diff mbox series

[v7,1/3] ethdev: add level support for RSS offload types

Message ID 20200901032708.58247-1-kirankumark@marvell.com (mailing list archive)
State Superseded, archived
Delegated to: Ferruh Yigit
Headers show
Series [v7,1/3] ethdev: add level support for RSS offload types | expand

Checks

Context Check Description
ci/checkpatch success coding style OK

Commit Message

Kiran Kumar Kokkilagadda Sept. 1, 2020, 3:27 a.m. UTC
From: Kiran Kumar K <kirankumark@marvell.com>

This patch reserves 2 bits as input selection to select Inner and
outer encapsulation level for RSS computation. It is combined with existing
ETH_RSS_* to choose Inner or outer layers.
This functionality already exists in rte_flow through level parameter in
RSS action configuration rte_flow_action_rss.

Signed-off-by: Kiran Kumar K <kirankumark@marvell.com>
---
V7 Changes:
* Re-worked to keep it in sync with rte_flow_action_rss and support upto
3 levels.
* Addressed testpmd review comments.

 lib/librte_ethdev/rte_ethdev.h | 27 +++++++++++++++++++++++++++
 1 file changed, 27 insertions(+)

--
2.25.1

Comments

Ferruh Yigit Sept. 1, 2020, 1:37 p.m. UTC | #1
On 9/1/2020 4:27 AM, kirankumark@marvell.com wrote:
> From: Kiran Kumar K <kirankumark@marvell.com>
> 
> This patch reserves 2 bits as input selection to select Inner and
> outer encapsulation level for RSS computation. It is combined with existing
> ETH_RSS_* to choose Inner or outer layers.
> This functionality already exists in rte_flow through level parameter in
> RSS action configuration rte_flow_action_rss.
> 
> Signed-off-by: Kiran Kumar K <kirankumark@marvell.com>
> ---
> V7 Changes:
> * Re-worked to keep it in sync with rte_flow_action_rss and support upto
> 3 levels.
> * Addressed testpmd review comments.
> 
>   lib/librte_ethdev/rte_ethdev.h | 27 +++++++++++++++++++++++++++
>   1 file changed, 27 insertions(+)
> 
> diff --git a/lib/librte_ethdev/rte_ethdev.h b/lib/librte_ethdev/rte_ethdev.h
> index 70295d7ab..13e49bbd7 100644
> --- a/lib/librte_ethdev/rte_ethdev.h
> +++ b/lib/librte_ethdev/rte_ethdev.h
> @@ -552,6 +552,33 @@ struct rte_eth_rss_conf {
>   #define RTE_ETH_RSS_L3_PRE64	   (1ULL << 53)
>   #define RTE_ETH_RSS_L3_PRE96	   (1ULL << 52)
> 
> +/*
> + * We use the following macros to combine with the above layers to choose
> + * inner and outer layers or both for RSS computation.
> + * bit 50 and 51 are reserved for this.
> + */
> +
> +/** level 0, requests the default behavior. Depending on the packet
> + * type, it can mean outermost, innermost, anything in between or even no RSS.
> + * It basically stands for the innermost encapsulation level RSS
> + * can be performed on according to PMD and device capabilities.
> + */
> +#define ETH_RSS_LEVEL_0         (0ULL << 50)

I can see from history how this is involved, but the 'ETH_RSS_LEVEL_0' naming is 
not really clear what it is, the naming in v6 is more clear.

What about following one:
0 -> LEVEL_PMD_DEFAULT
1 -> LEVEL_OUTER
2 -> LEVEL_INNER
3 -> LEVEL_INNER_OUTER

This doesn't exactly match to rte_flow one, but closer than v6 one. This ends 
with max level 2. And defines a way to say both inner and outer.

> +
> +/** level 1,  requests RSS to be performed on the outermost packet
> + * encapsulation level.
> + */
> +#define ETH_RSS_LEVEL_1         (1ULL << 50)
> +
> +/** level 2,  requests RSS to be performed on the
> + * specified inner packet encapsulation level, from outermost to
> + * innermost (lower to higher values).
> + */
> +#define ETH_RSS_LEVEL_2	        (2ULL << 50)

I can see you are trying to copy rte_flow usage, but this doesn't really makes 
sense here. Where the value of the level is defined in this case? If not defined 
how the PMD knows which level to use?

> +#define ETH_RSS_LEVEL_MASK	(3ULL << 50)
> +
> +#define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50)
> +
>   /**
>    * For input set change of hash filter, if SRC_ONLY and DST_ONLY of
>    * the same level are used simultaneously, it is the same case as
> --
> 2.25.1
>
Kiran Kumar Kokkilagadda Sept. 1, 2020, 2:27 p.m. UTC | #2
> -----Original Message-----
> From: Ferruh Yigit <ferruh.yigit@intel.com>
> Sent: Tuesday, September 1, 2020 7:08 PM
> To: Kiran Kumar Kokkilagadda <kirankumark@marvell.com>; Thomas Monjalon
> <thomas@monjalon.net>; Andrew Rybchenko <arybchenko@solarflare.com>
> Cc: dev@dpdk.org; Jerin Jacob Kollanukkaran <jerinj@marvell.com>;
> orika@mellanox.com; xuanziyang2@huawei.com;
> cloud.wangxiaoyun@huawei.com; zhouguoyang@huawei.com;
> rosen.xu@intel.com; beilei.xing@intel.com; jia.guo@intel.com; Rasesh Mody
> <rmody@marvell.com>; Shahed Shaikh <shshaikh@marvell.com>; Nithin Kumar
> Dabilpuram <ndabilpuram@marvell.com>; qiming.yang@intel.com;
> qi.z.zhang@intel.com; keith.wiles@intel.com; hemant.agrawal@nxp.com;
> sachin.saxena@nxp.com; wei.zhao1@intel.com; johndale@cisco.com;
> hyonkim@cisco.com; chas3@att.com; matan@mellanox.com;
> shahafs@mellanox.com; viacheslavo@mellanox.com;
> rahul.lakkireddy@chelsio.com; grive@u256.net; Liron Himi
> <lironh@marvell.com>; jingjing.wu@intel.com; xavier.huwei@huawei.com;
> humin29@huawei.com; yisen.zhuang@huawei.com;
> ajit.khaparde@broadcom.com; somnath.kotur@broadcom.com;
> jasvinder.singh@intel.com; cristian.dumitrescu@intel.com
> Subject: [EXT] Re: [dpdk-dev][PATCH v7 1/3] ethdev: add level support for RSS
> offload types
> 
> External Email
> 
> ----------------------------------------------------------------------
> On 9/1/2020 4:27 AM, kirankumark@marvell.com wrote:
> > From: Kiran Kumar K <kirankumark@marvell.com>
> >
> > This patch reserves 2 bits as input selection to select Inner and
> > outer encapsulation level for RSS computation. It is combined with
> > existing
> > ETH_RSS_* to choose Inner or outer layers.
> > This functionality already exists in rte_flow through level parameter
> > in RSS action configuration rte_flow_action_rss.
> >
> > Signed-off-by: Kiran Kumar K <kirankumark@marvell.com>
> > ---
> > V7 Changes:
> > * Re-worked to keep it in sync with rte_flow_action_rss and support
> > upto
> > 3 levels.
> > * Addressed testpmd review comments.
> >
> >   lib/librte_ethdev/rte_ethdev.h | 27 +++++++++++++++++++++++++++
> >   1 file changed, 27 insertions(+)
> >
> > diff --git a/lib/librte_ethdev/rte_ethdev.h
> > b/lib/librte_ethdev/rte_ethdev.h index 70295d7ab..13e49bbd7 100644
> > --- a/lib/librte_ethdev/rte_ethdev.h
> > +++ b/lib/librte_ethdev/rte_ethdev.h
> > @@ -552,6 +552,33 @@ struct rte_eth_rss_conf {
> >   #define RTE_ETH_RSS_L3_PRE64	   (1ULL << 53)
> >   #define RTE_ETH_RSS_L3_PRE96	   (1ULL << 52)
> >
> > +/*
> > + * We use the following macros to combine with the above layers to
> > +choose
> > + * inner and outer layers or both for RSS computation.
> > + * bit 50 and 51 are reserved for this.
> > + */
> > +
> > +/** level 0, requests the default behavior. Depending on the packet
> > + * type, it can mean outermost, innermost, anything in between or even no
> RSS.
> > + * It basically stands for the innermost encapsulation level RSS
> > + * can be performed on according to PMD and device capabilities.
> > + */
> > +#define ETH_RSS_LEVEL_0         (0ULL << 50)
> 
> I can see from history how this is involved, but the 'ETH_RSS_LEVEL_0' naming is
> not really clear what it is, the naming in v6 is more clear.
> 
> What about following one:
> 0 -> LEVEL_PMD_DEFAULT
> 1 -> LEVEL_OUTER
> 2 -> LEVEL_INNER
> 3 -> LEVEL_INNER_OUTER
> 
> This doesn't exactly match to rte_flow one, but closer than v6 one. This ends
> with max level 2. And defines a way to say both inner and outer.

This one looks good to me. If everyone is ok with the proposed changes, I will send V8.

