Virtualization - Jun 2021 - [PATCH linux-next v3 2/6] vdpa: Introduce query of device config layout

> From: Jason Wang <jasowang at redhat.com>
> Sent: Thursday, June 24, 2021 11:13 AM
> 
> ? 2021/6/23 ??12:22, Parav Pandit ??:
> >
> >> From: Jason Wang <jasowang at redhat.com>
> >> Sent: Wednesday, June 23, 2021 9:39 AM
> >>
> >> ? 2021/6/22 ??10:03, Parav Pandit ??:
> >>>> Is it better to use a separate enum for net specific
attributes?
> >>>>
> >>> Yes, because they are only net specific.
> >>> I guess it is related to your below question.
> >>>
> >>>> Another question (sorry if it has been asked before). Can
we simply
> >>>> return the config (binary) to the userspace, then usespace
can use
> >>>> the existing uAPI like virtio_net_config plus the feature
to
> >>>> explain the
> >> config?
> >>> We did discuss in v2.
> >>> Usually returning the whole blob and parsing is not desired
via netlink.
> >>> Returning individual fields give the full flexibility to
return only
> >>> the valid
> >> fields.
> >>> Otherwise we need to implement another bitmask too to tell
which
> >>> fields
> >> from the struct are valid and share with user space.
> >>> Returning individual fields is the widely used approach.
> >>
> >> The main concerns are:
> >>
> >> 1) The blob will be self contained if it was passed with the
> >> negotiated features, so we don't need bitmask.
> > Which fields of the struct are valid is told by additional fields.
> >> 2) Using individual fields means it must duplicate the config
fields
> >> of every virtio devices
> >>
> > Mostly no. if there are common config fields across two device types,
> > they would be named as
> > VDPA_ATTR_DEV_CFG_*
> > Net specific will be,
> > VDPA_ATTR_DEV_NET_CFG_*
> > Block specific, will be,
> > VDPA_ATTR_DEV_BLK_CFG_*
> 
> 
> I meant it looks like VDPA_ATTR_DEV_NET will duplicate all the fields of:
> 
> struct virtio_net_config;
> 
> And VDPA_ATTR_DEV_BLOCK will duplicate all the fields of
> 
> struct virtio_blk_config; which has ~21 fields.
> 
> And we had a plenty of other types of virtio devices.
> 
> Consider we had a mature set of virtio specific uAPI for config space.
> It would be a burden if we need an unnecessary translation layer of netlink
in
> the middle:
> 
> [vDPA parent (virtio_net_config)] <-> [netlink
(VDPA_ATTR_DEV_NET_XX)]
> <-> [userspace (VDPA_ATTR_DEV_NET_XX)] 
>> <-> [ user (virtio_net_config)]
This translation is not there. We show relevant net config fields as
VDPA_ATTR_DEV_NET individually.
It is not a binary dump which is harder for users to parse and make any use of
it.

It is only one level of translation from virtio_net_config (kernel) ->
netlink vdpa fields.
It is similar to 'struct netdevice' -> rtnl info fields.
> 
> If we make netlink simply a transport, it would be much easier. And we had
> the chance to unify the logic of build_config() and set_config() in the
driver.
How? We need bit mask to tell that out of 21 fields which fields to update and
which not.
And that is further mixed with offset and length.
> 
> 
> >
> >> And actually, it's not the binary blob since uapi clearly
define the
> >> format (e.g struct virtio_net_config), can we find a way to use
that?
> >> E.g introduce device/net specific command and passing the blob
with
> >> length and negotiated features.
> > Length may change in future, mostly expand. And parsing based on
length
> is not such a clean way.
> 
> 
> Length is only for legal checking. The config is self contained with:
> 
Unlikely. When structure size increases later, the parsing will change based on
the length.
Because older kernel would return shorter length with older iproute2 tool.
So user space always have to deal and have nasty parsing/typecasting based on
the length.
> 1) device id
> 2) features
> 
> 
> > Parsing fields require knowledge of features as well and application
needs
> to make multiple netlink calls to parse the config space.
> 
> 
> I think we don't care about the performance in this case. It's
about three
> netlink calls:
> 
Its not about performance. By the time 1st call is made, features got updated
and it is out of sync with config.
> 1) get config
> 2) get device id
> 3) get features
> 
This requires using features from 3rd netlink output to decode output of 1st
netlink output.
Which is a bit odd of netlink.
Other netlink nla_put() probably sending whole structure doesn?t need to do it.
> For build config, it's only one
> 
> 1) build config
> 
> 
> > I prefer to follow rest of the kernel style to return self contained
> invidividual fields.
> 
> 
> But I saw a lot of kernel codes choose to use e.g nla_put() directly with
> module specific structure.
> 
It might be self-contained structure that probably has not found the need to
expand.

