas rapl snd_pcm sparse_keymap dell_wmi_descriptor intel_wmi_thunderbolt intel_cstate drm_kms_helper snd_timer wmi_bmof i2c_algo_bit pcspkr
Jul 19 09:40:05 pve1 kernel: snd syscopyarea mei_me soundcore sysfillrect ee1004 mei sysimgblt usblp acpi_pad joydev mac_hid vhost_net vhost vhost_iotlb tap vfio_pci vfio_pci_core irqbypass vfio_iommu_type1 vfio iommufd parport_pc ppdev lp parport drm efi_pstore dmi_sysfs ip_tables x_tables autofs4 zfs(PO) zunicode(PO) zzstd(O) zlua(O) zavl(PO) icp(PO) zcommon(PO) znvpair(PO) spl(O) hid_generic usbhid hid uas usb_storage btrfs blake2b_generic xor raid6_pq libcrc32c simplefb nvme xhci_pci crc32_pclmul xhci_pci_renesas nvme_core intel_lpss_pci i2c_i801 intel_lpss e1000e nvme_common i2c_smbus ahci xhci_hcd libahci idma64 bnx2 video wmi pinctrl_sunrisepoint
Jul 19 09:40:05 pve1 kernel: CPU: 5 PID: 53705 Comm: kvm-nx-lpage-re Tainted: P O 6.2.16-4-pve #1
Jul 19 09:40:05 pve1 kernel: Hardware name: Dell Inc. OptiPlex 7050/0NW6H5, BIOS 1.21.0 07/12/2022
Jul 19 09:40:05 pve1 kernel: RIP: 0010:kvm_nx_huge_page_recovery_worker+0x3c4/0x410 [kvm]
Jul 19 09:40:05 pve1 kernel: Code: ff 48 8b 45 c0 4c 39 e0 0f 85 e6 fd ff ff 48 89 df e8 20 e7 f9 ff e9 ed fd ff ff 49 bc ff ff ff ff ff ff ff 7f e9 c6 fc ff ff <0f> 0b e9 01 ff ff ff 48 8b 45 d0 65 48 2b 04 25 28 00 00 00 75 27
Jul 19 09:40:05 pve1 kernel: RSP: 0018:ffffb106b4b3be40 EFLAGS: 00010246
Jul 19 09:40:05 pve1 kernel: RAX: 0000000000000000 RBX: ffffb106b4b2d000 RCX: 0000000000000000
Jul 19 09:40:05 pve1 kernel: RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000000
Jul 19 09:40:05 pve1 kernel: RBP: ffffb106b4b3bec0 R08: 0000000000000000 R09: 0000000000000000
Jul 19 09:40:05 pve1 kernel: R10: ffff8b5ab239f7e8 R11: 0000000000000000 R12: ffffb106b4b3be80
Jul 19 09:40:05 pve1 kernel: R13: 0000000000000001 R14: 0000000000000007 R15: ffff8b5ab239f878
Jul 19 09:40:05 pve1 kernel: FS: 0000000000000000(0000) GS:ffff8b63cdf40000(0000) knlGS:0000000000000000
Jul 19 09:40:05 pve1 kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Jul 19 09:40:05 pve1 kernel: CR2: 00000260138ba200 CR3: 00000002ee80e002 CR4: 00000000003726e0
Jul 19 09:40:05 pve1 kernel: Call Trace:
Jul 19 09:40:05 pve1 kernel: <TASK>
Jul 19 09:40:05 pve1 kernel: ? __pfx_kvm_nx_huge_page_recovery_worker+0x10/0x10 [kvm]
Jul 19 09:40:05 pve1 kernel: kvm_vm_worker_thread+0x9d/0x1b0 [kvm]
Jul 19 09:40:05 pve1 kernel: ? __pfx_kvm_vm_worker_thread+0x10/0x10 [kvm]
Jul 19 09:40:05 pve1 kernel: kthread+0xe6/0x110
Jul 19 09:40:05 pve1 kernel: ? __pfx_kthread+0x10/0x10
Jul 19 09:40:05 pve1 kernel: ret_from_fork+0x29/0x50
Jul 19 09:40:05 pve1 kernel: </TASK>
Jul 19 09:40:05 pve1 kernel: ---[ end trace 0000000000000000 ]---

 

[AncientBriton]

Looking at the git repository, it looks like the fix was added 4 days ago, but I don't see any release following that. 6.2.16-4-pve was released 12 days ago.

https://git.proxmox.com/?p=pve-kernel.git;a=shortlog;h=refs/heads/master

 

[senkis]

Linux pve 6.2.16-4-pve #1 SMP PREEMPT_DYNAMIC PVE 6.2.16-5 (2023-07-14T17:53Z) x86_64 GNU/Linux

issue still present.

 

[juliokele]

Linux pve 6.2.16-4-pve #1 SMP PREEMPT_DYNAMIC PVE 6.2.16-5 (2023-07-14T17:53Z) x86_64 GNU/Linux

issue still present.

2023-07-17 10:33:31 CEST

Fix will be included with next kernel bump (at max a few weeks).

https://bugzilla.proxmox.com/show_bug.cgi?id=4833#c2

 

[senkis]

Linux pve 6.2.16-5-pve #1 SMP PREEMPT_DYNAMIC PVE 6.2.16-6 (2023-07-25T15:33Z) x86_64 GNU/Linux

7 days up, error seems to be not present

 

[AddiCgn]

Same here. Tried to mark this thread as "solved", but it seems I dont have the option to edit post #1 to edit the subject line (?).

 

[fiona]

Hi,

Same here. Tried to mark this thread as "solved", but it seems I dont have the option to edit post #1 to edit the subject line (?).

you should see an Edit Thread button on the right side above the first post, where you can select the SOLVED prefix.

 

[AddiCgn]

@fiona I'm only seeign this - above 1st post / right side?

 

[fiona]

@fiona I'm only seeign this - above 1st post / right side?
View attachment 53927

Hmm, maybe there's a limit until when it can be edited by the original poster. I selected the prefix now.