> 
> > +
> > +/** level 1,  requests RSS to be performed on the outermost packet
> > + * encapsulation level.
> > + */
> > +#define ETH_RSS_LEVEL_1         (1ULL << 50)
> > +
> > +/** level 2,  requests RSS to be performed on the
> > + * specified inner packet encapsulation level, from outermost to
> > + * innermost (lower to higher values).
> > + */
> > +#define ETH_RSS_LEVEL_2	        (2ULL << 50)
> 
> I can see you are trying to copy rte_flow usage, but this doesn't really makes
> sense here. Where the value of the level is defined in this case? If not defined
> how the PMD knows which level to use?
> 
> > +#define ETH_RSS_LEVEL_MASK	(3ULL << 50)
> > +
> > +#define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50)
> > +
> >   /**
> >    * For input set change of hash filter, if SRC_ONLY and DST_ONLY of
> >    * the same level are used simultaneously, it is the same case as
> > --
> > 2.25.1
> >
Ferruh Yigit Sept. 1, 2020, 2:44 p.m. UTC | #3
On 9/1/2020 3:27 PM, Kiran Kumar Kokkilagadda wrote:
> 
> 
>> -----Original Message-----
>> From: Ferruh Yigit <ferruh.yigit@intel.com>
>> Sent: Tuesday, September 1, 2020 7:08 PM
>> To: Kiran Kumar Kokkilagadda <kirankumark@marvell.com>; Thomas Monjalon
>> <thomas@monjalon.net>; Andrew Rybchenko <arybchenko@solarflare.com>
>> Cc: dev@dpdk.org; Jerin Jacob Kollanukkaran <jerinj@marvell.com>;
>> orika@mellanox.com; xuanziyang2@huawei.com;
>> cloud.wangxiaoyun@huawei.com; zhouguoyang@huawei.com;
>> rosen.xu@intel.com; beilei.xing@intel.com; jia.guo@intel.com; Rasesh Mody
>> <rmody@marvell.com>; Shahed Shaikh <shshaikh@marvell.com>; Nithin Kumar
>> Dabilpuram <ndabilpuram@marvell.com>; qiming.yang@intel.com;
>> qi.z.zhang@intel.com; keith.wiles@intel.com; hemant.agrawal@nxp.com;
>> sachin.saxena@nxp.com; wei.zhao1@intel.com; johndale@cisco.com;
>> hyonkim@cisco.com; chas3@att.com; matan@mellanox.com;
>> shahafs@mellanox.com; viacheslavo@mellanox.com;
>> rahul.lakkireddy@chelsio.com; grive@u256.net; Liron Himi
>> <lironh@marvell.com>; jingjing.wu@intel.com; xavier.huwei@huawei.com;
>> humin29@huawei.com; yisen.zhuang@huawei.com;
>> ajit.khaparde@broadcom.com; somnath.kotur@broadcom.com;
>> jasvinder.singh@intel.com; cristian.dumitrescu@intel.com
>> Subject: [EXT] Re: [dpdk-dev][PATCH v7 1/3] ethdev: add level support for RSS
>> offload types
>>
>> External Email
>>
>> ----------------------------------------------------------------------
>> On 9/1/2020 4:27 AM, kirankumark@marvell.com wrote:
>>> From: Kiran Kumar K <kirankumark@marvell.com>
>>>
>>> This patch reserves 2 bits as input selection to select Inner and
>>> outer encapsulation level for RSS computation. It is combined with
>>> existing
>>> ETH_RSS_* to choose Inner or outer layers.
>>> This functionality already exists in rte_flow through level parameter
>>> in RSS action configuration rte_flow_action_rss.
>>>
>>> Signed-off-by: Kiran Kumar K <kirankumark@marvell.com>
>>> ---
>>> V7 Changes:
>>> * Re-worked to keep it in sync with rte_flow_action_rss and support
>>> upto
>>> 3 levels.
>>> * Addressed testpmd review comments.
>>>
>>>    lib/librte_ethdev/rte_ethdev.h | 27 +++++++++++++++++++++++++++
>>>    1 file changed, 27 insertions(+)
>>>
>>> diff --git a/lib/librte_ethdev/rte_ethdev.h
>>> b/lib/librte_ethdev/rte_ethdev.h index 70295d7ab..13e49bbd7 100644
>>> --- a/lib/librte_ethdev/rte_ethdev.h
>>> +++ b/lib/librte_ethdev/rte_ethdev.h
>>> @@ -552,6 +552,33 @@ struct rte_eth_rss_conf {
>>>    #define RTE_ETH_RSS_L3_PRE64	   (1ULL << 53)
>>>    #define RTE_ETH_RSS_L3_PRE96	   (1ULL << 52)
>>>
>>> +/*
>>> + * We use the following macros to combine with the above layers to
>>> +choose
>>> + * inner and outer layers or both for RSS computation.
>>> + * bit 50 and 51 are reserved for this.
>>> + */
>>> +
>>> +/** level 0, requests the default behavior. Depending on the packet
>>> + * type, it can mean outermost, innermost, anything in between or even no
>> RSS.
>>> + * It basically stands for the innermost encapsulation level RSS
>>> + * can be performed on according to PMD and device capabilities.
>>> + */
>>> +#define ETH_RSS_LEVEL_0         (0ULL << 50)
>>
>> I can see from history how this is involved, but the 'ETH_RSS_LEVEL_0' naming is
>> not really clear what it is, the naming in v6 is more clear.
>>
>> What about following one:
>> 0 -> LEVEL_PMD_DEFAULT
>> 1 -> LEVEL_OUTER
>> 2 -> LEVEL_INNER
>> 3 -> LEVEL_INNER_OUTER
>>
>> This doesn't exactly match to rte_flow one, but closer than v6 one. This ends
>> with max level 2. And defines a way to say both inner and outer.
> 
> This one looks good to me. If everyone is ok with the proposed changes, I will send V8.
> 

Let's wait for more comments before v8.

>>
>>> +
>>> +/** level 1,  requests RSS to be performed on the outermost packet
>>> + * encapsulation level.
>>> + */
>>> +#define ETH_RSS_LEVEL_1         (1ULL << 50)
>>> +
>>> +/** level 2,  requests RSS to be performed on the
>>> + * specified inner packet encapsulation level, from outermost to
>>> + * innermost (lower to higher values).
>>> + */
>>> +#define ETH_RSS_LEVEL_2	        (2ULL << 50)
>>
>> I can see you are trying to copy rte_flow usage, but this doesn't really makes
>> sense here. Where the value of the level is defined in this case? If not defined
>> how the PMD knows which level to use?
>>
>>> +#define ETH_RSS_LEVEL_MASK	(3ULL << 50)
>>> +
>>> +#define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50)
>>> +
>>>    /**
>>>     * For input set change of hash filter, if SRC_ONLY and DST_ONLY of
>>>     * the same level are used simultaneously, it is the same case as
>>> --
>>> 2.25.1
>>>
>
Ferruh Yigit Sept. 1, 2020, 2:54 p.m. UTC | #4
On 9/1/2020 3:27 PM, Kiran Kumar Kokkilagadda wrote:
> 
> 
>> -----Original Message-----
>> From: Ferruh Yigit <ferruh.yigit@intel.com>
>> Sent: Tuesday, September 1, 2020 7:08 PM
>> To: Kiran Kumar Kokkilagadda <kirankumark@marvell.com>; Thomas Monjalon
>> <thomas@monjalon.net>; Andrew Rybchenko <arybchenko@solarflare.com>
>> Cc: dev@dpdk.org; Jerin Jacob Kollanukkaran <jerinj@marvell.com>;
>> orika@mellanox.com; xuanziyang2@huawei.com;
>> cloud.wangxiaoyun@huawei.com; zhouguoyang@huawei.com;
>> rosen.xu@intel.com; beilei.xing@intel.com; jia.guo@intel.com; Rasesh Mody
>> <rmody@marvell.com>; Shahed Shaikh <shshaikh@marvell.com>; Nithin Kumar
>> Dabilpuram <ndabilpuram@marvell.com>; qiming.yang@intel.com;
>> qi.z.zhang@intel.com; keith.wiles@intel.com; hemant.agrawal@nxp.com;
>> sachin.saxena@nxp.com; wei.zhao1@intel.com; johndale@cisco.com;
>> hyonkim@cisco.com; chas3@att.com; matan@mellanox.com;
>> shahafs@mellanox.com; viacheslavo@mellanox.com;
>> rahul.lakkireddy@chelsio.com; grive@u256.net; Liron Himi
>> <lironh@marvell.com>; jingjing.wu@intel.com; xavier.huwei@huawei.com;
>> humin29@huawei.com; yisen.zhuang@huawei.com;
>> ajit.khaparde@broadcom.com; somnath.kotur@broadcom.com;
>> jasvinder.singh@intel.com; cristian.dumitrescu@intel.com
>> Subject: [EXT] Re: [dpdk-dev][PATCH v7 1/3] ethdev: add level support for RSS
>> offload types
>>
>> External Email
>>
>> ----------------------------------------------------------------------
>> On 9/1/2020 4:27 AM, kirankumark@marvell.com wrote:
>>> From: Kiran Kumar K <kirankumark@marvell.com>
>>>
>>> This patch reserves 2 bits as input selection to select Inner and
>>> outer encapsulation level for RSS computation. It is combined with
>>> existing
>>> ETH_RSS_* to choose Inner or outer layers.
>>> This functionality already exists in rte_flow through level parameter
>>> in RSS action configuration rte_flow_action_rss.
>>>
>>> Signed-off-by: Kiran Kumar K <kirankumark@marvell.com>
>>> ---
>>> V7 Changes:
>>> * Re-worked to keep it in sync with rte_flow_action_rss and support
>>> upto
>>> 3 levels.
>>> * Addressed testpmd review comments.
>>>
>>>    lib/librte_ethdev/rte_ethdev.h | 27 +++++++++++++++++++++++++++
>>>    1 file changed, 27 insertions(+)
>>>
>>> diff --git a/lib/librte_ethdev/rte_ethdev.h
>>> b/lib/librte_ethdev/rte_ethdev.h index 70295d7ab..13e49bbd7 100644
>>> --- a/lib/librte_ethdev/rte_ethdev.h
>>> +++ b/lib/librte_ethdev/rte_ethdev.h
>>> @@ -552,6 +552,33 @@ struct rte_eth_rss_conf {
>>>    #define RTE_ETH_RSS_L3_PRE64	   (1ULL << 53)
>>>    #define RTE_ETH_RSS_L3_PRE96	   (1ULL << 52)
>>>
>>> +/*
>>> + * We use the following macros to combine with the above layers to
>>> +choose
>>> + * inner and outer layers or both for RSS computation.
>>> + * bit 50 and 51 are reserved for this.
>>> + */
>>> +
>>> +/** level 0, requests the default behavior. Depending on the packet
>>> + * type, it can mean outermost, innermost, anything in between or even no
>> RSS.
>>> + * It basically stands for the innermost encapsulation level RSS
>>> + * can be performed on according to PMD and device capabilities.
>>> + */
>>> +#define ETH_RSS_LEVEL_0         (0ULL << 50)
>>
>> I can see from history how this is involved, but the 'ETH_RSS_LEVEL_0' naming is
>> not really clear what it is, the naming in v6 is more clear.
>>
>> What about following one:
>> 0 -> LEVEL_PMD_DEFAULT
>> 1 -> LEVEL_OUTER
>> 2 -> LEVEL_INNER
>> 3 -> LEVEL_INNER_OUTER

Or we can use 'LEVEL_OUTER | LEVEL_INNER' instead of explicitly defining 
'LEVEL_INNER_OUTER'.

