[DMA Read NO_PASID] Request device [00:02.0] fault addr 0xcce25000 [fault reason 0x05] PTE Write access is not set

[floppy86]

Hi I'm new to promox and virtualization.

I followed the guides that I could find for pcie passthrough and I don't seem to make it to work.
I get the bellow:


[ 0.012286] ACPI: DMAR 0x00000000C9EC9000 0000A8 (v01 INTEL SKL 00000001 INTL 00000001)
[ 0.012324] ACPI: Reserving DMAR table memory at [mem 0xc9ec9000-0xc9ec90a7]
[ 0.036337] DMAR: IOMMU enabled
[ 0.117203] DMAR: Host address width 39
[ 0.117204] DMAR: DRHD base: 0x000000fed90000 flags: 0x0
[ 0.117211] DMAR: dmar0: reg_base_addr fed90000 ver 1:0 cap 1c0000c40660462 ecap 7e3ff0505e
[ 0.117214] DMAR: DRHD base: 0x000000fed91000 flags: 0x1
[ 0.117217] DMAR: dmar1: reg_base_addr fed91000 ver 1:0 cap d2008c40660462 ecap f050da
[ 0.117219] DMAR: RMRR base: 0x000000c9bd7000 end: 0x000000c9bf6fff
[ 0.117221] DMAR: RMRR base: 0x000000cc000000 end: 0x000000ce7fffff
[ 0.117224] DMAR-IR: IOAPIC id 2 under DRHD base 0xfed91000 IOMMU 1
[ 0.117225] DMAR-IR: HPET id 0 under DRHD base 0xfed91000
[ 0.117226] DMAR-IR: Queued invalidation will be enabled to support x2apic and Intr-remapping.
[ 0.118811] DMAR-IR: Enabled IRQ remapping in x2apic mode
[ 0.369296] DMAR: No ATSR found
[ 0.369297] DMAR: No SATC found
[ 0.369298] DMAR: IOMMU feature fl1gp_support inconsistent
[ 0.369300] DMAR: IOMMU feature pgsel_inv inconsistent
[ 0.369301] DMAR: IOMMU feature nwfs inconsistent
[ 0.369302] DMAR: IOMMU feature eafs inconsistent
[ 0.369302] DMAR: IOMMU feature prs inconsistent
[ 0.369303] DMAR: IOMMU feature nest inconsistent
[ 0.369304] DMAR: IOMMU feature mts inconsistent
[ 0.369305] DMAR: IOMMU feature sc_support inconsistent
[ 0.369306] DMAR: IOMMU feature dev_iotlb_support inconsistent
[ 0.369307] DMAR: dmar0: Using Queued invalidation
[ 0.369310] DMAR: dmar1: Using Queued invalidation
[ 0.370889] DMAR: Intel(R) Virtualization Technology for Directed I/O
[ 752.444684] DMAR: DRHD: handling fault status reg 2
[ 752.444813] DMAR: [DMA Write NO_PASID] Request device [00:02.0] fault addr 0x0 [fault reason 0x02] Present bit in context entry is clear
[ 752.640834] DMAR: DRHD: handling fault status reg 3
[ 752.640961] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xcca4e000 [fault reason 0x05] PTE Write access is not set
[ 752.844881] DMAR: DRHD: handling fault status reg 3
[ 752.845012] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xcca8a000 [fault reason 0x05] PTE Write access is not set
[ 754.888796] DMAR: DRHD: handling fault status reg 3
[ 757.540754] DMAR: DRHD: handling fault status reg 3
[ 757.540865] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xccae4000 [fault reason 0x05] PTE Write access is not set
[ 757.744762] DMAR: DRHD: handling fault status reg 3
[ 757.744764] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xccb20000 [fault reason 0x05] PTE Write access is not set
[ 757.744867] DMAR: DRHD: handling fault status reg 3
[ 757.744868] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xccb0d000 [fault reason 0x05] PTE Write access is not set
[ 757.745105] DMAR: DRHD: handling fault status reg 3
[ 762.648716] DMAR: DRHD: handling fault status reg 3
[ 762.648835] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xccbb4000 [fault reason 0x05] PTE Write access is not set
[ 762.852893] DMAR: DRHD: handling fault status reg 3
[ 762.853007] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xccbf0000 [fault reason 0x05] PTE Write access is not set
[ 763.056708] DMAR: DRHD: handling fault status reg 3
[ 763.056818] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xccc2e000 [fault reason 0x05] PTE Write access is not set
[ 763.260651] DMAR: DRHD: handling fault status reg 3
[ 767.756635] DMAR: DRHD: handling fault status reg 3
[ 767.756748] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xccc74000 [fault reason 0x05] PTE Write access is not set
[ 767.960646] DMAR: DRHD: handling fault status reg 3
[ 767.960649] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xcccc0000 [fault reason 0x05] PTE Write access is not set
[ 767.960747] DMAR: DRHD: handling fault status reg 3
[ 767.960748] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xcccc2000 [fault reason 0x05] PTE Write access is not set
[ 767.961004] DMAR: DRHD: handling fault status reg 3
[ 772.876713] DMAR: DRHD: handling fault status reg 3
[ 772.876808] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xccdb4000 [fault reason 0x05] PTE Write access is not set
[ 773.080557] DMAR: DRHD: handling fault status reg 3
[ 773.080684] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xccdee000 [fault reason 0x05] PTE Write access is not set
[ 773.284521] DMAR: DRHD: handling fault status reg 3
[ 773.284633] DMAR: [DMA Read NO_PASID] Request device [00:02.0] fault addr 0xcce25000 [fault reason 0x05] PTE Write access is not set
[ 773.488554] DMAR: DRHD: handling fault status reg 3
[ 1017.793575] DMAR: DRHD: handling fault status reg 2
[ 1017.793625] DMAR: [DMA Write NO_PASID] Request device [00:02.0] fault addr 0x0 [fault reason 0x02] Present bit in context entry is clear


