From 246779dd090bd1b74d2652b3a6ca7759f593b27a Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Thu, 18 Aug 2016 16:50:56 +0800 Subject: rhashtable: Remove GFP flag from rhashtable_walk_init The commit 8f6fd83c6c5ec66a4a70c728535ddcdfef4f3697 ("rhashtable: accept GFP flags in rhashtable_walk_init") added a GFP flag argument to rhashtable_walk_init because some users wish to use the walker in an unsleepable context. In fact we don't need to allocate memory in rhashtable_walk_init at all. The walker is always paired with an iterator so we could just stash ourselves there. This patch does that by introducing a new enter function to replace the existing init function. This way we don't have to churn all the existing users again. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller --- lib/rhashtable.c | 46 ++++++++++++++++++---------------------------- 1 file changed, 18 insertions(+), 28 deletions(-) (limited to 'lib') diff --git a/lib/rhashtable.c b/lib/rhashtable.c index 5ba520b544d7..97e3cf08142c 100644 --- a/lib/rhashtable.c +++ b/lib/rhashtable.c @@ -489,10 +489,9 @@ exit: EXPORT_SYMBOL_GPL(rhashtable_insert_slow); /** - * rhashtable_walk_init - Initialise an iterator + * rhashtable_walk_enter - Initialise an iterator * @ht: Table to walk over * @iter: Hash table Iterator - * @gfp: GFP flags for allocations * * This function prepares a hash table walk. * @@ -507,30 +506,22 @@ EXPORT_SYMBOL_GPL(rhashtable_insert_slow); * This function may sleep so you must not call it from interrupt * context or with spin locks held. * - * You must call rhashtable_walk_exit if this function returns - * successfully. + * You must call rhashtable_walk_exit after this function returns. */ -int rhashtable_walk_init(struct rhashtable *ht, struct rhashtable_iter *iter, - gfp_t gfp) +void rhashtable_walk_enter(struct rhashtable *ht, struct rhashtable_iter *iter) { iter->ht = ht; iter->p = NULL; iter->slot = 0; iter->skip = 0; - iter->walker = kmalloc(sizeof(*iter->walker), gfp); - if (!iter->walker) - return -ENOMEM; - spin_lock(&ht->lock); - iter->walker->tbl = + iter->walker.tbl = rcu_dereference_protected(ht->tbl, lockdep_is_held(&ht->lock)); - list_add(&iter->walker->list, &iter->walker->tbl->walkers); + list_add(&iter->walker.list, &iter->walker.tbl->walkers); spin_unlock(&ht->lock); - - return 0; } -EXPORT_SYMBOL_GPL(rhashtable_walk_init); +EXPORT_SYMBOL_GPL(rhashtable_walk_enter); /** * rhashtable_walk_exit - Free an iterator @@ -541,10 +532,9 @@ EXPORT_SYMBOL_GPL(rhashtable_walk_init); void rhashtable_walk_exit(struct rhashtable_iter *iter) { spin_lock(&iter->ht->lock); - if (iter->walker->tbl) - list_del(&iter->walker->list); + if (iter->walker.tbl) + list_del(&iter->walker.list); spin_unlock(&iter->ht->lock); - kfree(iter->walker); } EXPORT_SYMBOL_GPL(rhashtable_walk_exit); @@ -570,12 +560,12 @@ int rhashtable_walk_start(struct rhashtable_iter *iter) rcu_read_lock(); spin_lock(&ht->lock); - if (iter->walker->tbl) - list_del(&iter->walker->list); + if (iter->walker.tbl) + list_del(&iter->walker.list); spin_unlock(&ht->lock); - if (!iter->walker->tbl) { - iter->walker->tbl = rht_dereference_rcu(ht->tbl, ht); + if (!iter->walker.tbl) { + iter->walker.tbl = rht_dereference_rcu(ht->tbl, ht); return -EAGAIN; } @@ -597,7 +587,7 @@ EXPORT_SYMBOL_GPL(rhashtable_walk_start); */ void *rhashtable_walk_next(struct rhashtable_iter *iter) { - struct bucket_table *tbl = iter->walker->tbl; + struct bucket_table *tbl = iter->walker.tbl; struct rhashtable *ht = iter->ht; struct rhash_head *p = iter->p; @@ -630,8 +620,8 @@ next: /* Ensure we see any new tables. */ smp_rmb(); - iter->walker->tbl = rht_dereference_rcu(tbl->future_tbl, ht); - if (iter->walker->tbl) { + iter->walker.tbl = rht_dereference_rcu(tbl->future_tbl, ht); + if (iter->walker.tbl) { iter->slot = 0; iter->skip = 0; return ERR_PTR(-EAGAIN); @@ -651,7 +641,7 @@ void rhashtable_walk_stop(struct rhashtable_iter *iter) __releases(RCU) { struct rhashtable *ht; - struct bucket_table *tbl = iter->walker->tbl; + struct bucket_table *tbl = iter->walker.tbl; if (!tbl) goto out; @@ -660,9 +650,9 @@ void rhashtable_walk_stop(struct rhashtable_iter *iter) spin_lock(&ht->lock); if (tbl->rehash < tbl->size) - list_add(&iter->walker->list, &tbl->walkers); + list_add(&iter->walker.list, &tbl->walkers); else - iter->walker->tbl = NULL; + iter->walker.tbl = NULL; spin_unlock(&ht->lock); iter->p = NULL; -- cgit v1.2.3-59-g8ed1b From 5ca8cc5bf11faed257c762018aea9106d529232f Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Wed, 24 Aug 2016 12:31:31 +0200 Subject: rhashtable: add rhashtable_lookup_get_insert_key() This patch modifies __rhashtable_insert_fast() so it returns the existing object that clashes with the one that you want to insert. In case the object is successfully inserted, NULL is returned. Otherwise, you get an error via ERR_PTR(). This patch adapts the existing callers of __rhashtable_insert_fast() so they handle this new logic, and it adds a new rhashtable_lookup_get_insert_key() interface to fetch this existing object. nf_tables needs this change to improve handling of EEXIST cases via honoring the NLM_F_EXCL flag and by checking if the data part of the mapping matches what we have. Cc: Herbert Xu Cc: Thomas Graf Signed-off-by: Pablo Neira Ayuso Acked-by: Herbert Xu --- include/linux/rhashtable.h | 70 +++++++++++++++++++++++++++++++++++++--------- lib/rhashtable.c | 10 +++++-- 2 files changed, 64 insertions(+), 16 deletions(-) (limited to 'lib') diff --git a/include/linux/rhashtable.h b/include/linux/rhashtable.h index 3eef0802a0cd..26b7a059c65e 100644 --- a/include/linux/rhashtable.h +++ b/include/linux/rhashtable.h @@ -343,7 +343,8 @@ int rhashtable_init(struct rhashtable *ht, struct bucket_table *rhashtable_insert_slow(struct rhashtable *ht, const void *key, struct rhash_head *obj, - struct bucket_table *old_tbl); + struct bucket_table *old_tbl, + void **data); int rhashtable_insert_rehash(struct rhashtable *ht, struct bucket_table *tbl); int rhashtable_walk_init(struct rhashtable *ht, struct rhashtable_iter *iter, @@ -563,8 +564,11 @@ restart: return NULL; } -/* Internal function, please use rhashtable_insert_fast() instead */ -static inline int __rhashtable_insert_fast( +/* Internal function, please use rhashtable_insert_fast() instead. This + * function returns the existing element already in hashes in there is a clash, + * otherwise it returns an error via ERR_PTR(). + */ +static inline void *__rhashtable_insert_fast( struct rhashtable *ht, const void *key, struct rhash_head *obj, const struct rhashtable_params params) { @@ -577,6 +581,7 @@ static inline int __rhashtable_insert_fast( spinlock_t *lock; unsigned int elasticity; unsigned int hash; + void *data = NULL; int err; restart: @@ -601,11 +606,14 @@ restart: new_tbl = rht_dereference_rcu(tbl->future_tbl, ht); if (unlikely(new_tbl)) { - tbl = rhashtable_insert_slow(ht, key, obj, new_tbl); + tbl = rhashtable_insert_slow(ht, key, obj, new_tbl, &data); if (!IS_ERR_OR_NULL(tbl)) goto slow_path; err = PTR_ERR(tbl); + if (err == -EEXIST) + err = 0; + goto out; } @@ -619,25 +627,25 @@ slow_path: err = rhashtable_insert_rehash(ht, tbl); rcu_read_unlock(); if (err) - return err; + return ERR_PTR(err); goto restart; } - err = -EEXIST; + err = 0; elasticity = ht->elasticity; rht_for_each(head, tbl, hash) { if (key && unlikely(!