[dpdk-dev,v4,14/17] mempool: add support to non-EAL thread

  For non-EAL thread, bypass per lcore cache, directly use ring pool.
It allows using rte_mempool in either EAL thread or any user pthread.
As in non-EAL thread, it directly rely on rte_ring and it's none preemptive.
It doesn't suggest to run multi-pthread/cpu which compete the rte_mempool.
It will get bad performance and has critical risk if scheduling policy is RT.

Signed-off-by: Cunming Liang <cunming.liang@intel.com>
---
 lib/librte_mempool/rte_mempool.h | 18 +++++++++++-------
 1 file changed, 11 insertions(+), 7 deletions(-)
  

Hi,

On 02/02/2015 03:02 AM, Cunming Liang wrote:
> For non-EAL thread, bypass per lcore cache, directly use ring pool.
> It allows using rte_mempool in either EAL thread or any user pthread.
> As in non-EAL thread, it directly rely on rte_ring and it's none preemptive.
> It doesn't suggest to run multi-pthread/cpu which compete the rte_mempool.
> It will get bad performance and has critical risk if scheduling policy is RT.
> 
> Signed-off-by: Cunming Liang <cunming.liang@intel.com>
> ---
>  lib/librte_mempool/rte_mempool.h | 18 +++++++++++-------
>  1 file changed, 11 insertions(+), 7 deletions(-)
> 
> diff --git a/lib/librte_mempool/rte_mempool.h b/lib/librte_mempool/rte_mempool.h
> index 3314651..4845f27 100644
> --- a/lib/librte_mempool/rte_mempool.h
> +++ b/lib/librte_mempool/rte_mempool.h
> @@ -198,10 +198,12 @@ struct rte_mempool {
>   *   Number to add to the object-oriented statistics.
>   */
>  #ifdef RTE_LIBRTE_MEMPOOL_DEBUG
> -#define __MEMPOOL_STAT_ADD(mp, name, n) do {			\
> -		unsigned __lcore_id = rte_lcore_id();		\
> -		mp->stats[__lcore_id].name##_objs += n;		\
> -		mp->stats[__lcore_id].name##_bulk += 1;		\
> +#define __MEMPOOL_STAT_ADD(mp, name, n) do {                    \
> +		unsigned __lcore_id = rte_lcore_id();           \
> +		if (__lcore_id < RTE_MAX_LCORE) {               \
> +			mp->stats[__lcore_id].name##_objs += n;	\
> +			mp->stats[__lcore_id].name##_bulk += 1;	\
> +		}                                               \

Does it mean that we have no statistics for non-EAL threads?
(same question for rings and timers in the next patches)


>  	} while(0)
>  #else
>  #define __MEMPOOL_STAT_ADD(mp, name, n) do {} while(0)
> @@ -767,8 +769,9 @@ __mempool_put_bulk(struct rte_mempool *mp, void * const *obj_table,
>  	__MEMPOOL_STAT_ADD(mp, put, n);
>  
>  #if RTE_MEMPOOL_CACHE_MAX_SIZE > 0
> -	/* cache is not enabled or single producer */
> -	if (unlikely(cache_size == 0 || is_mp == 0))
> +	/* cache is not enabled or single producer or none EAL thread */
> +	if (unlikely(cache_size == 0 || is_mp == 0 ||
> +		     lcore_id >= RTE_MAX_LCORE))
>  		goto ring_enqueue;
>  
>  	/* Go straight to ring if put would overflow mem allocated for cache */
> @@ -952,7 +955,8 @@ __mempool_get_bulk(struct rte_mempool *mp, void **obj_table,
>  	uint32_t cache_size = mp->cache_size;
>  
>  	/* cache is not enabled or single consumer */
> -	if (unlikely(cache_size == 0 || is_mc == 0 || n >= cache_size))
> +	if (unlikely(cache_size == 0 || is_mc == 0 ||
> +		     n >= cache_size || lcore_id >= RTE_MAX_LCORE))
>  		goto ring_dequeue;
>  
>  	cache = &mp->local_cache[lcore_id];
> 

What is the performance impact of adding this test?