>>
>> This doesn't exactly match to rte_flow one, but closer than v6 one. This ends
>> with max level 2. And defines a way to say both inner and outer.
> 
> This one looks good to me. If everyone is ok with the proposed changes, I will send V8.
> 
>>
>>> +
>>> +/** level 1,  requests RSS to be performed on the outermost packet
>>> + * encapsulation level.
>>> + */
>>> +#define ETH_RSS_LEVEL_1         (1ULL << 50)
>>> +
>>> +/** level 2,  requests RSS to be performed on the
>>> + * specified inner packet encapsulation level, from outermost to
>>> + * innermost (lower to higher values).
>>> + */
>>> +#define ETH_RSS_LEVEL_2	        (2ULL << 50)
>>
>> I can see you are trying to copy rte_flow usage, but this doesn't really makes
>> sense here. Where the value of the level is defined in this case? If not defined
>> how the PMD knows which level to use?
>>
>>> +#define ETH_RSS_LEVEL_MASK	(3ULL << 50)
>>> +
>>> +#define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50)
>>> +
>>>    /**
>>>     * For input set change of hash filter, if SRC_ONLY and DST_ONLY of
>>>     * the same level are used simultaneously, it is the same case as
>>> --
>>> 2.25.1
>>>
>
Ajit Khaparde Sept. 1, 2020, 5:11 p.m. UTC | #5
On Tue, Sep 1, 2020 at 7:27 AM Kiran Kumar Kokkilagadda <
kirankumark@marvell.com> wrote:

>
>
> > -----Original Message-----
> > From: Ferruh Yigit <ferruh.yigit@intel.com>
> > Sent: Tuesday, September 1, 2020 7:08 PM
> > To: Kiran Kumar Kokkilagadda <kirankumark@marvell.com>; Thomas Monjalon
> > <thomas@monjalon.net>; Andrew Rybchenko <arybchenko@solarflare.com>
> > Cc: dev@dpdk.org; Jerin Jacob Kollanukkaran <jerinj@marvell.com>;
> > orika@mellanox.com; xuanziyang2@huawei.com;
> > cloud.wangxiaoyun@huawei.com; zhouguoyang@huawei.com;
> > rosen.xu@intel.com; beilei.xing@intel.com; jia.guo@intel.com; Rasesh
> Mody
> > <rmody@marvell.com>; Shahed Shaikh <shshaikh@marvell.com>; Nithin Kumar
> > Dabilpuram <ndabilpuram@marvell.com>; qiming.yang@intel.com;
> > qi.z.zhang@intel.com; keith.wiles@intel.com; hemant.agrawal@nxp.com;
> > sachin.saxena@nxp.com; wei.zhao1@intel.com; johndale@cisco.com;
> > hyonkim@cisco.com; chas3@att.com; matan@mellanox.com;
> > shahafs@mellanox.com; viacheslavo@mellanox.com;
> > rahul.lakkireddy@chelsio.com; grive@u256.net; Liron Himi
> > <lironh@marvell.com>; jingjing.wu@intel.com; xavier.huwei@huawei.com;
> > humin29@huawei.com; yisen.zhuang@huawei.com;
> > ajit.khaparde@broadcom.com; somnath.kotur@broadcom.com;
> > jasvinder.singh@intel.com; cristian.dumitrescu@intel.com
> > Subject: [EXT] Re: [dpdk-dev][PATCH v7 1/3] ethdev: add level support
> for RSS
> > offload types
> >
> > External Email
> >
> > ----------------------------------------------------------------------
> > On 9/1/2020 4:27 AM, kirankumark@marvell.com wrote:
> > > From: Kiran Kumar K <kirankumark@marvell.com>
> > >
> > > This patch reserves 2 bits as input selection to select Inner and
> > > outer encapsulation level for RSS computation. It is combined with
> > > existing
> > > ETH_RSS_* to choose Inner or outer layers.
> > > This functionality already exists in rte_flow through level parameter
> > > in RSS action configuration rte_flow_action_rss.
> > >
> > > Signed-off-by: Kiran Kumar K <kirankumark@marvell.com>
> > > ---
> > > V7 Changes:
> > > * Re-worked to keep it in sync with rte_flow_action_rss and support
> > > upto
> > > 3 levels.
> > > * Addressed testpmd review comments.
> > >
> > >   lib/librte_ethdev/rte_ethdev.h | 27 +++++++++++++++++++++++++++
> > >   1 file changed, 27 insertions(+)
> > >
> > > diff --git a/lib/librte_ethdev/rte_ethdev.h
> > > b/lib/librte_ethdev/rte_ethdev.h index 70295d7ab..13e49bbd7 100644
> > > --- a/lib/librte_ethdev/rte_ethdev.h
> > > +++ b/lib/librte_ethdev/rte_ethdev.h
> > > @@ -552,6 +552,33 @@ struct rte_eth_rss_conf {
> > >   #define RTE_ETH_RSS_L3_PRE64         (1ULL << 53)
> > >   #define RTE_ETH_RSS_L3_PRE96         (1ULL << 52)
> > >
> > > +/*
> > > + * We use the following macros to combine with the above layers to
> > > +choose
> > > + * inner and outer layers or both for RSS computation.
> > > + * bit 50 and 51 are reserved for this.
> > > + */
> > > +
> > > +/** level 0, requests the default behavior. Depending on the packet
> > > + * type, it can mean outermost, innermost, anything in between or
> even no
> > RSS.
> > > + * It basically stands for the innermost encapsulation level RSS
> > > + * can be performed on according to PMD and device capabilities.
> > > + */
> > > +#define ETH_RSS_LEVEL_0         (0ULL << 50)
> >
> > I can see from history how this is involved, but the 'ETH_RSS_LEVEL_0'
> naming is
> > not really clear what it is, the naming in v6 is more clear.
> >
> > What about following one:
> > 0 -> LEVEL_PMD_DEFAULT
> > 1 -> LEVEL_OUTER
> > 2 -> LEVEL_INNER
> > 3 -> LEVEL_INNER_OUTER
> >
> > This doesn't exactly match to rte_flow one, but closer than v6 one. This
> ends
> > with max level 2. And defines a way to say both inner and outer.
>
> This one looks good to me. If everyone is ok with the proposed changes, I
> will send V8.
>


How about following one:
0 -> LEVEL_PMD_DEFAULT
1 -> LEVEL_OUTERMOST
2 -> LEVEL_INNERMOST
This way we can avoid any ambiguity especially if stacked tunnel headers
become real.

3 -> LEVEL_INNER_OUTER
But I am not sure if INNER_OUTER has a use case.

Alternatively,
why not just add uint32_t level;
just like in case of rte_flow_action_rss?

It will break ABI but its 20.11.

Thanks
-Ajit



>
> >
> > > +
> > > +/** level 1,  requests RSS to be performed on the outermost packet
> > > + * encapsulation level.
> > > + */
> > > +#define ETH_RSS_LEVEL_1         (1ULL << 50)
> > > +
> > > +/** level 2,  requests RSS to be performed on the
> > > + * specified inner packet encapsulation level, from outermost to
> > > + * innermost (lower to higher values).
> > > + */
> > > +#define ETH_RSS_LEVEL_2            (2ULL << 50)
> >
> > I can see you are trying to copy rte_flow usage, but this doesn't really
> makes
> > sense here. Where the value of the level is defined in this case? If not
> defined
> > how the PMD knows which level to use?
> >
> > > +#define ETH_RSS_LEVEL_MASK (3ULL << 50)
> > > +
> > > +#define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50)
> > > +
> > >   /**
> > >    * For input set change of hash filter, if SRC_ONLY and DST_ONLY of
> > >    * the same level are used simultaneously, it is the same case as
> > > --
> > > 2.25.1
> > >
>
>
Kiran Kumar Kokkilagadda Sept. 3, 2020, 10:11 a.m. UTC | #6
From: Ajit Khaparde <ajit.khaparde@broadcom.com>
Sent: Tuesday, September 1, 2020 10:42 PM
To: Kiran Kumar Kokkilagadda <kirankumark@marvell.com>
Cc: Ferruh Yigit <ferruh.yigit@intel.com>; Thomas Monjalon <thomas@monjalon.net>; Andrew Rybchenko <arybchenko@solarflare.com>; dev@dpdk.org; Jerin Jacob Kollanukkaran <jerinj@marvell.com>; orika@mellanox.com; xuanziyang2@huawei.com; cloud.wangxiaoyun@huawei.com; zhouguoyang@huawei.com; rosen.xu@intel.com; beilei.xing@intel.com; jia.guo@intel.com; Rasesh Mody <rmody@marvell.com>; Shahed Shaikh <shshaikh@marvell.com>; Nithin Kumar Dabilpuram <ndabilpuram@marvell.com>; qiming.yang@intel.com; qi.z.zhang@intel.com; keith.wiles@intel.com; hemant.agrawal@nxp.com; sachin.saxena@nxp.com; wei.zhao1@intel.com; johndale@cisco.com; hyonkim@cisco.com; chas3@att.com; matan@mellanox.com; shahafs@mellanox.com; viacheslavo@mellanox.com; rahul.lakkireddy@chelsio.com; grive@u256.net; Liron Himi <lironh@marvell.com>; jingjing.wu@intel.com; xavier.huwei@huawei.com; humin29@huawei.com; yisen.zhuang@huawei.com; somnath.kotur@broadcom.com; jasvinder.singh@intel.com; cristian.dumitrescu@intel.com
Subject: Re: [EXT] Re: [dpdk-dev][PATCH v7 1/3] ethdev: add level support for RSS offload types



