DPDK patches and discussions
 help / color / mirror / Atom feed
From: "Burakov, Anatoly" <anatoly.burakov@intel.com>
To: Ravi Kerur <rkerur@gmail.com>
Cc: dev@dpdk.org
Subject: Re: [dpdk-dev] IXGBE, IOMMU DMAR DRHD handling fault issue
Date: Wed, 31 Jan 2018 09:59:26 +0000	[thread overview]
Message-ID: <10911b54-57ee-370b-a4f7-f34accf4811e@intel.com> (raw)
In-Reply-To: <CAFb4SLAOTDnz0bK=9D-dLQJ+x6rVOqMH5zDaP_nAoEXLUByWRQ@mail.gmail.com>

On 29-Jan-18 10:35 PM, Ravi Kerur wrote:
> Hi Burakov,
> 
> When using vfio-pci on host both VF and PF interfaces works fine with 
> dpdk i.e. I don't see DMAR fault messages anymore. However, when I 
> attach a VF interface to a VM and start DPDK with vfio-pci inside VM I 
> still see DMAR fault messages on host. Both host and VM are booted with 
> 'intel-iommu=on' on GRUB. Ping from VM with DPDK/vfio-pci doesn't work 
> (I think it's expected because of DMAR faults), however, when VF 
> interface uses ixgbevf driver ping works.
> 
> Following are some details
> 
> /*****************On VM***************/
> dpdk-devbind -s
> 
> Network devices using DPDK-compatible driver
> ============================================
> 0000:00:07.0 '82599 Ethernet Controller Virtual Function' drv=vfio-pci 
> unused=ixgbevf
> 
> Network devices using kernel driver
> ===================================
> 0000:03:00.0 'Device 1041' if=eth0 drv=virtio-pci unused=vfio-pci *Active*
> 0000:04:00.0 'Device 1041' if=eth1 drv=virtio-pci unused=vfio-pci
> 0000:05:00.0 'Device 1041' if=eth2 drv=virtio-pci unused=vfio-pci
> 
> Other network devices
> =====================
> <none>
> 
> Crypto devices using DPDK-compatible driver
> ===========================================
> <none>
> 
> Crypto devices using kernel driver
> ==================================
> <none>
> 
> Other crypto devices
> ====================
> <none>
> 
> 
> 00:07.0 Ethernet controller: Intel Corporation 82599 Ethernet Controller 
> Virtual Function (rev 01)
>          Subsystem: Intel Corporation 82599 Ethernet Controller Virtual 
> Function
>          Control: I/O- Mem+ BusMaster- SpecCycle- MemWINV- VGASnoop- 
> ParErr- Stepping- SERR- FastB2B- DisINTx+
>          Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- 
> <TAbort- <MAbort- >SERR- <PERR- INTx-
>          Region 0: Memory at fda00000 (64-bit, prefetchable) [size=16K]
>          Region 3: Memory at fda04000 (64-bit, prefetchable) [size=16K]
>          Capabilities: [70] MSI-X: Enable+ Count=3 Masked-
>                  Vector table: BAR=3 offset=00000000
>                  PBA: BAR=3 offset=00002000
>          Capabilities: [a0] Express (v1) Root Complex Integrated 
> Endpoint, MSI 00
>                  DevCap: MaxPayload 128 bytes, PhantFunc 0
>                          ExtTag- RBE-
>                  DevCtl: Report errors: Correctable- Non-Fatal- Fatal- 
> Unsupported-
>                          RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop-
>                          MaxPayload 128 bytes, MaxReadReq 128 bytes
>                  DevSta: CorrErr- UncorrErr- FatalErr- UnsuppReq- 
> AuxPwr- TransPend-
>          Capabilities: [100 v1] Advanced Error Reporting
>                  UESta:  DLP- SDES- TLP- FCP- CmpltTO- CmpltAbrt- 
> UnxCmplt- RxOF- MalfTLP- ECRC- UnsupReq- ACSViol-
>                  UEMsk:  DLP- SDES- TLP- FCP- CmpltTO- CmpltAbrt- 
> UnxCmplt- RxOF- MalfTLP- ECRC- UnsupReq- ACSViol-
>                  UESvrt: DLP- SDES- TLP- FCP- CmpltTO- CmpltAbrt- 
> UnxCmplt- RxOF- MalfTLP- ECRC- UnsupReq- ACSViol-
>                  CESta:  RxErr- BadTLP- BadDLLP- Rollover- Timeout- 
> NonFatalErr-
>                  CEMsk:  RxErr- BadTLP- BadDLLP- Rollover- Timeout- 
> NonFatalErr-
>                  AERCap: First Error Pointer: 00, GenCap- CGenEn- 
> ChkCap- ChkEn-
>          Kernel driver in use: vfio-pci
>          Kernel modules: ixgbevf
> 
> /***************on Host*************/
> dmesg | grep DMAR
> ...
