From: Chinmaya Dwibedy <ckdwibedy@gmail.com>
To: users@dpdk.org
Subject: [dpdk-users] Issue with Physical Function passthrough on VM Guest (Fedora 20) using qatmux.l.2.6.0-60
Date: Fri, 10 Jun 2016 12:46:37 +0530 [thread overview]
Message-ID: <CAMHS8RF_=TgPxt8KqV-oCHqDieBEHHXb=e7aos54zVCQamQGBw@mail.gmail.com> (raw)
Hi,
I have assigned Intel QAT device to VMs as a full device (PF - physical
function, PCI passthrough). Below is the output of lspci in host (Cent OS
7.2).
[root@localhost nova(keystone_admin)]# lspci -nn | grep 0435
83:00.0 Co-processor [0b40]: Intel Corporation DH895XCC Series QAT [
8086:0435]
88:00.0 Co-processor [0b40]: Intel Corporation DH895XCC Series QAT [
8086:0435]
[root@localhost nova(keystone_admin)]#
[root@localhost ~]# lspci -vnd 8086:0435
83:00.0 0b40: 8086:0435
Subsystem: 8086:35c5
Physical Slot: 0-1
Flags: bus master, fast devsel, latency 0, IRQ 35
Memory at 387fffd00000 (64-bit, prefetchable) [size=512K]
Memory at c8200000 (64-bit, non-prefetchable) [size=256K]
Memory at c8240000 (64-bit, non-prefetchable) [size=256K]
Capabilities: [b0] MSI: Enable- Count=1/1 Maskable+ 64bit+
Capabilities: [60] MSI-X: Enable- Count=33 Masked-
Capabilities: [6c] Power Management version 3
Capabilities: [74] Express Endpoint, MSI 00
Capabilities: [100] Advanced Error Reporting
Capabilities: [138] Alternative Routing-ID Interpretation (ARI)
Capabilities: [140] Single Root I/O Virtualization (SR-IOV)
Kernel driver in use: vfio-pci
88:00.0 0b40: 8086:0435
Subsystem: 8086:35c5
Physical Slot: 0-2
Flags: fast devsel, IRQ 39
Memory at 387fffe00000 (64-bit, prefetchable) [size=512K]
Memory at c8000000 (64-bit, non-prefetchable) [size=256K]
Memory at c8040000 (64-bit, non-prefetchable) [size=256K]
Capabilities: [b0] MSI: Enable- Count=1/1 Maskable+ 64bit+
Capabilities: [60] MSI-X: Enable- Count=33 Masked-
Capabilities: [6c] Power Management version 3
Capabilities: [74] Express Endpoint, MSI 00
Capabilities: [100] Advanced Error Reporting
Capabilities: [138] Alternative Routing-ID Interpretation (ARI)
Capabilities: [140] Single Root I/O Virtualization (SR-IOV)
Kernel driver in use: vfio-pci
[root@localhost ~]#
In VM, the Intel QAT device shows up in lspci as "Intel Corporation Coleto
Creek PCIe Endpoint".
[root@test1 QAT]# lspci -nn | grep 0435
00:05.0 Co-processor [0b40]: Intel Corporation Coleto Creek PCIe Endpoint
[8086:0435]
[root@test1 QAT]#
Afterward I installed Intel® QuickAssist Technology Software i.e.,
qatmux.l.2.6.0-60.tgz ( downloaded at
https://01.org/packet-processing/intel%C2%AE-quickassist-technology-drivers-and-patches)
on the Guest VM . Here are the details of VM.
1) Fedora release 20 (Heisenbug).
2) Linux Kernel version: 3.12.9-301.fc20.x86_64.
3) BOOT_IMAGE=/boot/vmlinuz-3.12.9-301.fc20.x86_64
root=UUID=ab47cbc9-68ee-403c-96a5-184e68238e65 ro console=ttyS0,115200n8
intel_iommu=off
*Notes: *
1) We are not using SR-IOV and are instead passing through a PF for
acceleration services on guest only. Hence we have not installed the
Intel® QuickAssist Technology Software package on the host.
2) Found an Intel QAT driver with this kernel . Upon doing #lsmod |
grep qa, showed icp_qa_al_vf. Thus did rmmod icp_qa_al_vf and service
qat_service shutdown. This was done before building Intel QuickAssist
Technology Software package.
For passthrough of PF, opted the below Install Acceleration option
a) export ICP_SRIOV=0
b) export ICP_WITHOUT_IOMMU=1
c) ./installer.sh build_accel QAT1.6
d) ./installer.sh install QAT1.6 guest
e) insmod /root/QAT /QAT1.6/build/icp_qa_al.ko
When I issue the below command to start the acceleration software, it fails
to start the driver.
[root@test1 ~]# service qat_service start
Processing file: /etc/dh895xcc_qa_dev0.conf
ADF_CONFIG_CTL err: adf_set_bank_core_affinity: Invalid core affinity
settings for accelerator 0 bank 8. Max cpu number is 3. Defaulting to 0
ADF_CONFIG_CTL err: adf_set_bank_core_affinity: Invalid core affinity
settings for accelerator 0 bank 9. Max cpu number is 3. Defaulting to 0
Parity err reporting is disabled.