Regards,
Olivier
  

> -----Original Message-----
> From: Olivier MATZ [mailto:olivier.matz@6wind.com]
> Sent: Monday, February 09, 2015 4:01 AM
> To: Liang, Cunming; dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH v4 14/17] mempool: add support to non-EAL
> thread
> 
> Hi,
> 
> On 02/02/2015 03:02 AM, Cunming Liang wrote:
> > For non-EAL thread, bypass per lcore cache, directly use ring pool.
> > It allows using rte_mempool in either EAL thread or any user pthread.
> > As in non-EAL thread, it directly rely on rte_ring and it's none preemptive.
> > It doesn't suggest to run multi-pthread/cpu which compete the rte_mempool.
> > It will get bad performance and has critical risk if scheduling policy is RT.
> >
> > Signed-off-by: Cunming Liang <cunming.liang@intel.com>
> > ---
> >  lib/librte_mempool/rte_mempool.h | 18 +++++++++++-------
> >  1 file changed, 11 insertions(+), 7 deletions(-)
> >
> > diff --git a/lib/librte_mempool/rte_mempool.h
> b/lib/librte_mempool/rte_mempool.h
> > index 3314651..4845f27 100644
> > --- a/lib/librte_mempool/rte_mempool.h
> > +++ b/lib/librte_mempool/rte_mempool.h
> > @@ -198,10 +198,12 @@ struct rte_mempool {
> >   *   Number to add to the object-oriented statistics.
> >   */
> >  #ifdef RTE_LIBRTE_MEMPOOL_DEBUG
> > -#define __MEMPOOL_STAT_ADD(mp, name, n) do {			\
> > -		unsigned __lcore_id = rte_lcore_id();		\
> > -		mp->stats[__lcore_id].name##_objs += n;		\
> > -		mp->stats[__lcore_id].name##_bulk += 1;		\
> > +#define __MEMPOOL_STAT_ADD(mp, name, n) do {                    \
> > +		unsigned __lcore_id = rte_lcore_id();           \
> > +		if (__lcore_id < RTE_MAX_LCORE) {               \
> > +			mp->stats[__lcore_id].name##_objs += n;	\
> > +			mp->stats[__lcore_id].name##_bulk += 1;	\
> > +		}                                               \
> 
> Does it mean that we have no statistics for non-EAL threads?
> (same question for rings and timers in the next patches)
[LCM] Yes, it is in this patch set, mainly focus on EAL thread and make sure no running issue on non-EAL thread.
For full non-EAL function, will have other patch set to enhance non-EAL thread as the 2nd step.
> 
> 
> >  	} while(0)
> >  #else
> >  #define __MEMPOOL_STAT_ADD(mp, name, n) do {} while(0)
> > @@ -767,8 +769,9 @@ __mempool_put_bulk(struct rte_mempool *mp, void
> * const *obj_table,
> >  	__MEMPOOL_STAT_ADD(mp, put, n);
> >
> >  #if RTE_MEMPOOL_CACHE_MAX_SIZE > 0
> > -	/* cache is not enabled or single producer */
> > -	if (unlikely(cache_size == 0 || is_mp == 0))
> > +	/* cache is not enabled or single producer or none EAL thread */
> > +	if (unlikely(cache_size == 0 || is_mp == 0 ||
> > +		     lcore_id >= RTE_MAX_LCORE))
> >  		goto ring_enqueue;
> >
> >  	/* Go straight to ring if put would overflow mem allocated for cache */
> > @@ -952,7 +955,8 @@ __mempool_get_bulk(struct rte_mempool *mp, void
> **obj_table,
> >  	uint32_t cache_size = mp->cache_size;
> >
> >  	/* cache is not enabled or single consumer */
> > -	if (unlikely(cache_size == 0 || is_mc == 0 || n >= cache_size))
> > +	if (unlikely(cache_size == 0 || is_mc == 0 ||
> > +		     n >= cache_size || lcore_id >= RTE_MAX_LCORE))
> >  		goto ring_dequeue;
> >
> >  	cache = &mp->local_cache[lcore_id];
> >
> 
> What is the performance impact of adding this test?
[LCM] By perf in unit test, it's almost the same. But haven't measure EAL thread and non-EAL thread share the same mempool.
> 
> 
> Regards,
> Olivier
  