> [  978.268143] DMAR: DRHD: handling fault status reg 2
> [  978.268147] DMAR: [DMA Read] *Request device [04:10.0]* fault addr 
> 33a128000 [fault reason 06] PTE Read access is not set
> [ 1286.677726] DMAR: DRHD: handling fault status reg 102
> [ 1286.677730] DMAR: [DMA Read] Request device [04:10.0] fault addr 
> fb663000 [fault reason 06] PTE Read access is not set
> [ 1676.436145] DMAR: DRHD: handling fault status reg 202
> [ 1676.436149] DMAR: [DMA Read] Request device [04:10.0] fault addr 
> 33a128000 [fault reason 06] PTE Read access is not set
> [ 1734.433649] DMAR: DRHD: handling fault status reg 302
> [ 1734.433652] DMAR: [DMA Read] Request device [04:10.0] fault addr 
> 33a128000 [fault reason 06] PTE Read access is not set
> [ 2324.428938] DMAR: DRHD: handling fault status reg 402
> [ 2324.428942] DMAR: [DMA Read] Request device [04:10.0] fault addr 
> 7770c000 [fault reason 06] PTE Read access is not set
> [ 2388.553640] DMAR: DRHD: handling fault status reg 502
> [ 2388.553643] DMAR: [DMA Read] *Request device [04:10.0]* fault addr 
> 33a128000 [fault reason 06] PTE Read access is not set
> 
> 
> VM is started with
> 
> qemu-system-x86_64 -enable-kvm -M q35,accel=kvm,kernel-irqchip=split 
> -object iothread,id=iothread0 -device 
> intel-iommu,intremap=on,device-iotlb=on,caching-mode=on -cpu host 
> -daemonize -m 16G -smp 14 -uuid 0fc91c66-f0b1-11e7-acf4-525400123456 
> -name 212748-sriov-ravi-smac-alpha-SMAC10 -device 
> ioh3420,id=root.1,chassis=1 -device ioh3420,id=root.2,chassis=2 -netdev 
> tap,vhost=on,queues=2,ifname=vn-vn2_1_,downscript=no,id=vn-vn2_1_,script=no 
> -device ioh3420,id=root.3,chassis=3 -device 
> virtio-net-pci,netdev=vn-vn2_1_,bus=root.3,ats=on,mq=on,vectors=6,mac=DE:AD:02:88:10:37,id=vn-vn2_1__dev 
> -netdev 
> tap,vhost=on,queues=2,ifname=vn-vn92_1_,downscript=no,id=vn-vn92_1_,script=no 
> -device ioh3420,id=root.4,chassis=4 -device 
> virtio-net-pci,mac=DE:AD:02:88:10:38,netdev=vn-vn92_1_,bus=root.4,ats=on,mq=on,vectors=6,id=vn-vn92_1__dev 
> -netdev 
> tap,vhost=on,queues=2,ifname=vn-vn93_1_,downscript=no,id=vn-vn93_1_,script=no 
> -device ioh3420,id=root.5,chassis=5 -device 
> virtio-net-pci,mac=DE:AD:02:88:10:39,netdev=vn-vn93_1_,bus=root.5,ats=on,mq=on,vectors=6,id=vn-vn93_1__dev 
> -vnc :16,websocket=15916 -qmp tcp:127.0.0.1:12001 
> <http://127.0.0.1:12001>,server,nowait -chardev 
> socket,id=charmonitor,path=/tmp/mon.12001,server,nowait -mon 
> chardev=charmonitor,id=monitor -cdrom 
> /var/venom/cloud_init/0fc91c66-f0b1-11e7-acf4-525400123456.iso -*device 
> vfio-pci,host=04:10.0* -drive 
> file=/var/venom/instance_repo/test.img,if=none,id=drive-virtio-disk0,format=raw,aio=native,cache=none 
> -balloon none -device 
> virtio-blk-pci,scsi=off,iothread=iothread0,drive=drive-virtio-disk0,id=virtio-disk0,bus=root.1,ats=on,bootindex=1
> 
> Thanks.
> 
> 
> On Thu, Jan 25, 2018 at 2:49 AM, Burakov, Anatoly 
> <anatoly.burakov@intel.com <mailto:anatoly.burakov@intel.com>> wrote:
> 
>     On 24-Jan-18 7:13 PM, Ravi Kerur wrote:
> 
>         Hi Burakov, Thank you. I will try with vfio-pci driver. I am
>         assuming it will work for both PF and VF interfaces since I am
>         using both in my setup?