Error: failed to start device 0
There is 1 acceleration device(s) in the system:
icp_dev0 - type=dh895xcc, inst_id=0, node_id=0, bdf=00:05:0, #accel=6,
#engines=12, state=down
[root@test1 ~]#
Here are the kernel logs
[root@test1 ~]# dmesg -c
[ 3104.979666] Reading config file.
[ 3104.980250] Starting acceleration device icp_dev0.
[ 3104.980731] Resetting device icp_dev0
[ 3105.126920] qat_1_6_adf 0000:00:05.0: irq 45 for MSI/MSI-X
[ 3105.126939] qat_1_6_adf 0000:00:05.0: irq 46 for MSI/MSI-X
[ 3105.126956] qat_1_6_adf 0000:00:05.0: irq 47 for MSI/MSI-X
[ 3105.126973] qat_1_6_adf 0000:00:05.0: irq 48 for MSI/MSI-X
[ 3105.126989] qat_1_6_adf 0000:00:05.0: irq 49 for MSI/MSI-X
[ 3105.127029] qat_1_6_adf 0000:00:05.0: irq 50 for MSI/MSI-X
[ 3105.127048] qat_1_6_adf 0000:00:05.0: irq 51 for MSI/MSI-X
[ 3105.127065] qat_1_6_adf 0000:00:05.0: irq 52 for MSI/MSI-X
[ 3105.127082] qat_1_6_adf 0000:00:05.0: irq 53 for MSI/MSI-X
[ 3105.127098] qat_1_6_adf 0000:00:05.0: irq 54 for MSI/MSI-X
[ 3105.127115] qat_1_6_adf 0000:00:05.0: irq 55 for MSI/MSI-X
[ 3105.127131] qat_1_6_adf 0000:00:05.0: irq 56 for MSI/MSI-X
[ 3105.127148] qat_1_6_adf 0000:00:05.0: irq 57 for MSI/MSI-X
[ 3105.127164] qat_1_6_adf 0000:00:05.0: irq 58 for MSI/MSI-X
[ 3105.127181] qat_1_6_adf 0000:00:05.0: irq 59 for MSI/MSI-X
[ 3105.127198] qat_1_6_adf 0000:00:05.0: irq 60 for MSI/MSI-X
[ 3105.127215] qat_1_6_adf 0000:00:05.0: irq 61 for MSI/MSI-X
[ 3105.127231] qat_1_6_adf 0000:00:05.0: irq 62 for MSI/MSI-X
[ 3105.127248] qat_1_6_adf 0000:00:05.0: irq 63 for MSI/MSI-X
[ 3105.127265] qat_1_6_adf 0000:00:05.0: irq 64 for MSI/MSI-X
[ 3105.127282] qat_1_6_adf 0000:00:05.0: irq 65 for MSI/MSI-X
[ 3105.127298] qat_1_6_adf 0000:00:05.0: irq 66 for MSI/MSI-X
[ 3105.127315] qat_1_6_adf 0000:00:05.0: irq 67 for MSI/MSI-X
[ 3105.127331] qat_1_6_adf 0000:00:05.0: irq 68 for MSI/MSI-X
[ 3105.127348] qat_1_6_adf 0000:00:05.0: irq 69 for MSI/MSI-X
[ 3105.127364] qat_1_6_adf 0000:00:05.0: irq 70 for MSI/MSI-X
[ 3105.127380] qat_1_6_adf 0000:00:05.0: irq 71 for MSI/MSI-X
[ 3105.127397] qat_1_6_adf 0000:00:05.0: irq 72 for MSI/MSI-X
[ 3105.127413] qat_1_6_adf 0000:00:05.0: irq 73 for MSI/MSI-X
[ 3105.127429] qat_1_6_adf 0000:00:05.0: irq 74 for MSI/MSI-X
[ 3105.127446] qat_1_6_adf 0000:00:05.0: irq 75 for MSI/MSI-X
[ 3105.127462] qat_1_6_adf 0000:00:05.0: irq 76 for MSI/MSI-X
[ 3105.127487] qat_1_6_adf 0000:00:05.0: irq 77 for MSI/MSI-X
[ 3105.157973] icp_qa_al err: adf_subsystemInit: Invalid SRIOV_Enabled
value, driver is built with SRIOV.
[ 3105.158891] icp_qa_al err: adf_do_init: adf_subsystemInit error,
stopping and shutting down
[ 3106.162962] icp_qa_al err: adf_init_all: failed to initialize accelerator
[ 3106.163670] icp_qa_al err: adf_init_devices: failed init accel 0
[ 3106.164256] icp_qa_al err: adf_ctl_ioctl_dev_start: failed call to
adf_init_devices
[root@test1 ~]#
Can anyone please let me know what might be the issue and its remedy ?
please feel free to correct me if I am wrong. If it is not the right forum
, please let me know the correct one. Thank you in advance for time and
support.
Regards,
Chinmaya
next reply other threads:[~2016-06-10 7:16 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-06-10 7:16 Chinmaya Dwibedy [this message]
2016-06-10 7:33 ` SAKTHIVEL ANAND S
2016-06-10 8:57 ` Chinmaya Dwibedy
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='CAMHS8RF_=TgPxt8KqV-oCHqDieBEHHXb=e7aos54zVCQamQGBw@mail.gmail.com' \
--to=ckdwibedy@gmail.com \
--cc=users@dpdk.org \
/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).