[BlueZ,v2,0/4] Allow to reattach with new composition data

Message

Jakub Witowski Jan. 30, 2020, 2:34 p.m. UTC

This patch allows the application to modify the CID, PID and VID in the composition data.

Version 2: Do not allow to change CRPL in the composition data.
Additionaly verify the device key when updating comp data and remove
unused function in the 3rd patch.

Version 1: According the Mesh Profile (2.3.4 Elements) the modification of fields
other than "Elements" is not prohibited.

Also in my opinion (as you can see in the 1st patch), there is no need to use pointer to
the node_composition struct. The static is more clear and less problematic.

Jakub Witowski (4):
  mesh: use static node_comp instead of the pointer
  mesh: add cid/pid/vid setter
  mesh: remove unused node_set_device_key()
  mesh: allow to reattach with new composition data

 mesh/mesh-config-json.c |  40 ++++++++++++----
 mesh/mesh-config.h      |   2 +
 mesh/node.c             | 100 +++++++++++++++++++++++++---------------
 mesh/node.h             |   1 -
 4 files changed, 96 insertions(+), 47 deletions(-)

Comments

Brian Gix Jan. 31, 2020, 6:30 p.m. UTC | #1

Hi Jakub,
On Thu, 2020-01-30 at 15:34 +0100, Jakub Witowski wrote:
> This patch allows the application to modify the CID, PID and VID in the composition data.
> 
> Version 2: Do not allow to change CRPL in the composition data.
> Additionaly verify the device key when updating comp data and remove
> unused function in the 3rd patch.
> 
> Version 1: According the Mesh Profile (2.3.4 Elements) the modification of fields
> other than "Elements" is not prohibited.
> 
> Also in my opinion (as you can see in the 1st patch), there is no need to use pointer to
> the node_composition struct. The static is more clear and less problematic.
> 
> Jakub Witowski (4):
>   mesh: use static node_comp instead of the pointer
>   mesh: add cid/pid/vid setter
>   mesh: remove unused node_set_device_key()
>   mesh: allow to reattach with new composition data

Patches 1 and 3 of this patchset have been applied, as they are non-controversial.

I would like to wait a little while, as the Working group weighs in, on modifying composition data.

I am actually ready today to allow an App to Attach to an existing node, with modifications to it's CID/PID/VID
(drop daemon validation) with the understanding that the composition stored in node.json is not changed.  But
anything that changes OTA behavior, I would like blessed by the WG and the SIG.


> 
>  mesh/mesh-config-json.c |  40 ++++++++++++----
>  mesh/mesh-config.h      |   2 +
>  mesh/node.c             | 100 +++++++++++++++++++++++++---------------
>  mesh/node.h             |   1 -
>  4 files changed, 96 insertions(+), 47 deletions(-)
>