From: "Wang, Yipeng1" <yipeng1.wang@intel.com>
To: "Medvedkin, Vladimir" <vladimir.medvedkin@intel.com>,
"dev@dpdk.org" <dev@dpdk.org>
Cc: "Gobriel, Sameh" <sameh.gobriel@intel.com>,
"Richardson, Bruce" <bruce.richardson@intel.com>
Subject: Re: [dpdk-dev] [PATCH v2 1/2] hash: add hash bulk lookup with hash signatures array
Date: Mon, 30 Mar 2020 20:20:35 +0000 [thread overview]
Message-ID: <MN2PR11MB4158EA5C591BB1AB0DF633B0C3CB0@MN2PR11MB4158.namprd11.prod.outlook.com> (raw)
In-Reply-To: <1585241243-339118-1-git-send-email-vladimir.medvedkin@intel.com>
> -----Original Message-----
> From: Medvedkin, Vladimir <vladimir.medvedkin@intel.com>
> Sent: Thursday, March 26, 2020 9:47 AM
> To: dev@dpdk.org
> Cc: Wang, Yipeng1 <yipeng1.wang@intel.com>; Gobriel, Sameh
> <sameh.gobriel@intel.com>; Richardson, Bruce
> <bruce.richardson@intel.com>
> Subject: [PATCH v2 1/2] hash: add hash bulk lookup with hash signatures
> array
>
> Implement rte_hash_lookup_with_hash_bulk_data() - lookup function with
> precomputed hash signatures.
>
> Signed-off-by: Vladimir Medvedkin <vladimir.medvedkin@intel.com>
> ---
> lib/librte_hash/rte_cuckoo_hash.c | 296 +++++++++++++++++++++++-----
> -------
> lib/librte_hash/rte_hash.h | 27 ++++
> lib/librte_hash/rte_hash_version.map | 1 +
> 3 files changed, 227 insertions(+), 97 deletions(-)
>
> diff --git a/lib/librte_hash/rte_cuckoo_hash.c
> b/lib/librte_hash/rte_cuckoo_hash.c
> index 6af8ca4..24a0756 100644
> --- a/lib/librte_hash/rte_cuckoo_hash.c
> +++ b/lib/librte_hash/rte_cuckoo_hash.c
> @@ -1711,64 +1711,20 @@ compare_signatures(uint32_t
> *prim_hash_matches, uint32_t *sec_hash_matches,
> }
> }
>
> -#define PREFETCH_OFFSET 4
> static inline void
> -__rte_hash_lookup_bulk_l(const struct rte_hash *h, const void **keys,
> - int32_t num_keys, int32_t *positions,
> - uint64_t *hit_mask, void *data[])
> +__bulk_lookup_l(const struct rte_hash *h, const void **keys,
> + const struct rte_hash_bucket **primary_bkt,
> + const struct rte_hash_bucket **secondary_bkt,
> + uint16_t *sig, int32_t num_keys, int32_t *positions,
> + uint64_t *hit_mask, void *data[])
> {
> uint64_t hits = 0;
> int32_t i;
> int32_t ret;
> - uint32_t prim_hash[RTE_HASH_LOOKUP_BULK_MAX];
> - uint32_t prim_index[RTE_HASH_LOOKUP_BULK_MAX];
> - uint32_t sec_index[RTE_HASH_LOOKUP_BULK_MAX];
> - uint16_t sig[RTE_HASH_LOOKUP_BULK_MAX];
> - const struct rte_hash_bucket
> *primary_bkt[RTE_HASH_LOOKUP_BULK_MAX];
> - const struct rte_hash_bucket
> *secondary_bkt[RTE_HASH_LOOKUP_BULK_MAX];
> uint32_t prim_hitmask[RTE_HASH_LOOKUP_BULK_MAX] = {0};
> uint32_t sec_hitmask[RTE_HASH_LOOKUP_BULK_MAX] = {0};
> struct rte_hash_bucket *cur_bkt, *next_bkt;
>
> - /* Prefetch first keys */
> - for (i = 0; i < PREFETCH_OFFSET && i < num_keys; i++)
> - rte_prefetch0(keys[i]);
> -
> - /*
> - * Prefetch rest of the keys, calculate primary and
> - * secondary bucket and prefetch them
> - */
> - for (i = 0; i < (num_keys - PREFETCH_OFFSET); i++) {
> - rte_prefetch0(keys[i + PREFETCH_OFFSET]);
> -
> - prim_hash[i] = rte_hash_hash(h, keys[i]);
> -
> - sig[i] = get_short_sig(prim_hash[i]);
> - prim_index[i] = get_prim_bucket_index(h, prim_hash[i]);
> - sec_index[i] = get_alt_bucket_index(h, prim_index[i], sig[i]);
> -
> - primary_bkt[i] = &h->buckets[prim_index[i]];
> - secondary_bkt[i] = &h->buckets[sec_index[i]];
> -
> - rte_prefetch0(primary_bkt[i]);
> - rte_prefetch0(secondary_bkt[i]);
> - }
> -
