From: "Burakov, Anatoly" <anatoly.burakov@intel.com>
To: Bruce Richardson <bruce.richardson@intel.com>,
Sarosh Arif <sarosh.arif@emumba.com>
Cc: dev@dpdk.org
Subject: Re: [dpdk-dev] unable to bind to vfio-pci
Date: Thu, 17 Sep 2020 12:49:59 +0100 [thread overview]
Message-ID: <8fa1aae5-617c-64a7-a647-fd2508e0f662@intel.com> (raw)
In-Reply-To: <20200917102204.GE1568@bricha3-MOBL.ger.corp.intel.com>
On 17-Sep-20 11:22 AM, Bruce Richardson wrote:
> On Thu, Sep 17, 2020 at 02:21:27PM +0500, Sarosh Arif wrote:
>> On Thu, Sep 17, 2020 at 2:17 PM Bruce Richardson
>> <bruce.richardson@intel.com> wrote:
>>>
>>> On Thu, Sep 17, 2020 at 11:52:58AM +0500, Sarosh Arif wrote:
>>>> I have been trying to bind to vfio-pci using usertools/dpdk-devbind.py
>>>> but am unable to do so. The reason behind this is that I am unable to
>>>> write in /sys/bus/pci/drivers/vfio-pci/bind. Upon searching solutions
>>>> I tried a couple of things such as setting iommu=pt and intel_iommu=on
>>>> and ensured vt-d is enabled.
>>>> Along with this I have made sure that the vfio-pci module is correctly
>>>> loaded. I have also tried
>>>>
>>>> chmod 666 /sys/bus/pci/drivers/vfio-pci/bind
>>>>
>>>> So that I have permissions to write in this file.
>>>>
>>>> The error I get when I use usertools/dpdk-devbind.py to bind is this:
>>>> Error: bind failed for 0000:b7:00.1 - Cannot bind to driver vfio-pci
>>>>
>>>> The details of 0000:b7:00.1 are as follows:
>>>> Ethernet Connection X722 for 10GBASE-T 37d2' if=eno6 drv=i40e
>>>>
>>>> I have also unbinded The pci bridge to which 0000:b7:00.1 was connected.
>>>>
>>>> What more can be done to resolve this?
>>>>
>>> Since you describe changing permissions on the "bind" file, are you trying
>>> to run dpdk-devbind.py as a non-root user? Does it work as root?
>> I am running it as a root user. It does not work as a root user.
>
> One possible problem that it could be, is that you will need to ensure that
> any other ports on the same device are either similarly bound to vfio-pci
> or not bound to any driver. You can't have e.g. a 2-port X722 NIC where one
> port is bound to the kernel driver, while another is bound to vfio in
> userspace.
>
I don't think this would result in a bind error; rather, you'd see the
IOMMU group as non-viable when attempting to use it with VFIO, but the
binding itself should succeed.
--
Thanks,
Anatoly
next prev parent reply other threads:[~2020-09-17 11:50 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-09-17 6:52 Sarosh Arif
2020-09-17 7:31 ` Dybkowski, AdamX
2020-09-17 8:39 ` Sarosh Arif
2020-09-17 9:17 ` Bruce Richardson
2020-09-17 9:21 ` Sarosh Arif
2020-09-17 9:35 ` Burakov, Anatoly
2020-09-21 11:54 ` Sarosh Arif
2020-09-21 11:58 ` Burakov, Anatoly
2020-09-17 10:22 ` Bruce Richardson
2020-09-17 11:49 ` Burakov, Anatoly [this message]
2020-09-21 11:10 ` Sarosh Arif
2020-09-21 11:36 ` Burakov, Anatoly
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=8fa1aae5-617c-64a7-a647-fd2508e0f662@intel.com \
--to=anatoly.burakov@intel.com \
--cc=bruce.richardson@intel.com \
--cc=dev@dpdk.org \
--cc=sarosh.arif@emumba.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).