> 
>         Thanks.
> 
> 
>     Yes, it should work for both PF and VF devices.
> 
> 
>         On Wed, Jan 24, 2018 at 2:31 AM, Burakov, Anatoly
>         <anatoly.burakov@intel.com <mailto:anatoly.burakov@intel.com>
>         <mailto:anatoly.burakov@intel.com
>         <mailto:anatoly.burakov@intel.com>>> wrote:
> 
>              On 23-Jan-18 5:25 PM, Ravi Kerur wrote:
> 
>                  Hi,
> 
>                  I am running into an issue when DPDK is started with
>         iommu on
>                  via GRUB
>                  command. Problem is not seen with regular kernel
>         driver, error
>                  messages
>                  show when DPDK is started and happens for both PF and
>         VF interfaces.
> 
>                  I am using DPDK 17.05 so the patch proposed in the
>         following link is
>                  available
>         http://dpdk.org/ml/archives/dev/2017-February/057048.html
>         <http://dpdk.org/ml/archives/dev/2017-February/057048.html>
>                 
>         <http://dpdk.org/ml/archives/dev/2017-February/057048.html
>         <http://dpdk.org/ml/archives/dev/2017-February/057048.html>>
> 
>                  Workaround is to use "iommu=pt" but I want iommu
>         enabled in my
>                  setup. I
>                  checked BIOS for reserved memory(DMA RMRR for IXGBE)
>         didn't get
>                  any details
>                  on it.
> 
>                  Kindly let me know how to resolve this issue.
> 
>                  Following are the details
> 
>                  (1) Linux kernel 4.9
>                  (2) DPDK 17.05
> 
>                  (3) IXGBE details
>                  ethtool -i enp4s0f0  (PF driver)
>                  driver: ixgbe
>                  version: 5.3.3
>                  firmware-version: 0x800007b8, 1.1018.0
>                  bus-info: 0000:04:00.0
>                  supports-statistics: yes
>                  supports-test: yes
>                  supports-eeprom-access: yes
>                  supports-register-dump: yes
>                  supports-priv-flags: yes
> 
>                  ethtool -i enp4s16f2 (VF driver)
>                  driver: ixgbevf
>                  version: 4.3.2
>                  firmware-version:
>                  bus-info: 0000:04:10.2
>                  supports-statistics: yes
>                  supports-test: yes
>                  supports-eeprom-access: no
>                  supports-register-dump: yes
>                  supports-priv-flags: no
> 
>                  Bus info          Device       Class          Description
>                  =========================================================
>                  pci@0000:01:00.0  ens11f0      network        82599ES
>         10-Gigabit
>                  SFI/SFP+
>                  Network Connection
>                  pci@0000:01:00.1  ens11f1      network        82599ES
>         10-Gigabit
>                  SFI/SFP+
>                  Network Connection
>                  pci@0000:04:00.0  enp4s0f0     network        82599ES
>         10-Gigabit
>                  SFI/SFP+
>                  Network Connection
>                  pci@0000:04:00.1  enp4s0f1     network        82599ES
>         10-Gigabit
>                  SFI/SFP+
>                  Network Connection
>                  pci@0000:04:10.0  enp4s16      network        Illegal
>         Vendor ID
>                  pci@0000:04:10.2  enp4s16f2    network        Illegal
>         Vendor ID
> 
>                  (4) DPDK bind interfaces
> 
>                  # dpdk-devbind -s
> 
>                  Network devices using DPDK-compatible driver
>                  ============================================
>                  0000:01:00.0 '82599ES 10-Gigabit SFI/SFP+ Network
>         Connection 10fb'
>                  drv=igb_uio unused=vfio-pci
>                  0000:04:10.2 '82599 Ethernet Controller Virtual
>         Function 10ed'
>                  drv=igb_uio
>                  unused=vfio-pci
> 
>                  Network devices using kernel driver
>                  ===================================
>                  0000:01:00.1 '82599ES 10-Gigabit SFI/SFP+ Network
>         Connection 10fb'
>                  if=ens11f1 drv=ixgbe unused=igb_uio,vfio-pci
>                  0000:04:00.0 '82599ES 10-Gigabit SFI/SFP+ Network
>         Connection 10fb'
>                  if=enp4s0f0 drv=ixgbe unused=igb_uio,vfio-pci
>                  0000:04:00.1 '82599ES 10-Gigabit SFI/SFP+ Network
>         Connection 10fb'
>                  if=enp4s0f1 drv=ixgbe unused=igb_uio,vfio-pci
>                  0000:04:10.0 '82599 Ethernet Controller Virtual
>         Function 10ed'
>                  if=enp4s16
>                  drv=ixgbevf unused=igb_uio,vfio-pci
>                  0000:06:00.0 'I210 Gigabit Network Connection 1533'
>         if=eno1 drv=igb
>                  unused=igb_uio,vfio-pci *Active*
> 
>                  Other Network devices
>                  =====================
>                  <none>
> 
>                  ...