> - /* Calculate and prefetch rest of the buckets */
> - for (; i < num_keys; i++) {
> - prim_hash[i] = rte_hash_hash(h, keys[i]);
> -
> - sig[i] = get_short_sig(prim_hash[i]);
> - prim_index[i] = get_prim_bucket_index(h, prim_hash[i]);
> - sec_index[i] = get_alt_bucket_index(h, prim_index[i], sig[i]);
> -
> - primary_bkt[i] = &h->buckets[prim_index[i]];
> - secondary_bkt[i] = &h->buckets[sec_index[i]];
> -
> - rte_prefetch0(primary_bkt[i]);
> - rte_prefetch0(secondary_bkt[i]);
> - }
> -
> __hash_rw_reader_lock(h);
>
> /* Compare signatures and prefetch key slot of first hit */ @@ -
> 1903,63 +1859,20 @@ __rte_hash_lookup_bulk_l(const struct rte_hash *h,
> const void **keys, }
>
> static inline void
> -__rte_hash_lookup_bulk_lf(const struct rte_hash *h, const void **keys,
> - int32_t num_keys, int32_t *positions,
> - uint64_t *hit_mask, void *data[])
> +__bulk_lookup_lf(const struct rte_hash *h, const void **keys,
> + const struct rte_hash_bucket **primary_bkt,
> + const struct rte_hash_bucket **secondary_bkt,
> + uint16_t *sig, int32_t num_keys, int32_t *positions,
> + uint64_t *hit_mask, void *data[])
> {
> uint64_t hits = 0;
> int32_t i;
> int32_t ret;
> - uint32_t prim_hash[RTE_HASH_LOOKUP_BULK_MAX];
> - uint32_t prim_index[RTE_HASH_LOOKUP_BULK_MAX];
> - uint32_t sec_index[RTE_HASH_LOOKUP_BULK_MAX];
> - uint16_t sig[RTE_HASH_LOOKUP_BULK_MAX];
> - const struct rte_hash_bucket
> *primary_bkt[RTE_HASH_LOOKUP_BULK_MAX];
> - const struct rte_hash_bucket
> *secondary_bkt[RTE_HASH_LOOKUP_BULK_MAX];
> uint32_t prim_hitmask[RTE_HASH_LOOKUP_BULK_MAX] = {0};
> uint32_t sec_hitmask[RTE_HASH_LOOKUP_BULK_MAX] = {0};
> struct rte_hash_bucket *cur_bkt, *next_bkt;
> uint32_t cnt_b, cnt_a;
>
> - /* Prefetch first keys */
> - for (i = 0; i < PREFETCH_OFFSET && i < num_keys; i++)
> - rte_prefetch0(keys[i]);
> -
> - /*
> - * Prefetch rest of the keys, calculate primary and
> - * secondary bucket and prefetch them
> - */
> - for (i = 0; i < (num_keys - PREFETCH_OFFSET); i++) {
> - rte_prefetch0(keys[i + PREFETCH_OFFSET]);
> -
> - prim_hash[i] = rte_hash_hash(h, keys[i]);
> -
> - sig[i] = get_short_sig(prim_hash[i]);
> - prim_index[i] = get_prim_bucket_index(h, prim_hash[i]);
> - sec_index[i] = get_alt_bucket_index(h, prim_index[i], sig[i]);
> -
> - primary_bkt[i] = &h->buckets[prim_index[i]];
> - secondary_bkt[i] = &h->buckets[sec_index[i]];
> -
> - rte_prefetch0(primary_bkt[i]);
> - rte_prefetch0(secondary_bkt[i]);
> - }
> -
> - /* Calculate and prefetch rest of the buckets */
> - for (; i < num_keys; i++) {
> - prim_hash[i] = rte_hash_hash(h, keys[i]);
> -
> - sig[i] = get_short_sig(prim_hash[i]);
> - prim_index[i] = get_prim_bucket_index(h, prim_hash[i]);
> - sec_index[i] = get_alt_bucket_index(h, prim_index[i], sig[i]);
> -
> - primary_bkt[i] = &h->buckets[prim_index[i]];
> - secondary_bkt[i] = &h->buckets[sec_index[i]];
> -
> - rte_prefetch0(primary_bkt[i]);
> - rte_prefetch0(secondary_bkt[i]);
> - }
> -
> for (i = 0; i < num_keys; i++)
> positions[i] = -ENOENT;
>
> @@ -2124,6 +2037,92 @@ __rte_hash_lookup_bulk_lf(const struct rte_hash
> *h, const void **keys,
> *hit_mask = hits;
> }
>
> +#define PREFETCH_OFFSET 4
> +static inline void
> +__bulk_lookup_prefetching_loop(const struct rte_hash *h,
> + const void **keys, int32_t num_keys,
> + uint16_t *sig,
> + const struct rte_hash_bucket **primary_bkt,
> + const struct rte_hash_bucket **secondary_bkt) {
> + int32_t i;