On Tue, Sep 1, 2020 at 7:27 AM Kiran Kumar Kokkilagadda <kirankumark@marvell.com<mailto:kirankumark@marvell.com>> wrote:


> -----Original Message-----
> From: Ferruh Yigit <ferruh.yigit@intel.com<mailto:ferruh.yigit@intel.com>>
> Sent: Tuesday, September 1, 2020 7:08 PM
> To: Kiran Kumar Kokkilagadda <kirankumark@marvell.com<mailto:kirankumark@marvell.com>>; Thomas Monjalon
> <thomas@monjalon.net<mailto:thomas@monjalon.net>>; Andrew Rybchenko <arybchenko@solarflare.com<mailto:arybchenko@solarflare.com>>
> Cc: dev@dpdk.org<mailto:dev@dpdk.org>; Jerin Jacob Kollanukkaran <jerinj@marvell.com<mailto:jerinj@marvell.com>>;
> orika@mellanox.com<mailto:orika@mellanox.com>; xuanziyang2@huawei.com<mailto:xuanziyang2@huawei.com>;
> cloud.wangxiaoyun@huawei.com<mailto:cloud.wangxiaoyun@huawei.com>; zhouguoyang@huawei.com<mailto:zhouguoyang@huawei.com>;
> rosen.xu@intel.com<mailto:rosen.xu@intel.com>; beilei.xing@intel.com<mailto:beilei.xing@intel.com>; jia.guo@intel.com<mailto:jia.guo@intel.com>; Rasesh Mody
> <rmody@marvell.com<mailto:rmody@marvell.com>>; Shahed Shaikh <shshaikh@marvell.com<mailto:shshaikh@marvell.com>>; Nithin Kumar
> Dabilpuram <ndabilpuram@marvell.com<mailto:ndabilpuram@marvell.com>>; qiming.yang@intel.com<mailto:qiming.yang@intel.com>;
> qi.z.zhang@intel.com<mailto:qi.z.zhang@intel.com>; keith.wiles@intel.com<mailto:keith.wiles@intel.com>; hemant.agrawal@nxp.com<mailto:hemant.agrawal@nxp.com>;
> sachin.saxena@nxp.com<mailto:sachin.saxena@nxp.com>; wei.zhao1@intel.com<mailto:wei.zhao1@intel.com>; johndale@cisco.com<mailto:johndale@cisco.com>;
> hyonkim@cisco.com<mailto:hyonkim@cisco.com>; chas3@att.com<mailto:chas3@att.com>; matan@mellanox.com<mailto:matan@mellanox.com>;
> shahafs@mellanox.com<mailto:shahafs@mellanox.com>; viacheslavo@mellanox.com<mailto:viacheslavo@mellanox.com>;
> rahul.lakkireddy@chelsio.com<mailto:rahul.lakkireddy@chelsio.com>; grive@u256.net<mailto:grive@u256.net>; Liron Himi
> <lironh@marvell.com<mailto:lironh@marvell.com>>; jingjing.wu@intel.com<mailto:jingjing.wu@intel.com>; xavier.huwei@huawei.com<mailto:xavier.huwei@huawei.com>;
> humin29@huawei.com<mailto:humin29@huawei.com>; yisen.zhuang@huawei.com<mailto:yisen.zhuang@huawei.com>;
> ajit.khaparde@broadcom.com<mailto:ajit.khaparde@broadcom.com>; somnath.kotur@broadcom.com<mailto:somnath.kotur@broadcom.com>;
> jasvinder.singh@intel.com<mailto:jasvinder.singh@intel.com>; cristian.dumitrescu@intel.com<mailto:cristian.dumitrescu@intel.com>
> Subject: [EXT] Re: [dpdk-dev][PATCH v7 1/3] ethdev: add level support for RSS
> offload types
>
> External Email
>
> ----------------------------------------------------------------------
> On 9/1/2020 4:27 AM, kirankumark@marvell.com<mailto:kirankumark@marvell.com> wrote:
> > From: Kiran Kumar K <kirankumark@marvell.com<mailto:kirankumark@marvell.com>>
> >
> > This patch reserves 2 bits as input selection to select Inner and
> > outer encapsulation level for RSS computation. It is combined with
> > existing
> > ETH_RSS_* to choose Inner or outer layers.
> > This functionality already exists in rte_flow through level parameter
> > in RSS action configuration rte_flow_action_rss.
> >
> > Signed-off-by: Kiran Kumar K <kirankumark@marvell.com<mailto:kirankumark@marvell.com>>
> > ---
> > V7 Changes:
> > * Re-worked to keep it in sync with rte_flow_action_rss and support
> > upto
> > 3 levels.
> > * Addressed testpmd review comments.
> >
> >   lib/librte_ethdev/rte_ethdev.h | 27 +++++++++++++++++++++++++++
> >   1 file changed, 27 insertions(+)
> >
> > diff --git a/lib/librte_ethdev/rte_ethdev.h
> > b/lib/librte_ethdev/rte_ethdev.h index 70295d7ab..13e49bbd7 100644
> > --- a/lib/librte_ethdev/rte_ethdev.h
> > +++ b/lib/librte_ethdev/rte_ethdev.h
> > @@ -552,6 +552,33 @@ struct rte_eth_rss_conf {
> >   #define RTE_ETH_RSS_L3_PRE64         (1ULL << 53)
> >   #define RTE_ETH_RSS_L3_PRE96         (1ULL << 52)
> >
> > +/*
> > + * We use the following macros to combine with the above layers to
> > +choose
> > + * inner and outer layers or both for RSS computation.
> > + * bit 50 and 51 are reserved for this.
> > + */
> > +
> > +/** level 0, requests the default behavior. Depending on the packet
> > + * type, it can mean outermost, innermost, anything in between or even no
> RSS.
> > + * It basically stands for the innermost encapsulation level RSS
> > + * can be performed on according to PMD and device capabilities.
> > + */
> > +#define ETH_RSS_LEVEL_0         (0ULL << 50)
>
> I can see from history how this is involved, but the 'ETH_RSS_LEVEL_0' naming is
> not really clear what it is, the naming in v6 is more clear.
>
> What about following one:
> 0 -> LEVEL_PMD_DEFAULT
> 1 -> LEVEL_OUTER
> 2 -> LEVEL_INNER
> 3 -> LEVEL_INNER_OUTER
>
> This doesn't exactly match to rte_flow one, but closer than v6 one. This ends
> with max level 2. And defines a way to say both inner and outer.

This one looks good to me. If everyone is ok with the proposed changes, I will send V8.


How about following one:
0 -> LEVEL_PMD_DEFAULT
1 -> LEVEL_OUTERMOST
2 -> LEVEL_INNERMOST
This way we can avoid any ambiguity especially if stacked tunnel headers become real.

3 -> LEVEL_INNER_OUTER
But I am not sure if INNER_OUTER has a use case.

Alternatively,
why not just add uint32_t level;
just like in case of rte_flow_action_rss?

It will break ABI but its 20.11.

Thanks
-Ajit

Can I send V8 with this proposal?
0 -> LEVEL_PMD_DEFAULT
1 -> LEVEL_OUTERMOST
2 -> LEVEL_INNERMOST
If anyone want INNER_OUTER, they can specify LEVEL_OUTERMOST | LEVEL_INNERMOST