(params.obj_cmpfn ? params.obj_cmpfn(&arg, rht_obj(ht, head)) : - rhashtable_compare(&arg, rht_obj(ht, head))))) + rhashtable_compare(&arg, rht_obj(ht, head))))) { + data = rht_obj(ht, head); goto out; + } if (!--elasticity) goto slow_path; } - err = 0; - head = rht_dereference_bucket(tbl->buckets[hash], tbl, hash); RCU_INIT_POINTER(obj->next, head); @@ -652,7 +660,7 @@ out: spin_unlock_bh(lock); rcu_read_unlock(); - return err; + return err ? ERR_PTR(err) : data; } /** @@ -675,7 +683,13 @@ static inline int rhashtable_insert_fast( struct rhashtable *ht, struct rhash_head *obj, const struct rhashtable_params params) { - return __rhashtable_insert_fast(ht, NULL, obj, params); + void *ret; + + ret = __rhashtable_insert_fast(ht, NULL, obj, params); + if (IS_ERR(ret)) + return PTR_ERR(ret); + + return ret == NULL ? 0 : -EEXIST; } /** @@ -704,11 +718,15 @@ static inline int rhashtable_lookup_insert_fast( const struct rhashtable_params params) { const char *key = rht_obj(ht, obj); + void *ret; BUG_ON(ht->p.obj_hashfn); - return __rhashtable_insert_fast(ht, key + ht->p.key_offset, obj, - params); + ret = __rhashtable_insert_fast(ht, key + ht->p.key_offset, obj, params); + if (IS_ERR(ret)) + return PTR_ERR(ret); + + return ret == NULL ? 0 : -EEXIST; } /** @@ -736,6 +754,32 @@ static inline int rhashtable_lookup_insert_fast( static inline int rhashtable_lookup_insert_key( struct rhashtable *ht, const void *key, struct rhash_head *obj, const struct rhashtable_params params) +{ + void *ret; + + BUG_ON(!ht->p.obj_hashfn || !key); + + ret = __rhashtable_insert_fast(ht, key, obj, params); + if (IS_ERR(ret)) + return PTR_ERR(ret); + + return ret == NULL ? 0 : -EEXIST; +} + +/** + * rhashtable_lookup_get_insert_key - lookup and insert object into hash table + * @ht: hash table + * @obj: pointer to hash head inside object + * @params: hash table parameters + * @data: pointer to element data already in hashes + * + * Just like rhashtable_lookup_insert_key(), but this function returns the + * object if it exists, NULL if it does not and the insertion was successful, + * and an ERR_PTR otherwise. + */ +static inline void *rhashtable_lookup_get_insert_key( + struct rhashtable *ht, const void *key, struct rhash_head *obj, + const struct rhashtable_params params) { BUG_ON(!ht->p.obj_hashfn || !key); diff --git a/lib/rhashtable.c b/lib/rhashtable.c index 5d845ffd7982..7a940d92f17e 100644 --- a/lib/rhashtable.c +++ b/lib/rhashtable.c @@ -438,7 +438,8 @@ EXPORT_SYMBOL_GPL(rhashtable_insert_rehash); struct bucket_table *rhashtable_insert_slow(struct rhashtable *ht, const void *key, struct rhash_head *obj, - struct bucket_table *tbl) + struct bucket_table *tbl, + void **data) { struct rhash_head *head; unsigned int hash; @@ -449,8 +450,11 @@ struct bucket_table *rhashtable_insert_slow(struct rhashtable *ht, spin_lock_nested(rht_bucket_lock(tbl, hash), SINGLE_DEPTH_NESTING); err = -EEXIST; - if (key && rhashtable_lookup_fast(ht, key, ht->p)) - goto exit; + if (key) { + *data = rhashtable_lookup_fast(ht, key, ht->p); + if (*data) + goto exit; + } err = -E2BIG; if (unlikely(rht_grow_above_max(ht, tbl))) -- cgit v1.2.3-59-g8ed1b From 5c0ca3f566d7a19e9bf9671dfc2108fad1b7b9b2 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Mon, 12 Sep 2016 13:04:57 +0100 Subject: test_bpf: fix the dummy skb after dissector changes Commit d5709f7ab776 ("flow_dissector: For stripped vlan, get vlan info from skb->vlan_tci") made flow dissector look at vlan_proto when vlan is present. Since test_bpf sets skb->vlan_tci to ~0 (including VLAN_TAG_PRESENT) we have to populate skb->vlan_proto. Fixes false negative on test #24: test_bpf: #24 LD_PAYLOAD_OFF jited:0 175 ret 0 != 42 FAIL (1 times) Signed-off-by: Jakub Kicinski Reviewed-by: Dinan Gunawardena