> + uint32_t prim_hash[RTE_HASH_LOOKUP_BULK_MAX];
> + uint32_t prim_index[RTE_HASH_LOOKUP_BULK_MAX];
> + uint32_t sec_index[RTE_HASH_LOOKUP_BULK_MAX];
> +
> + /* Prefetch first keys */
> + for (i = 0; i < PREFETCH_OFFSET && i < num_keys; i++)
> + rte_prefetch0(keys[i]);
> +
> + /*
> + * Prefetch rest of the keys, calculate primary and
> + * secondary bucket and prefetch them
> + */
> + for (i = 0; i < (num_keys - PREFETCH_OFFSET); i++) {
> + rte_prefetch0(keys[i + PREFETCH_OFFSET]);
> +
> + prim_hash[i] = rte_hash_hash(h, keys[i]);
> +
> + sig[i] = get_short_sig(prim_hash[i]);
> + prim_index[i] = get_prim_bucket_index(h, prim_hash[i]);
> + sec_index[i] = get_alt_bucket_index(h, prim_index[i], sig[i]);
> +
> + primary_bkt[i] = &h->buckets[prim_index[i]];
> + secondary_bkt[i] = &h->buckets[sec_index[i]];
> +
> + rte_prefetch0(primary_bkt[i]);
> + rte_prefetch0(secondary_bkt[i]);
> + }
> +
> + /* Calculate and prefetch rest of the buckets */
> + for (; i < num_keys; i++) {
> + prim_hash[i] = rte_hash_hash(h, keys[i]);
> +
> + sig[i] = get_short_sig(prim_hash[i]);
> + prim_index[i] = get_prim_bucket_index(h, prim_hash[i]);
> + sec_index[i] = get_alt_bucket_index(h, prim_index[i], sig[i]);
> +
> + primary_bkt[i] = &h->buckets[prim_index[i]];
> + secondary_bkt[i] = &h->buckets[sec_index[i]];
> +
> + rte_prefetch0(primary_bkt[i]);
> + rte_prefetch0(secondary_bkt[i]);
> + }
> +}
> +
> +
[Wang, Yipeng] Here is an unnecessary blank line.
> +static inline void
> +__rte_hash_lookup_bulk_l(const struct rte_hash *h, const void **keys,
> + int32_t num_keys, int32_t *positions,
> + uint64_t *hit_mask, void *data[])
> +{
> + uint16_t sig[RTE_HASH_LOOKUP_BULK_MAX];
> + const struct rte_hash_bucket
> *primary_bkt[RTE_HASH_LOOKUP_BULK_MAX];
[Wang, Yipeng]
Thanks for revising the code to be more concise.
It looks so far so good to me.
BTW, would you like to add also the rte_hash_lookup_with_hash_bulk function in this patchset?
Just for the completeness.
Thanks
Yipeng
next prev parent reply other threads:[~2020-03-30 20:20 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-03-09 12:44 [dpdk-dev] [PATCH " Vladimir Medvedkin
2020-03-09 12:44 ` [dpdk-dev] [PATCH 2/2] test: update hash performance tests Vladimir Medvedkin
2020-03-17 17:27 ` [dpdk-dev] [PATCH 1/2] hash: add hash bulk lookup with hash signatures array Wang, Yipeng1
2020-03-26 14:16 ` Medvedkin, Vladimir
2020-03-26 16:47 ` [dpdk-dev] [PATCH v2 " Vladimir Medvedkin
2020-03-30 20:20 ` Wang, Yipeng1 [this message]
2020-04-08 18:32 ` [dpdk-dev] [PATCH v3 " Vladimir Medvedkin
2020-04-16 9:47 ` Thomas Monjalon
2020-04-16 15:00 ` [dpdk-dev] [PATCH v4] " Vladimir Medvedkin
2020-04-16 15:07 ` [dpdk-dev] [PATCH v5] " Vladimir Medvedkin
2020-04-16 23:46 ` Wang, Yipeng1
2020-04-25 13:30 ` Thomas Monjalon
2020-04-25 13:37 ` Thomas Monjalon
2020-04-08 18:32 ` [dpdk-dev] [PATCH v3 2/2] test: update hash performance tests Vladimir Medvedkin
2020-04-16 9:44 ` Thomas Monjalon
2020-04-16 14:47 ` Medvedkin, Vladimir
2020-04-16 14:50 ` Thomas Monjalon
2020-03-26 16:47 ` [dpdk-dev] [PATCH v2 " Vladimir Medvedkin
2020-03-30 20:33 ` Wang, Yipeng1
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=MN2PR11MB4158EA5C591BB1AB0DF633B0C3CB0@MN2PR11MB4158.namprd11.prod.outlook.com \
--to=yipeng1.wang@intel.com \
--cc=bruce.richardson@intel.com \
--cc=dev@dpdk.org \
--cc=sameh.gobriel@intel.com \
--cc=vladimir.medvedkin@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).