>
> > +
> > +/** level 1,  requests RSS to be performed on the outermost packet
> > + * encapsulation level.
> > + */
> > +#define ETH_RSS_LEVEL_1         (1ULL << 50)
> > +
> > +/** level 2,  requests RSS to be performed on the
> > + * specified inner packet encapsulation level, from outermost to
> > + * innermost (lower to higher values).
> > + */
> > +#define ETH_RSS_LEVEL_2            (2ULL << 50)
>
> I can see you are trying to copy rte_flow usage, but this doesn't really makes
> sense here. Where the value of the level is defined in this case? If not defined
> how the PMD knows which level to use?
>
> > +#define ETH_RSS_LEVEL_MASK (3ULL << 50)
> > +
> > +#define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50)
> > +
> >   /**
> >    * For input set change of hash filter, if SRC_ONLY and DST_ONLY of
> >    * the same level are used simultaneously, it is the same case as
> > --
> > 2.25.1
> >
Ferruh Yigit Sept. 3, 2020, 1:14 p.m. UTC | #7
On 9/3/2020 11:11 AM, Kiran Kumar Kokkilagadda wrote:
> *From:* Ajit Khaparde <ajit.khaparde@broadcom.com>
> *Sent:* Tuesday, September 1, 2020 10:42 PM
> *To:* Kiran Kumar Kokkilagadda <kirankumark@marvell.com>
> *Cc:* Ferruh Yigit <ferruh.yigit@intel.com>; Thomas Monjalon 
> <thomas@monjalon.net>; Andrew Rybchenko <arybchenko@solarflare.com>; 
> dev@dpdk.org; Jerin Jacob Kollanukkaran <jerinj@marvell.com>; 
> orika@mellanox.com; xuanziyang2@huawei.com; cloud.wangxiaoyun@huawei.com; 
> zhouguoyang@huawei.com; rosen.xu@intel.com; beilei.xing@intel.com; 
> jia.guo@intel.com; Rasesh Mody <rmody@marvell.com>; Shahed Shaikh 
> <shshaikh@marvell.com>; Nithin Kumar Dabilpuram <ndabilpuram@marvell.com>; 
> qiming.yang@intel.com; qi.z.zhang@intel.com; keith.wiles@intel.com; 
> hemant.agrawal@nxp.com; sachin.saxena@nxp.com; wei.zhao1@intel.com; 
> johndale@cisco.com; hyonkim@cisco.com; chas3@att.com; matan@mellanox.com; 
> shahafs@mellanox.com; viacheslavo@mellanox.com; rahul.lakkireddy@chelsio.com; 
> grive@u256.net; Liron Himi <lironh@marvell.com>; jingjing.wu@intel.com; 
> xavier.huwei@huawei.com; humin29@huawei.com; yisen.zhuang@huawei.com; 
> somnath.kotur@broadcom.com; jasvinder.singh@intel.com; cristian.dumitrescu@intel.com
> *Subject:* Re: [EXT] Re: [dpdk-dev][PATCH v7 1/3] ethdev: add level support for 
> RSS offload types
> 
> On Tue, Sep 1, 2020 at 7:27 AM Kiran Kumar Kokkilagadda <kirankumark@marvell.com 
> <mailto:kirankumark@marvell.com>> wrote:
> 
> 
> 
>      > -----Original Message-----
>      > From: Ferruh Yigit <ferruh.yigit@intel.com <mailto:ferruh.yigit@intel.com>>
>      > Sent: Tuesday, September 1, 2020 7:08 PM
>      > To: Kiran Kumar Kokkilagadda <kirankumark@marvell.com
>     <mailto:kirankumark@marvell.com>>; Thomas Monjalon
>      > <thomas@monjalon.net <mailto:thomas@monjalon.net>>; Andrew Rybchenko
>     <arybchenko@solarflare.com <mailto:arybchenko@solarflare.com>>
>      > Cc: dev@dpdk.org <mailto:dev@dpdk.org>; Jerin Jacob Kollanukkaran
>     <jerinj@marvell.com <mailto:jerinj@marvell.com>>;
>      > orika@mellanox.com <mailto:orika@mellanox.com>; xuanziyang2@huawei.com
>     <mailto:xuanziyang2@huawei.com>;
>      > cloud.wangxiaoyun@huawei.com <mailto:cloud.wangxiaoyun@huawei.com>;
>     zhouguoyang@huawei.com <mailto:zhouguoyang@huawei.com>;
>      > rosen.xu@intel.com <mailto:rosen.xu@intel.com>; beilei.xing@intel.com
>     <mailto:beilei.xing@intel.com>; jia.guo@intel.com
>     <mailto:jia.guo@intel.com>; Rasesh Mody
>      > <rmody@marvell.com <mailto:rmody@marvell.com>>; Shahed Shaikh
>     <shshaikh@marvell.com <mailto:shshaikh@marvell.com>>; Nithin Kumar
>      > Dabilpuram <ndabilpuram@marvell.com <mailto:ndabilpuram@marvell.com>>;
>     qiming.yang@intel.com <mailto:qiming.yang@intel.com>;
>      > qi.z.zhang@intel.com <mailto:qi.z.zhang@intel.com>; keith.wiles@intel.com
>     <mailto:keith.wiles@intel.com>; hemant.agrawal@nxp.com
>     <mailto:hemant.agrawal@nxp.com>;
>      > sachin.saxena@nxp.com <mailto:sachin.saxena@nxp.com>; wei.zhao1@intel.com
>     <mailto:wei.zhao1@intel.com>; johndale@cisco.com <mailto:johndale@cisco.com>;
>      > hyonkim@cisco.com <mailto:hyonkim@cisco.com>; chas3@att.com
>     <mailto:chas3@att.com>; matan@mellanox.com <mailto:matan@mellanox.com>;
>      > shahafs@mellanox.com <mailto:shahafs@mellanox.com>;
>     viacheslavo@mellanox.com <mailto:viacheslavo@mellanox.com>;
>      > rahul.lakkireddy@chelsio.com <mailto:rahul.lakkireddy@chelsio.com>;
>     grive@u256.net <mailto:grive@u256.net>; Liron Himi
>      > <lironh@marvell.com <mailto:lironh@marvell.com>>; jingjing.wu@intel.com
>     <mailto:jingjing.wu@intel.com>; xavier.huwei@huawei.com
>     <mailto:xavier.huwei@huawei.com>;
>      > humin29@huawei.com <mailto:humin29@huawei.com>; yisen.zhuang@huawei.com
>     <mailto:yisen.zhuang@huawei.com>;
>      > ajit.khaparde@broadcom.com <mailto:ajit.khaparde@broadcom.com>;
>     somnath.kotur@broadcom.com <mailto:somnath.kotur@broadcom.com>;
>      > jasvinder.singh@intel.com <mailto:jasvinder.singh@intel.com>;
>     cristian.dumitrescu@intel.com <mailto:cristian.dumitrescu@intel.com>
>      > Subject: [EXT] Re: [dpdk-dev][PATCH v7 1/3] ethdev: add level support for RSS
>      > offload types
>      >
>      > External Email
>      >
>      > ----------------------------------------------------------------------
>      > On 9/1/2020 4:27 AM, kirankumark@marvell.com
>     <mailto:kirankumark@marvell.com> wrote:
>      > > From: Kiran Kumar K <kirankumark@marvell.com
>     <mailto:kirankumark@marvell.com>>
>      > >
>      > > This patch reserves 2 bits as input selection to select Inner and
>      > > outer encapsulation level for RSS computation. It is combined with
>      > > existing
>      > > ETH_RSS_* to choose Inner or outer layers.
>      > > This functionality already exists in rte_flow through level parameter
>      > > in RSS action configuration rte_flow_action_rss.
>      > >
>      > > Signed-off-by: Kiran Kumar K <kirankumark@marvell.com
>     <mailto:kirankumark@marvell.com>>
>      > > ---
>      > > V7 Changes:
>      > > * Re-worked to keep it in sync with rte_flow_action_rss and support
>      > > upto
>      > > 3 levels.
>      > > * Addressed testpmd review comments.
>      > >
>      > >   lib/librte_ethdev/rte_ethdev.h | 27 +++++++++++++++++++++++++++
>      > >   1 file changed, 27 insertions(+)
>      > >
>      > > diff --git a/lib/librte_ethdev/rte_ethdev.h
>      > > b/lib/librte_ethdev/rte_ethdev.h index 70295d7ab..13e49bbd7 100644
>      > > --- a/lib/librte_ethdev/rte_ethdev.h
>      > > +++ b/lib/librte_ethdev/rte_ethdev.h
>      > > @@ -552,6 +552,33 @@ struct rte_eth_rss_conf {
>      > >   #define RTE_ETH_RSS_L3_PRE64         (1ULL << 53)
>      > >   #define RTE_ETH_RSS_L3_PRE96         (1ULL << 52)
>      > >
>      > > +/*
>      > > + * We use the following macros to combine with the above layers to
>      > > +choose
>      > > + * inner and outer layers or both for RSS computation.
>      > > + * bit 50 and 51 are reserved for this.
>      > > + */
>      > > +
>      > > +/** level 0, requests the default behavior. Depending on the packet
>      > > + * type, it can mean outermost, innermost, anything in between or even no
>      > RSS.
>      > > + * It basically stands for the innermost encapsulation level RSS
>      > > + * can be performed on according to PMD and device capabilities.
>      > > + */
>      > > +#define ETH_RSS_LEVEL_0         (0ULL << 50)
>      >
>      > I can see from history how this is involved, but the 'ETH_RSS_LEVEL_0'
>     naming is
>      > not really clear what it is, the naming in v6 is more clear.
>      >
>      > What about following one:
>      > 0 -> LEVEL_PMD_DEFAULT
>      > 1 -> LEVEL_OUTER
>      > 2 -> LEVEL_INNER
>      > 3 -> LEVEL_INNER_OUTER
>      >
>      > This doesn't exactly match to rte_flow one, but closer than v6 one. This ends
>      > with max level 2. And defines a way to say both inner and outer.
> 
>     This one looks good to me. If everyone is ok with the proposed changes, I
>     will send V8.
> 
> How about following one:
> 0 -> LEVEL_PMD_DEFAULT
> 1 -> LEVEL_OUTERMOST
> 2 -> LEVEL_INNERMOST
> 
> This way we can avoid any ambiguity especially if stacked tunnel headersbecome real.
> 
> 
> 3 -> LEVEL_INNER_OUTER
> 
> But I am not sure if INNER_OUTER has a use case.
> 
> Alternatively,
> 
> why not just add uint32_t level;
> 
> just like in case of rte_flow_action_rss?
> 
> It will break ABI but its 20.11.
> 
> Thanks
> 
> -Ajit
> 
> Can I send V8 with this proposal?
> 
> 0 -> LEVEL_PMD_DEFAULT
> 1 -> LEVEL_OUTERMOST
> 2 -> LEVEL_INNERMOST
> 
> If anyone want INNER_OUTER, they can specify LEVEL_OUTERMOST| LEVEL_INNERMOST

+1 to INNERMOST & OUTERMOST, and use "LEVEL_OUTERMOST| LEVEL_INNERMOST" for 
INNER_OUTER.