Acked-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- lib/test_bpf.c | 1 + 1 file changed, 1 insertion(+) (limited to 'lib') diff --git a/lib/test_bpf.c b/lib/test_bpf.c index 93f45011a59d..94346b4d8984 100644 --- a/lib/test_bpf.c +++ b/lib/test_bpf.c @@ -5485,6 +5485,7 @@ static struct sk_buff *populate_skb(char *buf, int size) skb->hash = SKB_HASH; skb->queue_mapping = SKB_QUEUE_MAP; skb->vlan_tci = SKB_VLAN_TCI; + skb->vlan_proto = htons(ETH_P_IP); skb->dev = &dev; skb->dev->ifindex = SKB_DEV_IFINDEX; skb->dev->type = SKB_DEV_TYPE; -- cgit v1.2.3-59-g8ed1b From ca26893f05e86497a86732768ec53cd38c0819ca Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Mon, 19 Sep 2016 19:00:09 +0800 Subject: rhashtable: Add rhlist interface The insecure_elasticity setting is an ugly wart brought out by users who need to insert duplicate objects (that is, distinct objects with identical keys) into the same table. In fact, those users have a much bigger problem. Once those duplicate objects are inserted, they don't have an interface to find them (unless you count the walker interface which walks over the entire table). Some users have resorted to doing a manual walk over the hash table which is of course broken because they don't handle the potential existence of multiple hash tables. The result is that they will break sporadically when they encounter a hash table resize/rehash. This patch provides a way out for those users, at the expense of an extra pointer per object. Essentially each object is now a list of objects carrying the same key. The hash table will only see the lists so nothing changes as far as rhashtable is concerned. To use this new interface, you need to insert a struct rhlist_head into your objects instead of struct rhash_head. While the hash table is unchanged, for type-safety you'll need to use struct rhltable instead of struct rhashtable. All the existing interfaces have been duplicated for rhlist, including the hash table walker. One missing feature is nulls marking because AFAIK the only potential user of it does not need duplicate objects. Should anyone need this it shouldn't be too hard to add. Signed-off-by: Herbert Xu Acked-by: Thomas Graf Signed-off-by: David S. Miller --- include/linux/rhashtable.h | 491 ++++++++++++++++++++++++++++++++++----------- lib/rhashtable.c | 258 +++++++++++++++++++----- 2 files changed, 583 insertions(+), 166 deletions(-) (limited to 'lib') diff --git a/include/linux/rhashtable.h b/include/linux/rhashtable.h index fd82584acd48..5c132d3188be 100644 --- a/include/linux/rhashtable.h +++ b/include/linux/rhashtable.h @@ -1,7 +1,7 @@ /* * Resizable, Scalable, Concurrent Hash Table * - * Copyright (c) 2015 Herbert Xu + * Copyright (c) 2015-2016 Herbert Xu * Copyright (c) 2014-2015 Thomas Graf * Copyright (c) 2008-2014 Patrick McHardy * @@ -53,6 +53,11 @@ struct rhash_head { struct rhash_head __rcu *next; }; +struct rhlist_head { + struct rhash_head rhead; + struct rhlist_head __rcu *next; +}; + /** * struct bucket_table - Table of hash buckets * @size: Number of hash buckets @@ -137,6 +142,7 @@ struct rhashtable_params { * @key_len: Key length for hashfn * @elasticity: Maximum chain length before rehash * @p: Configuration parameters + * @rhlist: True if this is an rhltable * @run_work: Deferred worker to expand/shrink asynchronously * @mutex: Mutex to protect current/future table swapping * @lock: Spin lock to protect walker list @@ -147,11 +153,20 @@ struct rhashtable { unsigned int key_len; unsigned int elasticity; struct rhashtable_params p; + bool rhlist; struct work_struct run_work; struct mutex mutex; spinlock_t lock; }; +/** + * struct rhltable - Hash table with duplicate objects in a list + * @ht: Underlying rhtable + */ +struct rhltable { + struct rhashtable ht; +}; + /** * struct rhashtable_walker - Hash table walker * @list: List entry on list of walkers @@ -163,9 +178,10 @@ struct rhashtable_walker { }; /** - * struct rhashtable_iter - Hash table iterator, fits into netlink cb + * struct rhashtable_iter - Hash table iterator * @ht: Table to iterate through * @p: Current pointer + * @list: Current hash list pointer * @walker: Associated rhashtable walker * @slot: Current slot * @skip: Number of entries to skip in slot @@ -173,6 +189,7 @@ struct rhashtable_walker { struct rhashtable_iter { struct rhashtable *ht; struct rhash_head *p; + struct rhlist_head *list; struct rhashtable_walker walker; unsigned int slot; unsigned int skip; @@ -339,13 +356,11 @@ static inline int lockdep_rht_bucket_is_held(const struct bucket_table *tbl, int rhashtable_init(struct rhashtable *ht, const struct rhashtable_params *params); +int rhltable_init(struct rhltable *hlt, + const struct rhashtable_params *params); -struct bucket_table *rhashtable_insert_slow(struct rhashtable *ht, - const void *key, - struct rhash_head *obj, - struct bucket_table *old_tbl, - void **data); -int rhashtable_insert_rehash(struct rhashtable *ht, struct bucket_table *tbl); +void *rhashtable_insert_slow(struct rhashtable *ht, const void *key, + struct rhash_head *obj); void rhashtable_walk_enter(struct rhashtable *ht, struct rhashtable_iter *iter); @@ -507,6 +522,31 @@ void rhashtable_destroy(struct rhashtable *ht); rht_for_each_entry_rcu_continue(tpos, pos, (tbl)->buckets[hash],\ tbl, hash, member) +/** + * rhl_for_each_rcu - iterate over rcu hash table list + * @pos: the &struct rlist_head to use as a loop cursor. + * @list: the head of the list + * + * This hash chain list-traversal primitive should be used on the + * list returned by rhltable_lookup. + */ +#define rhl_for_each_rcu(pos, list) \ + for (pos = list; pos; pos = rcu_dereference_raw(pos->next)) + +/** + * rhl_for_each_entry_rcu - iterate over rcu hash table list of given type + * @tpos: the type * to use as a loop cursor. + * @pos: the &struct rlist_head to use as a loop cursor. + * @list: the head of the list + * @member: name of the &struct rlist_head within the hashable struct. + * + * This hash chain list-traversal primitive should be used on the + * list returned by rhltable_lookup. + */ +#define rhl_for_each_entry_rcu(tpos, pos, list, member) \ + for (pos = list; pos && rht_entry(tpos, pos, member); \ + pos = rcu_dereference_raw(pos->next)) + static inline int rhashtable_compare(struct rhashtable_compare_arg *arg, const void *obj) { @@ -516,18 +556,8 @@ static inline int rhashtable_compare(struct rhashtable_compare_arg *arg, return memcmp(ptr + ht->p.key_offset, arg->key, ht->p.key_len); } -/** - * rhashtable_lookup_fast - search hash table, inlined version - * @ht: hash table - * @key: the pointer to the key - * @params: hash table parameters - * - * Computes the hash value for the key and traverses the bucket chain looking - * for a entry with an identical key. The first matching entry is returned. - * - * Returns the first entry on which the compare function returned true. - */ -static inline void *rhashtable_lookup_fast( +/* Internal function, do not use. */ +static inline struct rhash_head *__rhashtable_lookup( struct rhashtable *ht, const void *key, const struct rhashtable_params params) { @@ -539,8 +569,6 @@ static inline void *rhashtable_lookup_fast( struct rhash_head *he; unsigned int hash; - rcu_read_lock(); - tbl = rht_dereference_rcu(ht->tbl, ht); restart: hash = rht_key_hashfn(ht, tbl, key, params); @@ -549,8 +577,7 @@ restart: params.obj_cmpfn(&arg, rht_obj(ht, he)) : rhashtable_compare(&arg, rht_obj(ht, he))) continue; - rcu_read_unlock(); - return