[MPTCP] Re: [PATCH net-next 0/8] mptcp: Add genl events for connection info

[MPTCP] Re: [PATCH net-next 0/8] mptcp: Add genl events for connection info

[Mat Martineau]

[-- Attachment #1: Type: text/plain, Size: 2745 bytes --]

On Fri, 12 Feb 2021, Mat Martineau wrote:

> This series from the MPTCP tree adds genl multicast events that are
> important for implementing a userspace path manager. In MPTCP, a path
> manager is responsible for adding or removing additional subflows on
> each MPTCP connection. The in-kernel path manager (already part of the
> kernel) is a better fit for many server use cases, but the additional
> flexibility of userspace path managers is often useful for client
> devices.
>
> Patches 1, 2, 4, 5, and 6 do some refactoring to streamline the netlink
> event implementation in the final patch.
>
> Patch 3 improves the timeliness of subflow destruction to ensure the
> 'subflow closed' event will be sent soon enough.
>
> Patch 7 allows use of the GENL_UNS_ADMIN_PERM flag on genl mcast groups
> to mandate CAP_NET_ADMIN, which is important to protect token information
> in the MPTCP events. This is a genetlink change.
>

David -

I noticed that this series got merged to net-next and shows as "Accepted" 
in patchwork:

https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next.git/commit/?id=0a2f6b32cc45e3918321779fe90c28f1ed27d2af


However, somehow patch 7 did not propagate through the netdev list and 
does not show up in patchwork or the merged series!


It did get archived on the cc'd mptcp list 
(https://lists.01.org/hyperkitty/list/mptcp(a)lists.01.org/message/KBY6UIFETMXCAOHNXQLYWKXNCHSGJ7AG/) 
so hopefully your directly-addressed copy also arrived. The missing patch 
won't cause a build error but does lead to a token getting exposed to 
non-admin users.

I will re-send it with an extra note about where it originally fit in.


Mat


> Patch 8 adds the MPTCP netlink events.
>
>
> Florian Westphal (8):
>  mptcp: move pm netlink work into pm_netlink
>  mptcp: split __mptcp_close_ssk helper
>  mptcp: schedule worker when subflow is closed
>  mptcp: move subflow close loop after sk close check
>  mptcp: pass subflow socket to a few helpers
>  mptcp: avoid lock_fast usage in accept path
>  genetlink: add CAP_NET_ADMIN test for multicast bind
>  mptcp: add netlink event support
>
> include/net/genetlink.h    |   1 +
> include/uapi/linux/mptcp.h |  74 +++++++++
> net/mptcp/options.c        |   2 +-
> net/mptcp/pm.c             |  24 ++-
> net/mptcp/pm_netlink.c     | 310 ++++++++++++++++++++++++++++++++++++-
> net/mptcp/protocol.c       |  72 ++++-----
> net/mptcp/protocol.h       |  20 +--
> net/mptcp/subflow.c        |  27 +++-
> net/netlink/genetlink.c    |  32 ++++
> 9 files changed, 491 insertions(+), 71 deletions(-)
>
>
> base-commit: c3ff3b02e99c691197a05556ef45f5c3dd2ed3d6
> -- 
> 2.30.1
>
>

--
Mat Martineau
Intel

[MPTCP] Re: [PATCH net-next 0/8] mptcp: Add genl events for connection info

[Mat Martineau]

[-- Attachment #1: Type: text/plain, Size: 2013 bytes --]


On Fri, 12 Feb 2021, Mat Martineau wrote:

> On Fri, 12 Feb 2021, Mat Martineau wrote:
>
>> This series from the MPTCP tree adds genl multicast events that are
>> important for implementing a userspace path manager. In MPTCP, a path
>> manager is responsible for adding or removing additional subflows on
>> each MPTCP connection. The in-kernel path manager (already part of the
>> kernel) is a better fit for many server use cases, but the additional
>> flexibility of userspace path managers is often useful for client
>> devices.
>> 
>> Patches 1, 2, 4, 5, and 6 do some refactoring to streamline the netlink
>> event implementation in the final patch.
>> 
>> Patch 3 improves the timeliness of subflow destruction to ensure the
>> 'subflow closed' event will be sent soon enough.
>> 
>> Patch 7 allows use of the GENL_UNS_ADMIN_PERM flag on genl mcast groups
>> to mandate CAP_NET_ADMIN, which is important to protect token information
>> in the MPTCP events. This is a genetlink change.
>> 
>
> David -
>
> I noticed that this series got merged to net-next and shows as "Accepted" in 
> patchwork:
>
> https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next.git/commit/?id=0a2f6b32cc45e3918321779fe90c28f1ed27d2af
>
>
> However, somehow patch 7 did not propagate through the netdev list and does 
> not show up in patchwork or the merged series!
>
>
> It did get archived on the cc'd mptcp list 
> (https://lists.01.org/hyperkitty/list/mptcp(a)lists.01.org/message/KBY6UIFETMXCAOHNXQLYWKXNCHSGJ7AG/) 
> so hopefully your directly-addressed copy also arrived. The missing patch 
> won't cause a build error but does lead to a token getting exposed to 
> non-admin users.
>
> I will re-send it with an extra note about where it originally fit in.

Ok, please disregard. When I went to prepare the patch to re-send I found 
that Dave was a step ahead of me, and the code is in fact in the tree. 
Thanks for handling the issue.


--
Mat Martineau
Intel

[MPTCP] Re: [PATCH net-next 0/8] mptcp: Add genl events for connection info

[David Miller]

[-- Attachment #1: Type: text/plain, Size: 1519 bytes --]

From: Mat Martineau <mathew.j.martineau(a)linux.intel.com>
Date: Fri, 12 Feb 2021 20:46:30 -0800 (PST)

> On Fri, 12 Feb 2021, Mat Martineau wrote:
> 
>> This series from the MPTCP tree adds genl multicast events that are
>> important for implementing a userspace path manager. In MPTCP, a path
>> manager is responsible for adding or removing additional subflows on
>> each MPTCP connection. The in-kernel path manager (already part of the
>> kernel) is a better fit for many server use cases, but the additional
>> flexibility of userspace path managers is often useful for client
>> devices.
>>
>> Patches 1, 2, 4, 5, and 6 do some refactoring to streamline the
>> netlink
>> event implementation in the final patch.
>>
>> Patch 3 improves the timeliness of subflow destruction to ensure the
>> 'subflow closed' event will be sent soon enough.
>>
>> Patch 7 allows use of the GENL_UNS_ADMIN_PERM flag on genl mcast
>> groups
>> to mandate CAP_NET_ADMIN, which is important to protect token
>> information
>> in the MPTCP events. This is a genetlink change.
>>
> 
> David -
> 
> I noticed that this series got merged to net-next and shows as
> "Accepted" in patchwork:
> 
> https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next.git/commit/?id=0a2f6b32cc45e3918321779fe90c28f1ed27d2af
> 
> 
> However, somehow patch 7 did not propagate through the netdev list and
> does not show up in patchwork or the merged series!

I did put patch 7 into the commit, so no worries...