> 
>                  (5) Kernel dmesg
> 
>                  # dmesg | grep -e DMAR
>                  [    0.000000] ACPI: DMAR 0x000000007999BAD0 0000E0
>         (v01 ALASKA
>                  A M I
>                  00000001 INTL 20091013)
>                  [    0.000000] DMAR: IOMMU enabled
>                  [    0.518747] DMAR: Host address width 46
>                  [    0.526616] DMAR: DRHD base: 0x000000fbffc000 flags: 0x0
>                  [    0.537447] DMAR: dmar0: reg_base_addr fbffc000 ver
>         1:0 cap
>                  d2078c106f0466 ecap f020df
>                  [    0.553620] DMAR: DRHD base: 0x000000c7ffc000 flags: 0x1
>                  [    0.564445] DMAR: dmar1: reg_base_addr c7ffc000 ver
>         1:0 cap
>                  d2078c106f0466 ecap f020df
>                  [    0.580611] DMAR: RMRR base: 0x0000007bbc6000 end:
>                  0x0000007bbd4fff
>                  [    0.593344] DMAR: ATSR flags: 0x0
>                  [    0.600178] DMAR: RHSA base: 0x000000c7ffc000 proximity
>                  domain: 0x0
>                  [    0.612905] DMAR: RHSA base: 0x000000fbffc000 proximity
>                  domain: 0x1
>                  [    0.625632] DMAR-IR: IOAPIC id 3 under DRHD base 
>         0xfbffc000
>                  IOMMU 0
>                  [    0.638522] DMAR-IR: IOAPIC id 1 under DRHD base 
>         0xc7ffc000
>                  IOMMU 1
>                  [    0.651426] DMAR-IR: IOAPIC id 2 under DRHD base 
>         0xc7ffc000
>                  IOMMU 1
>                  [    0.664324] DMAR-IR: HPET id 0 under DRHD base
>         0xc7ffc000
>                  [    0.675326] DMAR-IR: Queued invalidation will be
>         enabled to
>                  support
>                  x2apic and Intr-remapping.
>                  [    0.693805] DMAR-IR: Enabled IRQ remapping in x2apic
>         mode
>                  [    9.395170] DMAR: dmar1: Using Queued invalidation
>                  [    9.405011] DMAR: Setting RMRR:
>                  [    9.412006] DMAR: Setting identity map for device
>         0000:00:1d.0
>                  [0x7bbc6000 - 0x7bbd4fff]
>                  [    9.428569] DMAR: Prepare 0-16MiB unity mapping for LPC
>                  [    9.439712] DMAR: Setting identity map for device
>                  0000:00:1f.0 [0x0 -
>                  0xffffff]
>                  [    9.454684] DMAR: Intel(R) Virtualization Technology for
>                  Directed I/O
>                  [  287.023068] DMAR: DRHD: handling fault status reg 2
>                  [  287.023073] DMAR: [DMA Read] Request device
>         [01:00.0] fault addr
>                  18a260a000 [fault reason 06] PTE Read access is not set
>                  [  287.023180] DMAR: DRHD: handling fault status reg 102
>                  [  287.023183] DMAR: [DMA Read] Request device
>         [01:00.0] fault addr
>                  18a3010000 [fault reason 06] PTE Read access is not set
>                  [  287.038250] DMAR: DRHD: handling fault status reg 202
>                  [  287.038252] DMAR: [DMA Read] Request device
>         [01:00.0] fault addr
>                  18a3010000 [fault reason 06] PTE Read access is not set
>                  [  288.170165] DMAR: DRHD: handling fault status reg 302
>                  [  288.170170] DMAR: [DMA Read] Request device
>         [01:00.0] fault addr
>                  1890754000 [fault reason 06] PTE Read access is not set
>                  [  288.694496] DMAR: DRHD: handling fault status reg 402
>                  [  288.694499] DMAR: [DMA Read] Request device
>         [04:10.2] fault addr
>                  189069c000 [fault reason 06] PTE Read access is not set
>                  [  289.927113] DMAR: DRHD: handling fault status reg 502
>                  [  289.927116] DMAR: [DMA Read] Request device
>         [01:00.0] fault addr
>                  1890754000 [fault reason 06] PTE Read access is not set
>                  [  290.174275] DMAR: DRHD: handling fault status reg 602
>                  [  290.174279] DMAR: [DMA Read] Request device
>         [01:00.0] fault addr
>                  1890754000 [fault reason 06] PTE Read access is not set
>                  [  292.174247] DMAR: DRHD: handling fault status reg 702
>                  [  292.174251] DMAR: [DMA Read] Request device
>         [01:00.0] fault addr
>                  1890754000 [fault reason 06] PTE Read access is not set
>                  [  294.174227] DMAR: DRHD: handling fault status reg 2
>                  [  294.174230] DMAR: [DMA Read] Request device
>         [01:00.0] fault addr
>                  1890754000 [fault reason 06] PTE Read access is not set
>                  [  296.174216] DMAR: DRHD: handling fault status reg 102
>                  [  296.174219] DMAR: [DMA Read] Request device
>         [01:00.0] fault addr
>                  1890754000 [fault reason 06] PTE Read access is not set
>                  [root@infradev-comp006.naw02.infradev.viasat.io
>         <mailto:root@infradev-comp006.naw02.infradev.viasat.io>
>                  <mailto:root@infradev-comp006.naw02.infradev.viasat.io
>         <mailto:root@infradev-comp006.naw02.infradev.viasat.io>> ~]
>                  #
> 
>                  Thanks.
