* [dpdk-dev] [PATCH v2] mempool: limit cache_size
@ 2015-05-18 15:35 Zoltan Kiss
2015-05-18 15:51 ` Ananyev, Konstantin
2015-05-20 8:32 ` Olivier MATZ
0 siblings, 2 replies; 4+ messages in thread
From: Zoltan Kiss @ 2015-05-18 15:35 UTC (permalink / raw)
To: dev
Otherwise cache_flushthresh can be bigger than n, and
a consumer can starve others by keeping every element
either in use or in the cache.
Signed-off-by: Zoltan Kiss <zoltan.kiss@linaro.org>
---
v2: use macro for calculation, with proper casting
lib/librte_mempool/rte_mempool.c | 8 +++++---
lib/librte_mempool/rte_mempool.h | 2 +-
2 files changed, 6 insertions(+), 4 deletions(-)
diff --git a/lib/librte_mempool/rte_mempool.c b/lib/librte_mempool/rte_mempool.c
index cf7ed76..5cfb96b 100644
--- a/lib/librte_mempool/rte_mempool.c
+++ b/lib/librte_mempool/rte_mempool.c
@@ -68,6 +68,8 @@ static struct rte_tailq_elem rte_mempool_tailq = {
EAL_REGISTER_TAILQ(rte_mempool_tailq)
#define CACHE_FLUSHTHRESH_MULTIPLIER 1.5
+#define CALC_CACHE_FLUSHTHRESH(c) \
+ ((typeof (c))((c) * CACHE_FLUSHTHRESH_MULTIPLIER))
/*
* return the greatest common divisor between a and b (fast algorithm)
@@ -440,7 +442,8 @@ rte_mempool_xmem_create(const char *name, unsigned n, unsigned elt_size,
mempool_list = RTE_TAILQ_CAST(rte_mempool_tailq.head, rte_mempool_list);
/* asked cache too big */
- if (cache_size > RTE_MEMPOOL_CACHE_MAX_SIZE) {
+ if (cache_size > RTE_MEMPOOL_CACHE_MAX_SIZE ||
+ CALC_CACHE_FLUSHTHRESH(cache_size) > n) {
rte_errno = EINVAL;
return NULL;
}
@@ -565,8 +568,7 @@ rte_mempool_xmem_create(const char *name, unsigned n, unsigned elt_size,
mp->header_size = objsz.header_size;
mp->trailer_size = objsz.trailer_size;
mp->cache_size = cache_size;
- mp->cache_flushthresh = (uint32_t)
- (cache_size * CACHE_FLUSHTHRESH_MULTIPLIER);
+ mp->cache_flushthresh = CALC_CACHE_FLUSHTHRESH(cache_size);
mp->private_data_size = private_data_size;
/* calculate address of the first element for continuous mempool. */
diff --git a/lib/librte_mempool/rte_mempool.h b/lib/librte_mempool/rte_mempool.h
index 9001312..a4a9610 100644
--- a/lib/librte_mempool/rte_mempool.h
+++ b/lib/librte_mempool/rte_mempool.h
@@ -468,7 +468,7 @@ typedef void (rte_mempool_ctor_t)(struct rte_mempool *, void *);
* If cache_size is non-zero, the rte_mempool library will try to
* limit the accesses to the common lockless pool, by maintaining a
* per-lcore object cache. This argument must be lower or equal to
- * CONFIG_RTE_MEMPOOL_CACHE_MAX_SIZE. It is advised to choose
+ * CONFIG_RTE_MEMPOOL_CACHE_MAX_SIZE and n / 1.5. It is advised to choose
* cache_size to have "n modulo cache_size == 0": if this is
* not the case, some elements will always stay in the pool and will
* never be used. The access to the per-lcore table is of course
--
1.9.1
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [dpdk-dev] [PATCH v2] mempool: limit cache_size
2015-05-18 15:35 [dpdk-dev] [PATCH v2] mempool: limit cache_size Zoltan Kiss
@ 2015-05-18 15:51 ` Ananyev, Konstantin
2015-05-20 8:32 ` Olivier MATZ
1 sibling, 0 replies; 4+ messages in thread
From: Ananyev, Konstantin @ 2015-05-18 15:51 UTC (permalink / raw)
To: Zoltan Kiss, dev
> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Zoltan Kiss
> Sent: Monday, May 18, 2015 4:35 PM
> To: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH v2] mempool: limit cache_size
>
> Otherwise cache_flushthresh can be bigger than n, and
> a consumer can starve others by keeping every element
> either in use or in the cache.