Hi,

On 02/09/2015 03:41 PM, Liang, Cunming wrote:
>>>  #ifdef RTE_LIBRTE_MEMPOOL_DEBUG
>>> -#define __MEMPOOL_STAT_ADD(mp, name, n) do {			\
>>> -		unsigned __lcore_id = rte_lcore_id();		\
>>> -		mp->stats[__lcore_id].name##_objs += n;		\
>>> -		mp->stats[__lcore_id].name##_bulk += 1;		\
>>> +#define __MEMPOOL_STAT_ADD(mp, name, n) do {                    \
>>> +		unsigned __lcore_id = rte_lcore_id();           \
>>> +		if (__lcore_id < RTE_MAX_LCORE) {               \
>>> +			mp->stats[__lcore_id].name##_objs += n;	\
>>> +			mp->stats[__lcore_id].name##_bulk += 1;	\
>>> +		}                                               \
>>
>> Does it mean that we have no statistics for non-EAL threads?
>> (same question for rings and timers in the next patches)
> [LCM] Yes, it is in this patch set, mainly focus on EAL thread and make sure no running issue on non-EAL thread.
> For full non-EAL function, will have other patch set to enhance non-EAL thread as the 2nd step.

OK

>>> @@ -952,7 +955,8 @@ __mempool_get_bulk(struct rte_mempool *mp, void
>> **obj_table,
>>>  	uint32_t cache_size = mp->cache_size;
>>>
>>>  	/* cache is not enabled or single consumer */
>>> -	if (unlikely(cache_size == 0 || is_mc == 0 || n >= cache_size))
>>> +	if (unlikely(cache_size == 0 || is_mc == 0 ||
>>> +		     n >= cache_size || lcore_id >= RTE_MAX_LCORE))
>>>  		goto ring_dequeue;
>>>
>>>  	cache = &mp->local_cache[lcore_id];
>>>
>>
>> What is the performance impact of adding this test?
> [LCM] By perf in unit test, it's almost the same. But haven't measure EAL thread and non-EAL thread share the same mempool.


When you say "unit test", are you talking about mempool tests from
"make test"? Do you have some numbers to share?
  

> -----Original Message-----
> From: Olivier MATZ [mailto:olivier.matz@6wind.com]
> Sent: Tuesday, February 10, 2015 1:52 AM
> To: Liang, Cunming; dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH v4 14/17] mempool: add support to non-EAL
> thread
> 
> Hi,
> 
> On 02/09/2015 03:41 PM, Liang, Cunming wrote:
> >>>  #ifdef RTE_LIBRTE_MEMPOOL_DEBUG
> >>> -#define __MEMPOOL_STAT_ADD(mp, name, n) do {			\
> >>> -		unsigned __lcore_id = rte_lcore_id();		\
> >>> -		mp->stats[__lcore_id].name##_objs += n;		\
> >>> -		mp->stats[__lcore_id].name##_bulk += 1;		\
> >>> +#define __MEMPOOL_STAT_ADD(mp, name, n) do {                    \
> >>> +		unsigned __lcore_id = rte_lcore_id();           \
> >>> +		if (__lcore_id < RTE_MAX_LCORE) {               \
> >>> +			mp->stats[__lcore_id].name##_objs += n;	\
> >>> +			mp->stats[__lcore_id].name##_bulk += 1;	\
> >>> +		}                                               \
> >>
> >> Does it mean that we have no statistics for non-EAL threads?
> >> (same question for rings and timers in the next patches)
> > [LCM] Yes, it is in this patch set, mainly focus on EAL thread and make sure no
> running issue on non-EAL thread.
> > For full non-EAL function, will have other patch set to enhance non-EAL thread
> as the 2nd step.
> 
> OK
> 
> >>> @@ -952,7 +955,8 @@ __mempool_get_bulk(struct rte_mempool *mp,
> void
> >> **obj_table,
> >>>  	uint32_t cache_size = mp->cache_size;
> >>>
> >>>  	/* cache is not enabled or single consumer */
> >>> -	if (unlikely(cache_size == 0 || is_mc == 0 || n >= cache_size))
> >>> +	if (unlikely(cache_size == 0 || is_mc == 0 ||
> >>> +		     n >= cache_size || lcore_id >= RTE_MAX_LCORE))
> >>>  		goto ring_dequeue;
> >>>
> >>>  	cache = &mp->local_cache[lcore_id];
> >>>
> >>
> >> What is the performance impact of adding this test?
> > [LCM] By perf in unit test, it's almost the same. But haven't measure EAL thread
> and non-EAL thread share the same mempool.
> 
> 
> When you say "unit test", are you talking about mempool tests from
> "make test"? Do you have some numbers to share?
[LCM] I means DPDK app/test, run mempool_perf_test. Will add numbers on v5.