> 
> 
>              Hi Ravi,
> 
>              The "iommu=pt" workaround applies only when you want to use
>         igb_uio
>              driver. VFIO driver is able to fully utilize IOMMU without
>         the need
>              for pass-through mode. From your log i can see that some
>         devices are
>              bound to igb_uio while others are bound to vfio-pci. Just
>         bind all
>              of the devices you want to use with DPDK to vfio-pci and these
>              errors should go away.
> 
>              --     Thanks,
>              Anatoly
> 
> 
> 
> 
>     -- 
>     Thanks,
>     Anatoly
> 
> 

Hi Ravi,

Using vfio-pci in IOMMU mode will only work if your VM provides IOMMU 
emulation (it's a fairly recent development, so your QEMU must be of an 
appropriate version - can't recall which one off the top of my head). 
Otherwise you'd have to treat your VM as if it was a machine without 
IOMMU, i.e. use noiommu mode for VFIO, or igb_uio driver.

-- 
Thanks,
Anatoly

  reply	other threads:[~2018-01-31  9:59 UTC|newest]

Thread overview: 33+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-01-23 17:25 Ravi Kerur
2018-01-24 10:31 ` Burakov, Anatoly
2018-01-24 19:13   ` Ravi Kerur
2018-01-25 10:49     ` Burakov, Anatoly
2018-01-29 22:35       ` Ravi Kerur
2018-01-31  9:59         ` Burakov, Anatoly [this message]
2018-01-31 21:51           ` Ravi Kerur
2018-02-01 10:10             ` Burakov, Anatoly
2018-02-01 19:26               ` Ravi Kerur
2018-02-02 10:28                 ` Burakov, Anatoly
2018-02-02 20:21                   ` Ravi Kerur
2018-02-02 20:51                     ` Ravi Kerur
2018-02-05 10:01                       ` Burakov, Anatoly
2018-02-06 17:55                         ` Ravi Kerur
2018-02-08 11:20                           ` Burakov, Anatoly
2018-02-09 17:41                             ` Ravi Kerur
2018-02-10 10:11                               ` Burakov, Anatoly
2018-02-10 10:58         ` Burakov, Anatoly
2018-02-10 17:53           ` Ravi Kerur
2018-02-12 10:13             ` Burakov, Anatoly
2018-02-12 22:00               ` Ravi Kerur
2018-02-13 14:31                 ` Burakov, Anatoly
2018-02-14 20:00                   ` Ravi Kerur
2018-02-15 10:28                     ` Burakov, Anatoly
2018-02-15 18:27                       ` Ravi Kerur
2018-02-15 20:53                         ` Ravi Kerur
2018-02-16  9:41                           ` Burakov, Anatoly
2019-01-15  7:07                             ` Hu, Xuekun
2019-01-15 11:22                               ` Burakov, Anatoly
2019-01-15 13:07                                 ` Hu, Xuekun
2019-01-21 13:18                                 ` Hu, Xuekun
2019-01-21 13:39                                   ` Burakov, Anatoly
2019-01-21 14:44                                     ` Thomas Monjalon

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=10911b54-57ee-370b-a4f7-f34accf4811e@intel.com \
    --to=anatoly.burakov@intel.com \
    --cc=dev@dpdk.org \
    --cc=rkerur@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).