>
> Signed-off-by: Zoltan Kiss <zoltan.kiss@linaro.org>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
> ---
> v2: use macro for calculation, with proper casting
>
> lib/librte_mempool/rte_mempool.c | 8 +++++---
> lib/librte_mempool/rte_mempool.h | 2 +-
> 2 files changed, 6 insertions(+), 4 deletions(-)
>
> diff --git a/lib/librte_mempool/rte_mempool.c b/lib/librte_mempool/rte_mempool.c
> index cf7ed76..5cfb96b 100644
> --- a/lib/librte_mempool/rte_mempool.c
> +++ b/lib/librte_mempool/rte_mempool.c
> @@ -68,6 +68,8 @@ static struct rte_tailq_elem rte_mempool_tailq = {
> EAL_REGISTER_TAILQ(rte_mempool_tailq)
>
> #define CACHE_FLUSHTHRESH_MULTIPLIER 1.5
> +#define CALC_CACHE_FLUSHTHRESH(c) \
> + ((typeof (c))((c) * CACHE_FLUSHTHRESH_MULTIPLIER))
>
> /*
> * return the greatest common divisor between a and b (fast algorithm)
> @@ -440,7 +442,8 @@ rte_mempool_xmem_create(const char *name, unsigned n, unsigned elt_size,
> mempool_list = RTE_TAILQ_CAST(rte_mempool_tailq.head, rte_mempool_list);
>
> /* asked cache too big */
> - if (cache_size > RTE_MEMPOOL_CACHE_MAX_SIZE) {
> + if (cache_size > RTE_MEMPOOL_CACHE_MAX_SIZE ||
> + CALC_CACHE_FLUSHTHRESH(cache_size) > n) {
> rte_errno = EINVAL;
> return NULL;
> }
> @@ -565,8 +568,7 @@ rte_mempool_xmem_create(const char *name, unsigned n, unsigned elt_size,
> mp->header_size = objsz.header_size;
> mp->trailer_size = objsz.trailer_size;
> mp->cache_size = cache_size;
> - mp->cache_flushthresh = (uint32_t)
> - (cache_size * CACHE_FLUSHTHRESH_MULTIPLIER);
> + mp->cache_flushthresh = CALC_CACHE_FLUSHTHRESH(cache_size);
> mp->private_data_size = private_data_size;
>
> /* calculate address of the first element for continuous mempool. */
> diff --git a/lib/librte_mempool/rte_mempool.h b/lib/librte_mempool/rte_mempool.h
> index 9001312..a4a9610 100644
> --- a/lib/librte_mempool/rte_mempool.h
> +++ b/lib/librte_mempool/rte_mempool.h
> @@ -468,7 +468,7 @@ typedef void (rte_mempool_ctor_t)(struct rte_mempool *, void *);
> * If cache_size is non-zero, the rte_mempool library will try to
> * limit the accesses to the common lockless pool, by maintaining a
> * per-lcore object cache. This argument must be lower or equal to
> - * CONFIG_RTE_MEMPOOL_CACHE_MAX_SIZE. It is advised to choose
> + * CONFIG_RTE_MEMPOOL_CACHE_MAX_SIZE and n / 1.5. It is advised to choose
> * cache_size to have "n modulo cache_size == 0": if this is
> * not the case, some elements will always stay in the pool and will
> * never be used. The access to the per-lcore table is of course
> --
> 1.9.1
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [dpdk-dev] [PATCH v2] mempool: limit cache_size
2015-05-18 15:35 [dpdk-dev] [PATCH v2] mempool: limit cache_size Zoltan Kiss
2015-05-18 15:51 ` Ananyev, Konstantin
@ 2015-05-20 8:32 ` Olivier MATZ
2015-05-20 8:46 ` Thomas Monjalon
1 sibling, 1 reply; 4+ messages in thread
From: Olivier MATZ @ 2015-05-20 8:32 UTC (permalink / raw)
To: Zoltan Kiss, dev
On 05/18/2015 05:35 PM, Zoltan Kiss wrote:
> Otherwise cache_flushthresh can be bigger than n, and
> a consumer can starve others by keeping every element
> either in use or in the cache.
