[dpdk-dev] doc: announce ivshmem support removal

[Mauricio Vasquez]

Hello All,

Here in Politecnico di Torino we use the ivshmem technology from a 
research point of view.

Our research efforts focus in optimizing the inter-Virtual Network 
Function communication, currently we have implemented two versions of 
our prototype, they are described in [1] and [2].

Unfortunately, we do not have the human resources to implement the 
improvements that are necessary for ivshmem in DPDK, however we could 
provide some feedback and testing for possible patches.

Best Regards,

Mauricio Vasquez.

[1] 
https://www.researchgate.net/publication/305699120_Transparent_Optimization_of_Inter-Virtual_Network_Function_Communication_in_Open_vSwitch

[2] 
https://www.researchgate.net/publication/305699122_A_Transparent_Highway_for_inter-Virtual_Network_Function_Communication_with_Open_vSwitch


On 07/28/2016 11:20 AM, Christian Ehrhardt wrote:
> Hi Thomas,
> just my two cents as Ubuntu DPDK maintainer (and part of the Debian Team
> that does the same).
> (It seems I can reuse that line for all posts about the deprecation notices
> :-) )
>
> While IVSHMEM was enabled (as it was the default) I never heard of any
> users of what we provided so far.
> But that is expected considering that not all qemu bits are landed either.
> Since it will follow the process of "deprecation notice -> grace period ->
> ABI bump" on removal, I think packaging and consuming applications should
> be fine.
>
> I'd agree to Hiroshi that, if really needed a clean re-implementation more
> properly separated from EAL likely is the best way to go.
>
> I think it is a good change to drop rather complex code in favor of
> stabilizing the main paths:
> Acked-by: Christian Ehrhardt <christian.ehrhardt at canonical.com>
>
>
> Christian Ehrhardt
> Software Engineer, Ubuntu Server
> Canonical Ltd
>
> On Wed, Jul 27, 2016 at 9:08 PM, Jan Viktorin <viktorin at rehivetech.com>
> wrote:
>
>> On Wed, 20 Jul 2016 18:35:46 +0200
>> Thomas Monjalon <thomas.monjalon at 6wind.com> wrote:
>>
>>> There was a prior call with an explanation of what needs to be done:
>>>        http://dpdk.org/ml/archives/dev/2016-June/040844.html
>>> - Qemu patch upstreamed
>>> - IVSHMEM PCI device managed by a PCI driver
>>> - No DPDK objects (ring/mempool) allocated by EAL
>>>
>>> As nobody seems interested, it is time to remove this code which
>>> makes EAL improvements harder.
>>>
>>> Signed-off-by: Thomas Monjalon <thomas.monjalon at 6wind.com>
>>> Acked-by: David Marchand <david.marchand at 6wind.com>
>>> Acked-by: Maxime Coquelin <maxime.coquelin at redhat.com>
>> Acked-by: Jan Viktorin <viktorin at rehivetech.com
>>