rht_obj(ht, he); + return he; } /* Ensure we see any new tables. */ @@ -559,96 +586,165 @@ restart: tbl = rht_dereference_rcu(tbl->future_tbl, ht); if (unlikely(tbl)) goto restart; - rcu_read_unlock(); return NULL; } +/** + * rhashtable_lookup - search hash table + * @ht: hash table + * @key: the pointer to the key + * @params: hash table parameters + * + * Computes the hash value for the key and traverses the bucket chain looking + * for a entry with an identical key. The first matching entry is returned. + * + * This must only be called under the RCU read lock. + * + * Returns the first entry on which the compare function returned true. + */ +static inline void *rhashtable_lookup( + struct rhashtable *ht, const void *key, + const struct rhashtable_params params) +{ + struct rhash_head *he = __rhashtable_lookup(ht, key, params); + + return he ? rht_obj(ht, he) : NULL; +} + +/** + * rhashtable_lookup_fast - search hash table, without RCU read lock + * @ht: hash table + * @key: the pointer to the key + * @params: hash table parameters + * + * Computes the hash value for the key and traverses the bucket chain looking + * for a entry with an identical key. The first matching entry is returned. + * + * Only use this function when you have other mechanisms guaranteeing + * that the object won't go away after the RCU read lock is released. + * + * Returns the first entry on which the compare function returned true. + */ +static inline void *rhashtable_lookup_fast( + struct rhashtable *ht, const void *key, + const struct rhashtable_params params) +{ + void *obj; + + rcu_read_lock(); + obj = rhashtable_lookup(ht, key, params); + rcu_read_unlock(); + + return obj; +} + +/** + * rhltable_lookup - search hash list table + * @hlt: hash table + * @key: the pointer to the key + * @params: hash table parameters + * + * Computes the hash value for the key and traverses the bucket chain looking + * for a entry with an identical key. All matching entries are returned + * in a list. + * + * This must only be called under the RCU read lock. + * + * Returns the list of entries that match the given key. + */ +static inline struct rhlist_head *rhltable_lookup( + struct rhltable *hlt, const void *key, + const struct rhashtable_params params) +{ + struct rhash_head *he = __rhashtable_lookup(&hlt->ht, key, params); + + return he ? container_of(he, struct rhlist_head, rhead) : NULL; +} + /* Internal function, please use rhashtable_insert_fast() instead. This * function returns the existing element already in hashes in there is a clash, * otherwise it returns an error via ERR_PTR(). */ static inline void *__rhashtable_insert_fast( struct rhashtable *ht, const void *key, struct rhash_head *obj, - const struct rhashtable_params params) + const struct rhashtable_params params, bool rhlist) { struct rhashtable_compare_arg arg = { .ht = ht, .key = key, }; - struct bucket_table *tbl, *new_tbl; + struct rhash_head __rcu **pprev; + struct bucket_table *tbl; struct rhash_head *head; spinlock_t *lock; - unsigned int elasticity; unsigned int hash; - void *data = NULL; - int err; + int elasticity; + void *data; -restart: rcu_read_lock(); tbl = rht_dereference_rcu(ht->tbl, ht); + hash = rht_head_hashfn(ht, tbl, obj, params); + lock = rht_bucket_lock(tbl, hash); + spin_lock_bh(lock); - /* All insertions must grab the oldest table containing - * the hashed bucket that is yet to be rehashed. - */ - for (;;) { - hash = rht_head_hashfn(ht, tbl, obj, params); - lock = rht_bucket_lock(tbl, hash); - spin_lock_bh(lock); - - if (tbl->rehash <= hash) - break; - + if (unlikely(rht_dereference_bucket(tbl->future_tbl, tbl, hash))) { +slow_path: spin_unlock_bh(lock); - tbl = rht_dereference_rcu(tbl->future_tbl, ht); + rcu_read_unlock(); + return rhashtable_insert_slow(ht, key, obj); } - new_tbl = rht_dereference_rcu(tbl->future_tbl, ht); - if (unlikely(new_tbl)) { - tbl = rhashtable_insert_slow(ht, key, obj, new_tbl, &data); - if (!IS_ERR_OR_NULL(tbl)) - goto slow_path; + elasticity = ht->elasticity; + pprev = &tbl->buckets[hash]; + rht_for_each(head, tbl, hash) { + struct rhlist_head *plist; + struct rhlist_head *list; + + elasticity--; + if (!key || + (params.obj_cmpfn ? + params.obj_cmpfn(&arg, rht_obj(ht, head)) : + rhashtable_compare(&arg, rht_obj(ht, head)))) + continue; + + data = rht_obj(ht, head); - err = PTR_ERR(tbl); - if (err == -EEXIST) - err = 0; + if (!rhlist) + goto out; - goto out; - } - err = -E2BIG; - if (unlikely(rht_grow_above_max(ht, tbl))) - goto out; + list = container_of(obj, struct rhlist_head, rhead); + plist = container_of(head, struct rhlist_head, rhead); - if (unlikely(rht_grow_above_100(ht, tbl))) { -slow_path: - spin_unlock_bh(lock); - err = rhashtable_insert_rehash(ht, tbl); - rcu_read_unlock(); - if (err) - return ERR_PTR(err); + RCU_INIT_POINTER(list->next, plist); + head = rht_dereference_bucket(head->next, tbl, hash); + RCU_INIT_POINTER(list->rhead.next, head); + rcu_assign_pointer(*pprev, obj); - goto restart; + goto good; } - err = 0; - elasticity = ht->elasticity; - rht_for_each(head, tbl, hash) { - if (key && - unlikely(!(params.obj_cmpfn ? - params.obj_cmpfn(&arg, rht_obj(ht, head)) : - rhashtable_compare(&arg, rht_obj(ht, head))))) { - data = rht_obj(ht, head); - goto out; - } - if (!--elasticity) - goto slow_path; - } + if (elasticity <= 0) + goto slow_path; + + data = ERR_PTR(-E2BIG); + if (unlikely(rht_grow_above_max(ht, tbl))) + goto out; + + if (unlikely(rht_grow_above_100(ht, tbl))) + goto slow_path; head = rht_dereference_bucket(tbl->buckets[hash], tbl, hash); RCU_INIT_POINTER(obj->next, head); + if (rhlist) { + struct rhlist_head *list; + + list = container_of(obj, struct rhlist_head, rhead); + RCU_INIT_POINTER(list->next, NULL); + } rcu_assign_pointer(tbl->buckets[hash], obj); @@ -656,11 +752,14 @@ slow_path: if (rht_grow_above_75(ht, tbl)) schedule_work(&ht->run_work); +good: + data = NULL; + out: spin_unlock_bh(lock); rcu_read_unlock(); - return err ? ERR_PTR(err) : data; + return data; } /** @@ -685,13 +784,65 @@ static inline int rhashtable_insert_fast( { void *ret; - ret = __rhashtable_insert_fast(ht, NULL, obj, params); + ret = __rhashtable_insert_fast(ht, NULL, obj, params, false); if (IS_ERR(ret)) return PTR_ERR(ret); return ret == NULL ? 0 : -EEXIST; } +/** + * rhltable_insert_key - insert object into hash list table + * @hlt: hash list table + * @key: the pointer to the key + * @list: pointer to hash list head inside object + * @params: hash table parameters + * + * Will take a per bucket spinlock to protect against mutual mutations + * on the same bucket. Multiple insertions may occur in parallel unless + * they map to the same bucket lock. + * + * It is safe to call this function from atomic context. + * + * Will trigger an automatic deferred table resizing if the size grows + * beyond the watermark indicated by grow_decision() which can be passed + * to rhashtable_init(). + */ +static inline int rhltable_insert_key( + struct rhltable *hlt, const void *key, struct rhlist_head *list, + const struct rhashtable_params params) +{ + return PTR_ERR(__rhashtable_insert_fast(&hlt->ht, key, &list->rhead, + params, true)); +} + +/** + * rhltable_insert - insert object into hash list table + * @hlt: hash list table + * @list: pointer to hash list head inside object + * @params: hash table parameters + * + * Will take a per bucket spinlock to protect against mutual