>
> Signed-off-by: Zoltan Kiss <zoltan.kiss@linaro.org>
Acked-by: Olivier Matz <olivier.matz@6wind.com>
> ---
> v2: use macro for calculation, with proper casting
>
> lib/librte_mempool/rte_mempool.c | 8 +++++---
> lib/librte_mempool/rte_mempool.h | 2 +-
> 2 files changed, 6 insertions(+), 4 deletions(-)
>
> diff --git a/lib/librte_mempool/rte_mempool.c b/lib/librte_mempool/rte_mempool.c
> index cf7ed76..5cfb96b 100644
> --- a/lib/librte_mempool/rte_mempool.c
> +++ b/lib/librte_mempool/rte_mempool.c
> @@ -68,6 +68,8 @@ static struct rte_tailq_elem rte_mempool_tailq = {
> EAL_REGISTER_TAILQ(rte_mempool_tailq)
>
> #define CACHE_FLUSHTHRESH_MULTIPLIER 1.5
> +#define CALC_CACHE_FLUSHTHRESH(c) \
> + ((typeof (c))((c) * CACHE_FLUSHTHRESH_MULTIPLIER))
>
> /*
> * return the greatest common divisor between a and b (fast algorithm)
> @@ -440,7 +442,8 @@ rte_mempool_xmem_create(const char *name, unsigned n, unsigned elt_size,
> mempool_list = RTE_TAILQ_CAST(rte_mempool_tailq.head, rte_mempool_list);
>
> /* asked cache too big */
> - if (cache_size > RTE_MEMPOOL_CACHE_MAX_SIZE) {
> + if (cache_size > RTE_MEMPOOL_CACHE_MAX_SIZE ||
> + CALC_CACHE_FLUSHTHRESH(cache_size) > n) {
> rte_errno = EINVAL;
> return NULL;
> }
> @@ -565,8 +568,7 @@ rte_mempool_xmem_create(const char *name, unsigned n, unsigned elt_size,
> mp->header_size = objsz.header_size;
> mp->trailer_size = objsz.trailer_size;
> mp->cache_size = cache_size;
> - mp->cache_flushthresh = (uint32_t)
> - (cache_size * CACHE_FLUSHTHRESH_MULTIPLIER);
> + mp->cache_flushthresh = CALC_CACHE_FLUSHTHRESH(cache_size);
> mp->private_data_size = private_data_size;
>
> /* calculate address of the first element for continuous mempool. */
> diff --git a/lib/librte_mempool/rte_mempool.h b/lib/librte_mempool/rte_mempool.h
> index 9001312..a4a9610 100644
> --- a/lib/librte_mempool/rte_mempool.h
> +++ b/lib/librte_mempool/rte_mempool.h
> @@ -468,7 +468,7 @@ typedef void (rte_mempool_ctor_t)(struct rte_mempool *, void *);
> * If cache_size is non-zero, the rte_mempool library will try to
> * limit the accesses to the common lockless pool, by maintaining a
> * per-lcore object cache. This argument must be lower or equal to
> - * CONFIG_RTE_MEMPOOL_CACHE_MAX_SIZE. It is advised to choose
> + * CONFIG_RTE_MEMPOOL_CACHE_MAX_SIZE and n / 1.5. It is advised to choose
> * cache_size to have "n modulo cache_size == 0": if this is
> * not the case, some elements will always stay in the pool and will
> * never be used. The access to the per-lcore table is of course
>
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [dpdk-dev] [PATCH v2] mempool: limit cache_size
2015-05-20 8:32 ` Olivier MATZ
@ 2015-05-20 8:46 ` Thomas Monjalon
0 siblings, 0 replies; 4+ messages in thread
From: Thomas Monjalon @ 2015-05-20 8:46 UTC (permalink / raw)
To: Zoltan Kiss; +Cc: dev
2015-05-20 10:32, Olivier MATZ:
> On 05/18/2015 05:35 PM, Zoltan Kiss wrote:
> > Otherwise cache_flushthresh can be bigger than n, and
> > a consumer can starve others by keeping every element
> > either in use or in the cache.
> >
> > Signed-off-by: Zoltan Kiss <zoltan.kiss@linaro.org>
>
> Acked-by: Olivier Matz <olivier.matz@6wind.com>
Applied, thanks
> > +#define CALC_CACHE_FLUSHTHRESH(c) \
> > + ((typeof (c))((c) * CACHE_FLUSHTHRESH_MULTIPLIER))
I fixed spacing here:
+ ((typeof(c))((c) * CACHE_FLUSHTHRESH_MULTIPLIER))
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2015-05-20 8:46 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2015-05-18 15:35 [dpdk-dev] [PATCH v2] mempool: limit cache_size Zoltan Kiss
2015-05-18 15:51 ` Ananyev, Konstantin
2015-05-20 8:32 ` Olivier MATZ
2015-05-20 8:46 ` Thomas Monjalon
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).