But the capability reporting is still problematic.
If @Andrew has no objection, I think it is ok to have a v8 and we can continue 
discussion on it.

> 
> 
>      >
>      > > +
>      > > +/** level 1,  requests RSS to be performed on the outermost packet
>      > > + * encapsulation level.
>      > > + */
>      > > +#define ETH_RSS_LEVEL_1         (1ULL << 50)
>      > > +
>      > > +/** level 2,  requests RSS to be performed on the
>      > > + * specified inner packet encapsulation level, from outermost to
>      > > + * innermost (lower to higher values).
>      > > + */
>      > > +#define ETH_RSS_LEVEL_2            (2ULL << 50)
>      >
>      > I can see you are trying to copy rte_flow usage, but this doesn't really
>     makes
>      > sense here. Where the value of the level is defined in this case? If not
>     defined
>      > how the PMD knows which level to use?
>      >
>      > > +#define ETH_RSS_LEVEL_MASK (3ULL << 50)
>      > > +
>      > > +#define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50)
>      > > +
>      > >   /**
>      > >    * For input set change of hash filter, if SRC_ONLY and DST_ONLY of
>      > >    * the same level are used simultaneously, it is the same case as
>      > > --
>      > > 2.25.1
>      > >
>
Andrew Rybchenko Sept. 7, 2020, 8:12 a.m. UTC | #8
On 9/3/20 4:14 PM, Ferruh Yigit wrote:
> On 9/3/2020 11:11 AM, Kiran Kumar Kokkilagadda wrote:
>> *From:* Ajit Khaparde <ajit.khaparde@broadcom.com>
>> *Sent:* Tuesday, September 1, 2020 10:42 PM
>> *To:* Kiran Kumar Kokkilagadda <kirankumark@marvell.com>
>> *Cc:* Ferruh Yigit <ferruh.yigit@intel.com>; Thomas Monjalon
>> <thomas@monjalon.net>; Andrew Rybchenko <arybchenko@solarflare.com>;
>> dev@dpdk.org; Jerin Jacob Kollanukkaran <jerinj@marvell.com>;
>> orika@mellanox.com; xuanziyang2@huawei.com;
>> cloud.wangxiaoyun@huawei.com; zhouguoyang@huawei.com;
>> rosen.xu@intel.com; beilei.xing@intel.com; jia.guo@intel.com; Rasesh
>> Mody <rmody@marvell.com>; Shahed Shaikh <shshaikh@marvell.com>; Nithin
>> Kumar Dabilpuram <ndabilpuram@marvell.com>; qiming.yang@intel.com;
>> qi.z.zhang@intel.com; keith.wiles@intel.com; hemant.agrawal@nxp.com;
>> sachin.saxena@nxp.com; wei.zhao1@intel.com; johndale@cisco.com;
>> hyonkim@cisco.com; chas3@att.com; matan@mellanox.com;
>> shahafs@mellanox.com; viacheslavo@mellanox.com;
>> rahul.lakkireddy@chelsio.com; grive@u256.net; Liron Himi
>> <lironh@marvell.com>; jingjing.wu@intel.com; xavier.huwei@huawei.com;
>> humin29@huawei.com; yisen.zhuang@huawei.com;
>> somnath.kotur@broadcom.com; jasvinder.singh@intel.com;
>> cristian.dumitrescu@intel.com
>> *Subject:* Re: [EXT] Re: [dpdk-dev][PATCH v7 1/3] ethdev: add level
>> support for RSS offload types
>>
>> On Tue, Sep 1, 2020 at 7:27 AM Kiran Kumar Kokkilagadda
>> <kirankumark@marvell.com <mailto:kirankumark@marvell.com>> wrote:
>>
>>
>>
>>      > -----Original Message-----
>>      > From: Ferruh Yigit <ferruh.yigit@intel.com
>> <mailto:ferruh.yigit@intel.com>>
>>      > Sent: Tuesday, September 1, 2020 7:08 PM
>>      > To: Kiran Kumar Kokkilagadda <kirankumark@marvell.com
>>     <mailto:kirankumark@marvell.com>>; Thomas Monjalon
>>      > <thomas@monjalon.net <mailto:thomas@monjalon.net>>; Andrew
>> Rybchenko
>>     <arybchenko@solarflare.com <mailto:arybchenko@solarflare.com>>
>>      > Cc: dev@dpdk.org <mailto:dev@dpdk.org>; Jerin Jacob Kollanukkaran
>>     <jerinj@marvell.com <mailto:jerinj@marvell.com>>;
>>      > orika@mellanox.com <mailto:orika@mellanox.com>;
>> xuanziyang2@huawei.com
>>     <mailto:xuanziyang2@huawei.com>;
>>      > cloud.wangxiaoyun@huawei.com
>> <mailto:cloud.wangxiaoyun@huawei.com>;
>>     zhouguoyang@huawei.com <mailto:zhouguoyang@huawei.com>;
>>      > rosen.xu@intel.com <mailto:rosen.xu@intel.com>;
>> beilei.xing@intel.com
>>     <mailto:beilei.xing@intel.com>; jia.guo@intel.com
>>     <mailto:jia.guo@intel.com>; Rasesh Mody
>>      > <rmody@marvell.com <mailto:rmody@marvell.com>>; Shahed Shaikh
>>     <shshaikh@marvell.com <mailto:shshaikh@marvell.com>>; Nithin Kumar
>>      > Dabilpuram <ndabilpuram@marvell.com
>> <mailto:ndabilpuram@marvell.com>>;
>>     qiming.yang@intel.com <mailto:qiming.yang@intel.com>;
>>      > qi.z.zhang@intel.com <mailto:qi.z.zhang@intel.com>;
>> keith.wiles@intel.com
>>     <mailto:keith.wiles@intel.com>; hemant.agrawal@nxp.com
>>     <mailto:hemant.agrawal@nxp.com>;
>>      > sachin.saxena@nxp.com <mailto:sachin.saxena@nxp.com>;
>> wei.zhao1@intel.com
>>     <mailto:wei.zhao1@intel.com>; johndale@cisco.com
>> <mailto:johndale@cisco.com>;
>>      > hyonkim@cisco.com <mailto:hyonkim@cisco.com>; chas3@att.com
>>     <mailto:chas3@att.com>; matan@mellanox.com
>> <mailto:matan@mellanox.com>;
>>      > shahafs@mellanox.com <mailto:shahafs@mellanox.com>;
>>     viacheslavo@mellanox.com <mailto:viacheslavo@mellanox.com>;
>>      > rahul.lakkireddy@chelsio.com
>> <mailto:rahul.lakkireddy@chelsio.com>;
>>     grive@u256.net <mailto:grive@u256.net>; Liron Himi
>>      > <lironh@marvell.com <mailto:lironh@marvell.com>>;
>> jingjing.wu@intel.com
>>     <mailto:jingjing.wu@intel.com>; xavier.huwei@huawei.com
>>     <mailto:xavier.huwei@huawei.com>;
>>      > humin29@huawei.com <mailto:humin29@huawei.com>;
>> yisen.zhuang@huawei.com
>>     <mailto:yisen.zhuang@huawei.com>;
>>      > ajit.khaparde@broadcom.com <mailto:ajit.khaparde@broadcom.com>;
>>     somnath.kotur@broadcom.com <mailto:somnath.kotur@broadcom.com>;
>>      > jasvinder.singh@intel.com <mailto:jasvinder.singh@intel.com>;
>>     cristian.dumitrescu@intel.com <mailto:cristian.dumitrescu@intel.com>
>>      > Subject: [EXT] Re: [dpdk-dev][PATCH v7 1/3] ethdev: add level
>> support for RSS
>>      > offload types
>>      >
>>      > External Email
>>      >
>>      >
>> ----------------------------------------------------------------------
>>      > On 9/1/2020 4:27 AM, kirankumark@marvell.com
>>     <mailto:kirankumark@marvell.com> wrote:
>>      > > From: Kiran Kumar K <kirankumark@marvell.com
>>     <mailto:kirankumark@marvell.com>>
>>      > >
>>      > > This patch reserves 2 bits as input selection to select Inner
>> and
>>      > > outer encapsulation level for RSS computation. It is combined
>> with
>>      > > existing
>>      > > ETH_RSS_* to choose Inner or outer layers.
>>      > > This functionality already exists in rte_flow through level
>> parameter
>>      > > in RSS action configuration rte_flow_action_rss.
>>      > >
>>      > > Signed-off-by: Kiran Kumar K <kirankumark@marvell.com
>>     <mailto:kirankumark@marvell.com>>
>>      > > ---
>>      > > V7 Changes:
>>      > > * Re-worked to keep it in sync with rte_flow_action_rss and
>> support
>>      > > upto
>>      > > 3 levels.
>>      > > * Addressed testpmd review comments.
>>      > >
>>      > >   lib/librte_ethdev/rte_ethdev.h | 27
>> +++++++++++++++++++++++++++
>>      > >   1 file changed, 27 insertions(+)
>>      > >
>>      > > diff --git a/lib/librte_ethdev/rte_ethdev.h
>>      > > b/lib/librte_ethdev/rte_ethdev.h index 70295d7ab..13e49bbd7
>> 100644
>>      > > --- a/lib/librte_ethdev/rte_ethdev.h
>>      > > +++ b/lib/librte_ethdev/rte_ethdev.h
>>      > > @@ -552,6 +552,33 @@ struct rte_eth_rss_conf {
>>      > >   #define RTE_ETH_RSS_L3_PRE64         (1ULL << 53)
>>      > >   #define RTE_ETH_RSS_L3_PRE96         (1ULL << 52)
>>      > >
>>      > > +/*
>>      > > + * We use the following macros to combine with the above
>> layers to
>>      > > +choose
>>      > > + * inner and outer layers or both for RSS computation.
>>      > > + * bit 50 and 51 are reserved for this.
>>      > > + */
>>      > > +
>>      > > +/** level 0, requests the default behavior. Depending on the
>> packet
>>      > > + * type, it can mean outermost, innermost, anything in
>> between or even no
>>      > RSS.
>>      > > + * It basically stands for the innermost encapsulation level
>> RSS
>>      > > + * can be performed on according to PMD and device
>> capabilities.
>>      > > + */
>>      > > +#define ETH_RSS_LEVEL_0         (0ULL << 50)
>>      >
>>      > I can see from history how this is involved, but the
>> 'ETH_RSS_LEVEL_0'
>>     naming is
>>      > not really clear what it is, the naming in v6 is more clear.
>>      >
>>      > What about following one:
>>      > 0 -> LEVEL_PMD_DEFAULT
>>      > 1 -> LEVEL_OUTER
>>      > 2 -> LEVEL_INNER
>>      > 3 -> LEVEL_INNER_OUTER
>>      >
>>      > This doesn't exactly match to rte_flow one, but closer than v6
>> one. This ends
>>      > with max level 2. And defines a way to say both inner and outer.
>>
>>     This one looks good to me. If everyone is ok with the proposed
>> changes, I
>>     will send V8.
>>
>> How about following one:
>> 0 -> LEVEL_PMD_DEFAULT
>> 1 -> LEVEL_OUTERMOST
>> 2 -> LEVEL_INNERMOST
>>
>> This way we can avoid any ambiguity especially if stacked tunnel
>> headersbecome real.
>>
>>
>> 3 -> LEVEL_INNER_OUTER
>>
>> But I am not sure if INNER_OUTER has a use case.
>>
>> Alternatively,
>>
>> why not just add uint32_t level;
>>
>> just like in case of rte_flow_action_rss?
>>
>> It will break ABI but its 20.11.
>>
>> Thanks
>>
>> -Ajit
>>
>> Can I send V8 with this proposal?
>>
>> 0 -> LEVEL_PMD_DEFAULT
>> 1 -> LEVEL_OUTERMOST
>> 2 -> LEVEL_INNERMOST
>>
>> If anyone want INNER_OUTER, they can specify LEVEL_OUTERMOST|
>> LEVEL_INNERMOST
> 
> +1 to INNERMOST & OUTERMOST, and use "LEVEL_OUTERMOST| LEVEL_INNERMOST"
> for INNER_OUTER.