? 2021/6/24 ??2:29, Parav Pandit ??:
>
>> From: Jason Wang <jasowang at redhat.com>
>> Sent: Thursday, June 24, 2021 11:13 AM
>>
>> ? 2021/6/23 ??12:22, Parav Pandit ??:
>>>> From: Jason Wang <jasowang at redhat.com>
>>>> Sent: Wednesday, June 23, 2021 9:39 AM
>>>>
>>>> ? 2021/6/22 ??10:03, Parav Pandit ??:
>>>>>> Is it better to use a separate enum for net specific
attributes?
>>>>>>
>>>>> Yes, because they are only net specific.
>>>>> I guess it is related to your below question.
>>>>>
>>>>>> Another question (sorry if it has been asked before).
Can we simply
>>>>>> return the config (binary) to the userspace, then
usespace can use
>>>>>> the existing uAPI like virtio_net_config plus the
feature to
>>>>>> explain the
>>>> config?
>>>>> We did discuss in v2.
>>>>> Usually returning the whole blob and parsing is not desired
via netlink.
>>>>> Returning individual fields give the full flexibility to
return only
>>>>> the valid
>>>> fields.
>>>>> Otherwise we need to implement another bitmask too to tell
which
>>>>> fields
>>>> from the struct are valid and share with user space.
>>>>> Returning individual fields is the widely used approach.
>>>> The main concerns are:
>>>>
>>>> 1) The blob will be self contained if it was passed with the
>>>> negotiated features, so we don't need bitmask.
>>> Which fields of the struct are valid is told by additional fields.
>>>> 2) Using individual fields means it must duplicate the config
fields
>>>> of every virtio devices
>>>>
>>> Mostly no. if there are common config fields across two device
types,
>>> they would be named as
>>> VDPA_ATTR_DEV_CFG_*
>>> Net specific will be,
>>> VDPA_ATTR_DEV_NET_CFG_*
>>> Block specific, will be,
>>> VDPA_ATTR_DEV_BLK_CFG_*
>>
>> I meant it looks like VDPA_ATTR_DEV_NET will duplicate all the fields
of:
>>
>> struct virtio_net_config;
>>
>> And VDPA_ATTR_DEV_BLOCK will duplicate all the fields of
>>
>> struct virtio_blk_config; which has ~21 fields.
>>
>> And we had a plenty of other types of virtio devices.
>>
>> Consider we had a mature set of virtio specific uAPI for config space.
>> It would be a burden if we need an unnecessary translation layer of
netlink in
>> the middle:
>>
>> [vDPA parent (virtio_net_config)] <-> [netlink
(VDPA_ATTR_DEV_NET_XX)]
>> <-> [userspace (VDPA_ATTR_DEV_NET_XX)]
>>> <-> [ user (virtio_net_config)]
> This translation is not there. We show relevant net config fields as
VDPA_ATTR_DEV_NET individually.
> It is not a binary dump which is harder for users to parse and make any use
of it.

The is done implicitly, user needs to understand the semantic of 
virtio_net_config and map the individual fields to the vdpa tool 
sub-command.

>
> It is only one level of translation from virtio_net_config (kernel) ->
netlink vdpa fields.
> It is similar to 'struct netdevice' -> rtnl info fields.

I think not, the problem is that the netdevice is not a part of uAPI but 
virtio_net_config is.

>
>> If we make netlink simply a transport, it would be much easier. And we
had
>> the chance to unify the logic of build_config() and set_config() in the
driver.
> How? We need bit mask to tell that out of 21 fields which fields to update
and which not.
> And that is further mixed with offset and length.

So set_config() could be called from userspace, so did build_config(). 
The only difference is:

1) they're using different transport, ioctl vs netlink
2) build_config() is only expected to be called by the management tool

If qemu works well via set_config ioctl, netlink should work as well.

Btw, what happens if management tool tries to modify the mac of vDPA 
when the device is already used by the driver?

>
>>
>>>> And actually, it's not the binary blob since uapi clearly
define the
>>>> format (e.g struct virtio_net_config), can we find a way to use
that?
>>>> E.g introduce device/net specific command and passing the blob
with
>>>> length and negotiated features.
>>> Length may change in future, mostly expand. And parsing based on
length
>> is not such a clean way.
>>
>>
>> Length is only for legal checking. The config is self contained with:
>>
> Unlikely. When structure size increases later, the parsing will change
based on the length.
> Because older kernel would return shorter length with older iproute2 tool.

This is fine since the older kernel only support less features. The only 
possible issue if the old iproute 2 runs on new kernel. With the current 
proposal, it may cause some config fields can't not be showed.

I think it might be useful to introduce a command to simply dump the 
config space.

> So user space always have to deal and have nasty parsing/typecasting based
on the length.

That's how userspace (Qemu) is expected to work now. The userspace 
should determine the semantic of the fields based on the features.

Differentiate config fields doesn't help much, e.g userspace still need 
to differ LINK_UP and ANNOUNCE for the status field.

>
>> 1) device id
>> 2) features
>>
>>
>>> Parsing fields require knowledge of features as well and
application needs
>> to make multiple netlink calls to parse the config space.
>>
>>
>> I think we don't care about the performance in this case. It's
about three
>> netlink calls:
>>
> Its not about performance. By the time 1st call is made, features got
updated and it is out of sync with config.
>
>> 1) get config
>> 2) get device id
>> 3) get features
>>
> This requires using features from 3rd netlink output to decode output of
1st netlink output.
> Which is a bit odd of netlink.
> Other netlink nla_put() probably sending whole structure doesn?t need to do
it.

Well, we can pack them all into a single skb isn't it? (probably with a 
config len).

>
>> For build config, it's only one
>>
>> 1) build config
>>
>>
>>> I prefer to follow rest of the kernel style to return self
contained
>> invidividual fields.
>>
>>
>> But I saw a lot of kernel codes choose to use e.g nla_put() directly
with
>> module specific structure.
>>
> It might be self-contained structure that probably has not found the need
to expand.

I think it's just a matter of putting the config length with the config 
data. Note that we've already had .get_config_size() ops for validating 
inputs through VHOST_SET_CONFIG/VHOST_GET_CONFIG.

Thanks