root@media:~# cat /etc/modprobe.d/vfio.conf
options vfio-pci ids=8086:1912

root@media:~# cat /etc/modprobe.d/iommu_unsafe_interrupts.conf
options vfio_iommu_type1 allow_unsafe_interrupts=1

root@media:~# lspci -nnv | grep VGA
00:02.0 VGA compatible controller [0300]: Intel Corporation HD Graphics 530 [8086:1912] (rev 06) (prog-if 00 [VGA controller])

root@media:~# cat /etc/modules
vfio
vfio_iommu_type1
vfio_pci
vfio_virqfd
# Modules required for Intel GVT
kvmgt
exngt
vfio-mdev

root@media:~# cat /etc/modprobe.d/blacklist.conf
blacklist snd_hda_intel
blacklist snd_hda_codec_hdmi
blacklist i915

root@media:~# cat /etc/modprobe.d/kvm.conf
options kvm ignore_msrs=1


root@media:~# cat /etc/default/grub | grep GRUB_CMDLINE_LINUX_DEFAULT
GRUB_CMDLINE_LINUX_DEFAULT="quiet intel_iommu=on video=efifbff"

 

[mirpol666]

Hi! no response and no positive news....
Seems that the issue is related @ new kernel of proxmox...in fact with 7.0 no error and perfect passtrough only with flag "intel_iommu=on", but after latest updates not luck!

 

[lxiosjao]

Hello,
did you find a solution about this ?
I have the same issue...with kernel 5.15.102

 

[fennecguy]

Heya, is there any solution to that? I'm just about to ditch proxmox because of this exact problem. Searching through the forum, there's no solution.

 

[maxiReglisse]

Same problem on my Ceph/Proxmox cluster of 5 servers, one Dell R540 and four HP DL380. The Dell server froze this morning, around 4 a.m., and in the kernel logs, kern.log, we had the same type of message, the day before the crash.

DMAR: [DMA Write NO_PASID] Request device [af:00.0] fault addr 0xd390d000 [fault reason 0x05] PTE Write access is not set

Linux kernel version: 6.8.4-3-pve
ServerID: 205F71E13CDB16F773FCB8B6A8DFABB2 (is the subscription key confidential? ... pve2c-30xxxxxxxx3f)
GRUB_CMDLINE_LINUX_DEFAULT="quiet"
No kernel module

Thank you in advance for your feedback

 

[brucexx]

Experienced it today with my 10years old R720 server taht I use for testing.

DMAR DRHD Request device [03:00.0] fault addr ... fault reason PTE read access is not set.

I had to replace failing Perc H710P mini and took laying around H310 mini, the server recognized it and did not make a fuss about swapping the 710 with 310. Everything seemed to work but the server kept not booting up , when it booted up it would work for 10-20 min and then became unresponsive which required me to reboot it.

the 03:00.0 address is MegaRaid controller in my case and that controller did not like that I had RAID1 array (2 drives). I have a total of 8 drives. 6 drives including the boot drive are JBOD and 2 drives are in RAID1. The moment I removed the 2 RAID1 drives from the chassis everything would start working. In the end I removed the RAID1 but left the 2 drives as JBOD and it's worked since then, no freezing and no issues with booting up.

I have no reason to believe that the Perc H310 is defective, I took it out from a working server. I have another Dell R430 server with 4 JBOD drives on the same 8.2.7 firmware and no issues. I also have R320 with JBOD drives with Perc H310 no issue on 8.2.7.