Frankly speaking I'd drop OUTERMOST | INNERMOST for now in requested RSS
hash config and defined OUTERMOST | INNERMOST in
capabilities as possibility to hash by either INNERMOST or
OUTERMOST headers correspondingly.

> 
> But the capability reporting is still problematic.
> If @Andrew has no objection, I think it is ok to have a v8 and we can
> continue discussion on it.

See above. Number of recognized tunnel levels could be reported
in dev_info, but looks insufficient, since it is interesting
which tunnels are supported (may be even on which level).

>>
>>
>>      >
>>      > > +
>>      > > +/** level 1,  requests RSS to be performed on the outermost
>> packet
>>      > > + * encapsulation level.
>>      > > + */
>>      > > +#define ETH_RSS_LEVEL_1         (1ULL << 50)
>>      > > +
>>      > > +/** level 2,  requests RSS to be performed on the
>>      > > + * specified inner packet encapsulation level, from
>> outermost to
>>      > > + * innermost (lower to higher values).
>>      > > + */
>>      > > +#define ETH_RSS_LEVEL_2            (2ULL << 50)
>>      >
>>      > I can see you are trying to copy rte_flow usage, but this
>> doesn't really
>>     makes
>>      > sense here. Where the value of the level is defined in this
>> case? If not
>>     defined
>>      > how the PMD knows which level to use?
>>      >
>>      > > +#define ETH_RSS_LEVEL_MASK (3ULL << 50)
>>      > > +
>>      > > +#define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK)
>> >> 50)
>>      > > +
>>      > >   /**
>>      > >    * For input set change of hash filter, if SRC_ONLY and
>> DST_ONLY of
>>      > >    * the same level are used simultaneously, it is the same
>> case as
>>      > > --
>>      > > 2.25.1
>>      > >
>>
Ajit Khaparde Sept. 8, 2020, 7:40 p.m. UTC | #9
On Mon, Sep 7, 2020 at 1:12 AM Andrew Rybchenko <arybchenko@solarflare.com>
wrote:

