Re: [PATCH v6 3/3] ixgbe: Add LRO support

[Vlad Zolotarov <vladz-RmZWMc9puTNJc61us3aD9laTQe2KTcn/@public.gmane.org>]

On 03/13/15 13:28, Ananyev, Konstantin wrote:
> Hi Olivier,
>
>> -----Original Message-----
>> From: Olivier MATZ [mailto:olivier.matz-pdR9zngts4EAvxtiuMwx3w@public.gmane.org]
>> Sent: Friday, March 13, 2015 9:08 AM
>> To: Vlad Zolotarov; Ananyev, Konstantin; dev-VfR2kkLFssw@public.gmane.org
>> Subject: Re: [dpdk-dev] [PATCH v6 3/3] ixgbe: Add LRO support
>>
>> Hi Vlad,
>>
>> On 03/11/2015 05:54 PM, Vlad Zolotarov wrote:
>>>>>> About the existing RX/TX functions and PPC support:
>>>>>> Note that all of them were created before PPC support for DPDK was
>>>>>> introduced.
>>>>>> At that moment only IA was supported.
>>>>>> That's why in some places where you would expect to see 'mb()' there
>>>>>> are 'volatile' and/or ' rte_compiler_barrier' instead.
>>>>>> Why all that places wasn't updated when PPC support was added -
>>>>>> that's another question.
>>>>>>    From my understanding - with current implementation some of DPDK
>>>>>> PMDs RX/TX functions and  rte_ring wouldn't work correctly
>>>>> on PPC.
>>>>>> So, I suppose we need to decide for ourselves - do we really want to
>>>>>> support PPC and other architectures with non-IA memory
>>>>> model or not?
>>>>>> If not, then I think we don't need any mb()s inside recv_pkts_lro()
>>>>>> - just rte_compiler_barrier seems enough, and no point to
>>>>> complain about
>>>>>> it in comments.
>>>>>> If yes - then why to introduce a new function with a known potential
>>>>>> bug?
>>>>> In order to introduce a new function with the proper implementation or
>>>>> to fix any other places with the similar weakness I would need a proper
>>>>> tools like a proper platform-dependent barrier-macros similar to
>>>>> smp_Xmb() Linux macros that reduce to a compiler barrier where
>>>>> appropriate or to a proper memory fence where needed.
>>>> I understand that.
>>>> Let's add new macro for that: rte_smp_Xmb() or something,
>>>> so it would be just rte_compiler_barrier() for x86 and a proper mb()
>>>> for PPC.
>>> There was an idea to use the C11 built-in memory barriers. I suggest we
>>> open a separate discussion about that and add these and the appropriate
>>> fixes in a separate series. There are quite a few places to fix anyway,
>>> which are currently broken on PPC so this patch doesn't make things any
>>> worse. However adding a new memory barrier doesn't belong to an LRO
>>> functionality and thus to this series.
>> This is an interesting discussion. Just for reference, I submitted a
>> patch on this topic but it was probably too early as only Intel
>> architecture was supported at that time.
>>
>> See http://dpdk.org/ml/archives/dev/2014-May/002597.html
> I do remember that conversation :)
> At that moment, as nothing except IA wasn't supported, I feel it was not needed.
> Though now, if we do want to support PPC and other architectures with weak memory model,
> I think we do need to introduce some platform dependent set of Xmb() macros.
> See http://dpdk.org/ml/archives/dev/2014-October/006729.html
>
> Actually while thinking about it once again:
> Is there any good use for rte_compiler_barrier() for PPC memory model?
> I can't think about any.
> So I wonder can't we just make for PPC:
>   #define rte_compiler_barrier    rte_mb
> While keeping it as it is for IA.
> Would save us from searching/replacing though all the code.

I wonder why should we invent a wheel? Like Avi has proposed we may use 
the existing standard C library primitives for that. See this 
http://en.cppreference.com/w/c/atomic. I don't know what's the state of 
icc in this area though... ;)

Pros:

  * Zero maintenance.
  * Multi-platform support.
  * It seems that this is the direction the industry is going to (as
    opposed to the discussed above mb(), rmb(), wmb() model).

Cons:

  * The model is a bit different from what most of the kernel
    programmers used to.
  * The current code adaptation would be a bit more painful (due to
    first "cons").


I think this could be a very nice move. For a user space for sure. The 
open question is a KNI component. I don't know how much code is shared 
between kernel and user space DPDK code but if there isn't much - then 
we may still go for a built-in C atomics primitives in a user space and 
do whatever we choose in the KNI...

>
>   Konstantin
>
>
>
>> Regards,
>> Olivier