> On 9/3/20 4:14 PM, Ferruh Yigit wrote:
> > On 9/3/2020 11:11 AM, Kiran Kumar Kokkilagadda wrote:
> >> *From:* Ajit Khaparde <ajit.khaparde@broadcom.com>
> >> *Sent:* Tuesday, September 1, 2020 10:42 PM
> >> *To:* Kiran Kumar Kokkilagadda <kirankumark@marvell.com>
> >> *Cc:* Ferruh Yigit <ferruh.yigit@intel.com>; Thomas Monjalon
> >> <thomas@monjalon.net>; Andrew Rybchenko <arybchenko@solarflare.com>;
> >> dev@dpdk.org; Jerin Jacob Kollanukkaran <jerinj@marvell.com>;
> >> orika@mellanox.com; xuanziyang2@huawei.com;
> >> cloud.wangxiaoyun@huawei.com; zhouguoyang@huawei.com;
> >> rosen.xu@intel.com; beilei.xing@intel.com; jia.guo@intel.com; Rasesh
> >> Mody <rmody@marvell.com>; Shahed Shaikh <shshaikh@marvell.com>; Nithin
> >> Kumar Dabilpuram <ndabilpuram@marvell.com>; qiming.yang@intel.com;
> >> qi.z.zhang@intel.com; keith.wiles@intel.com; hemant.agrawal@nxp.com;
> >> sachin.saxena@nxp.com; wei.zhao1@intel.com; johndale@cisco.com;
> >> hyonkim@cisco.com; chas3@att.com; matan@mellanox.com;
> >> shahafs@mellanox.com; viacheslavo@mellanox.com;
> >> rahul.lakkireddy@chelsio.com; grive@u256.net; Liron Himi
> >> <lironh@marvell.com>; jingjing.wu@intel.com; xavier.huwei@huawei.com;
> >> humin29@huawei.com; yisen.zhuang@huawei.com;
> >> somnath.kotur@broadcom.com; jasvinder.singh@intel.com;
> >> cristian.dumitrescu@intel.com
> >> *Subject:* Re: [EXT] Re: [dpdk-dev][PATCH v7 1/3] ethdev: add level
> >> support for RSS offload types
> >>
> >> On Tue, Sep 1, 2020 at 7:27 AM Kiran Kumar Kokkilagadda
> >> <kirankumark@marvell.com <mailto:kirankumark@marvell.com>> wrote:
> >>
> >>
> >>
> >>      > -----Original Message-----
> >>      > From: Ferruh Yigit <ferruh.yigit@intel.com
> >> <mailto:ferruh.yigit@intel.com>>
> >>      > Sent: Tuesday, September 1, 2020 7:08 PM
> >>      > To: Kiran Kumar Kokkilagadda <kirankumark@marvell.com
> >>     <mailto:kirankumark@marvell.com>>; Thomas Monjalon
> >>      > <thomas@monjalon.net <mailto:thomas@monjalon.net>>; Andrew
> >> Rybchenko
> >>     <arybchenko@solarflare.com <mailto:arybchenko@solarflare.com>>
> >>      > Cc: dev@dpdk.org <mailto:dev@dpdk.org>; Jerin Jacob
> Kollanukkaran
> >>     <jerinj@marvell.com <mailto:jerinj@marvell.com>>;
> >>      > orika@mellanox.com <mailto:orika@mellanox.com>;
> >> xuanziyang2@huawei.com
> >>     <mailto:xuanziyang2@huawei.com>;
> >>      > cloud.wangxiaoyun@huawei.com
> >> <mailto:cloud.wangxiaoyun@huawei.com>;
> >>     zhouguoyang@huawei.com <mailto:zhouguoyang@huawei.com>;
> >>      > rosen.xu@intel.com <mailto:rosen.xu@intel.com>;
> >> beilei.xing@intel.com
> >>     <mailto:beilei.xing@intel.com>; jia.guo@intel.com
> >>     <mailto:jia.guo@intel.com>; Rasesh Mody
> >>      > <rmody@marvell.com <mailto:rmody@marvell.com>>; Shahed Shaikh
> >>     <shshaikh@marvell.com <mailto:shshaikh@marvell.com>>; Nithin Kumar
> >>      > Dabilpuram <ndabilpuram@marvell.com
> >> <mailto:ndabilpuram@marvell.com>>;
> >>     qiming.yang@intel.com <mailto:qiming.yang@intel.com>;
> >>      > qi.z.zhang@intel.com <mailto:qi.z.zhang@intel.com>;
> >> keith.wiles@intel.com
> >>     <mailto:keith.wiles@intel.com>; hemant.agrawal@nxp.com
> >>     <mailto:hemant.agrawal@nxp.com>;
> >>      > sachin.saxena@nxp.com <mailto:sachin.saxena@nxp.com>;
> >> wei.zhao1@intel.com
> >>     <mailto:wei.zhao1@intel.com>; johndale@cisco.com
> >> <mailto:johndale@cisco.com>;
> >>      > hyonkim@cisco.com <mailto:hyonkim@cisco.com>; chas3@att.com
> >>     <mailto:chas3@att.com>; matan@mellanox.com
> >> <mailto:matan@mellanox.com>;
> >>      > shahafs@mellanox.com <mailto:shahafs@mellanox.com>;
> >>     viacheslavo@mellanox.com <mailto:viacheslavo@mellanox.com>;
> >>      > rahul.lakkireddy@chelsio.com
> >> <mailto:rahul.lakkireddy@chelsio.com>;
> >>     grive@u256.net <mailto:grive@u256.net>; Liron Himi
> >>      > <lironh@marvell.com <mailto:lironh@marvell.com>>;
> >> jingjing.wu@intel.com
> >>     <mailto:jingjing.wu@intel.com>; xavier.huwei@huawei.com
> >>     <mailto:xavier.huwei@huawei.com>;
> >>      > humin29@huawei.com <mailto:humin29@huawei.com>;
> >> yisen.zhuang@huawei.com
> >>     <mailto:yisen.zhuang@huawei.com>;
> >>      > ajit.khaparde@broadcom.com <mailto:ajit.khaparde@broadcom.com>;
> >>     somnath.kotur@broadcom.com <mailto:somnath.kotur@broadcom.com>;
> >>      > jasvinder.singh@intel.com <mailto:jasvinder.singh@intel.com>;
> >>     cristian.dumitrescu@intel.com <mailto:cristian.dumitrescu@intel.com
> >
> >>      > Subject: [EXT] Re: [dpdk-dev][PATCH v7 1/3] ethdev: add level
> >> support for RSS
> >>      > offload types
> >>      >
> >>      > External Email
> >>      >
> >>      >
> >> ----------------------------------------------------------------------
> >>      > On 9/1/2020 4:27 AM, kirankumark@marvell.com
> >>     <mailto:kirankumark@marvell.com> wrote:
> >>      > > From: Kiran Kumar K <kirankumark@marvell.com
> >>     <mailto:kirankumark@marvell.com>>
> >>      > >
> >>      > > This patch reserves 2 bits as input selection to select Inner
> >> and
> >>      > > outer encapsulation level for RSS computation. It is combined
> >> with
> >>      > > existing
> >>      > > ETH_RSS_* to choose Inner or outer layers.
> >>      > > This functionality already exists in rte_flow through level
> >> parameter
> >>      > > in RSS action configuration rte_flow_action_rss.
> >>      > >
> >>      > > Signed-off-by: Kiran Kumar K <kirankumark@marvell.com
> >>     <mailto:kirankumark@marvell.com>>
> >>      > > ---
> >>      > > V7 Changes:
> >>      > > * Re-worked to keep it in sync with rte_flow_action_rss and
> >> support
> >>      > > upto
> >>      > > 3 levels.
> >>      > > * Addressed testpmd review comments.
> >>      > >
> >>      > >   lib/librte_ethdev/rte_ethdev.h | 27
> >> +++++++++++++++++++++++++++
> >>      > >   1 file changed, 27 insertions(+)
> >>      > >
> >>      > > diff --git a/lib/librte_ethdev/rte_ethdev.h
> >>      > > b/lib/librte_ethdev/rte_ethdev.h index 70295d7ab..13e49bbd7
> >> 100644
> >>      > > --- a/lib/librte_ethdev/rte_ethdev.h
> >>      > > +++ b/lib/librte_ethdev/rte_ethdev.h
> >>      > > @@ -552,6 +552,33 @@ struct rte_eth_rss_conf {
> >>      > >   #define RTE_ETH_RSS_L3_PRE64         (1ULL << 53)
> >>      > >   #define RTE_ETH_RSS_L3_PRE96         (1ULL << 52)
> >>      > >
> >>      > > +/*
> >>      > > + * We use the following macros to combine with the above
> >> layers to
> >>      > > +choose
> >>      > > + * inner and outer layers or both for RSS computation.
> >>      > > + * bit 50 and 51 are reserved for this.
> >>      > > + */
> >>      > > +
> >>      > > +/** level 0, requests the default behavior. Depending on the
> >> packet
> >>      > > + * type, it can mean outermost, innermost, anything in
> >> between or even no
> >>      > RSS.
> >>      > > + * It basically stands for the innermost encapsulation level
> >> RSS
> >>      > > + * can be performed on according to PMD and device
> >> capabilities.
> >>      > > + */
> >>      > > +#define ETH_RSS_LEVEL_0         (0ULL << 50)
> >>      >
> >>      > I can see from history how this is involved, but the
> >> 'ETH_RSS_LEVEL_0'
> >>     naming is
> >>      > not really clear what it is, the naming in v6 is more clear.
> >>      >
> >>      > What about following one:
> >>      > 0 -> LEVEL_PMD_DEFAULT
> >>      > 1 -> LEVEL_OUTER
> >>      > 2 -> LEVEL_INNER
> >>      > 3 -> LEVEL_INNER_OUTER
> >>      >
> >>      > This doesn't exactly match to rte_flow one, but closer than v6
> >> one. This ends
> >>      > with max level 2. And defines a way to say both inner and outer.
> >>
> >>     This one looks good to me. If everyone is ok with the proposed
> >> changes, I
> >>     will send V8.
> >>
> >> How about following one:
> >> 0 -> LEVEL_PMD_DEFAULT
> >> 1 -> LEVEL_OUTERMOST
> >> 2 -> LEVEL_INNERMOST
> >>
> >> This way we can avoid any ambiguity especially if stacked tunnel
> >> headersbecome real.
> >>
> >>
> >> 3 -> LEVEL_INNER_OUTER
> >>
> >> But I am not sure if INNER_OUTER has a use case.
> >>
> >> Alternatively,
> >>
> >> why not just add uint32_t level;
> >>
> >> just like in case of rte_flow_action_rss?
> >>
> >> It will break ABI but its 20.11.
> >>
> >> Thanks
> >>
> >> -Ajit
> >>
> >> Can I send V8 with this proposal?
> >>
> >> 0 -> LEVEL_PMD_DEFAULT
> >> 1 -> LEVEL_OUTERMOST
> >> 2 -> LEVEL_INNERMOST
> >>
> >> If anyone want INNER_OUTER, they can specify LEVEL_OUTERMOST|
> >> LEVEL_INNERMOST
> >
> > +1 to INNERMOST & OUTERMOST, and use "LEVEL_OUTERMOST| LEVEL_INNERMOST"
> > for INNER_OUTER.
>
> Frankly speaking I'd drop OUTERMOST | INNERMOST for now in requested RSS
> hash config and defined OUTERMOST | INNERMOST in
> capabilities as possibility to hash by either INNERMOST or
> OUTERMOST headers correspondingly.
>
> >
> > But the capability reporting is still problematic.
> > If @Andrew has no objection, I think it is ok to have a v8 and we can
> > continue discussion on it.
>
> See above. Number of recognized tunnel levels could be reported
> in dev_info, but looks insufficient, since it is interesting
> which tunnels are supported (may be even on which level).
>
+1.


>
> >>
> >>
> >>      >
> >>      > > +
> >>      > > +/** level 1,  requests RSS to be performed on the outermost
> >> packet
> >>      > > + * encapsulation level.
> >>      > > + */
> >>      > > +#define ETH_RSS_LEVEL_1         (1ULL << 50)
> >>      > > +
> >>      > > +/** level 2,  requests RSS to be performed on the
> >>      > > + * specified inner packet encapsulation level, from
> >> outermost to
> >>      > > + * innermost (lower to higher values).
> >>      > > + */
> >>      > > +#define ETH_RSS_LEVEL_2            (2ULL << 50)
> >>      >
> >>      > I can see you are trying to copy rte_flow usage, but this
> >> doesn't really
> >>     makes
> >>      > sense here. Where the value of the level is defined in this
> >> case? If not
> >>     defined
> >>      > how the PMD knows which level to use?
> >>      >
> >>      > > +#define ETH_RSS_LEVEL_MASK (3ULL << 50)
> >>      > > +
> >>      > > +#define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK)
> >> >> 50)
> >>      > > +
> >>      > >   /**
> >>      > >    * For input set change of hash filter, if SRC_ONLY and
> >> DST_ONLY of
> >>      > >    * the same level are used simultaneously, it is the same
> >> case as
> >>      > > --
> >>      > > 2.25.1
> >>      > >
> >>
>
>
diff mbox series

Patch

diff --git a/lib/librte_ethdev/rte_ethdev.h b/lib/librte_ethdev/rte_ethdev.h
index 70295d7ab..13e49bbd7 100644
--- a/lib/librte_ethdev/rte_ethdev.h
+++ b/lib/librte_ethdev/rte_ethdev.h
@@ -552,6 +552,33 @@  struct rte_eth_rss_conf {
 #define RTE_ETH_RSS_L3_PRE64	   (1ULL << 53)
 #define RTE_ETH_RSS_L3_PRE96	   (1ULL << 52)

+/*
+ * We use the following macros to combine with the above layers to choose
+ * inner and outer layers or both for RSS computation.
+ * bit 50 and 51 are reserved for this.
+ */
+
+/** level 0, requests the default behavior. Depending on the packet
+ * type, it can mean outermost, innermost, anything in between or even no RSS.
+ * It basically stands for the innermost encapsulation level RSS
+ * can be performed on according to PMD and device capabilities.
+ */
+#define ETH_RSS_LEVEL_0         (0ULL << 50)
+
+/** level 1,  requests RSS to be performed on the outermost packet
+ * encapsulation level.
+ */
+#define ETH_RSS_LEVEL_1         (1ULL << 50)
+
+/** level 2,  requests RSS to be performed on the
+ * specified inner packet encapsulation level, from outermost to
+ * innermost (lower to higher values).
+ */
+#define ETH_RSS_LEVEL_2	        (2ULL << 50)
+#define ETH_RSS_LEVEL_MASK	(3ULL << 50)
+
+#define ETH_RSS_LEVEL(rss_hf) ((rss_hf & ETH_RSS_LEVEL_MASK) >> 50)
+
 /**
  * For input set change of hash filter, if SRC_ONLY and DST_ONLY of
  * the same level are used simultaneously, it is the same case as