mutations + * on the same bucket. Multiple insertions may occur in parallel unless + * they map to the same bucket lock. + * + * It is safe to call this function from atomic context. + * + * Will trigger an automatic deferred table resizing if the size grows + * beyond the watermark indicated by grow_decision() which can be passed + * to rhashtable_init(). + */ +static inline int rhltable_insert( + struct rhltable *hlt, struct rhlist_head *list, + const struct rhashtable_params params) +{ + const char *key = rht_obj(&hlt->ht, &list->rhead); + + key += params.key_offset; + + return rhltable_insert_key(hlt, key, list, params); +} + /** * rhashtable_lookup_insert_fast - lookup and insert object into hash table * @ht: hash table @@ -722,7 +873,8 @@ static inline int rhashtable_lookup_insert_fast( BUG_ON(ht->p.obj_hashfn); - ret = __rhashtable_insert_fast(ht, key + ht->p.key_offset, obj, params); + ret = __rhashtable_insert_fast(ht, key + ht->p.key_offset, obj, params, + false); if (IS_ERR(ret)) return PTR_ERR(ret); @@ -759,7 +911,7 @@ static inline int rhashtable_lookup_insert_key( BUG_ON(!ht->p.obj_hashfn || !key); - ret = __rhashtable_insert_fast(ht, key, obj, params); + ret = __rhashtable_insert_fast(ht, key, obj, params, false); if (IS_ERR(ret)) return PTR_ERR(ret); @@ -783,13 +935,14 @@ static inline void *rhashtable_lookup_get_insert_key( { BUG_ON(!ht->p.obj_hashfn || !key); - return __rhashtable_insert_fast(ht, key, obj, params); + return __rhashtable_insert_fast(ht, key, obj, params, false); } /* Internal function, please use rhashtable_remove_fast() instead */ -static inline int __rhashtable_remove_fast( +static inline int __rhashtable_remove_fast_one( struct rhashtable *ht, struct bucket_table *tbl, - struct rhash_head *obj, const struct rhashtable_params params) + struct rhash_head *obj, const struct rhashtable_params params, + bool rhlist) { struct rhash_head __rcu **pprev; struct rhash_head *he; @@ -804,39 +957,66 @@ static inline int __rhashtable_remove_fast( pprev = &tbl->buckets[hash]; rht_for_each(he, tbl, hash) { + struct rhlist_head *list; + + list = container_of(he, struct rhlist_head, rhead); + if (he != obj) { + struct rhlist_head __rcu **lpprev; + pprev = &he->next; - continue; + + if (!rhlist) + continue; + + do { + lpprev = &list->next; + list = rht_dereference_bucket(list->next, + tbl, hash); + } while (list && obj != &list->rhead); + + if (!list) + continue; + + list = rht_dereference_bucket(list->next, tbl, hash); + RCU_INIT_POINTER(*lpprev, list); + err = 0; + break; } - rcu_assign_pointer(*pprev, obj->next); - err = 0; + obj = rht_dereference_bucket(obj->next, tbl, hash); + err = 1; + + if (rhlist) { + list = rht_dereference_bucket(list->next, tbl, hash); + if (list) { + RCU_INIT_POINTER(list->rhead.next, obj); + obj = &list->rhead; + err = 0; + } + } + + rcu_assign_pointer(*pprev, obj); break; } spin_unlock_bh(lock); + if (err > 0) { + atomic_dec(&ht->nelems); + if (unlikely(ht->p.automatic_shrinking && + rht_shrink_below_30(ht, tbl))) + schedule_work(&ht->run_work); + err = 0; + } + return err; } -/** - * rhashtable_remove_fast - remove object from hash table - * @ht: hash table - * @obj: pointer to hash head inside object - * @params: hash table parameters - * - * Since the hash chain is single linked, the removal operation needs to - * walk the bucket chain upon removal. The removal operation is thus - * considerable slow if the hash table is not correctly sized. - * - * Will automatically shrink the table via rhashtable_expand() if the - * shrink_decision function specified at rhashtable_init() returns true. - * - * Returns zero on success, -ENOENT if the entry could not be found. - */ -static inline int rhashtable_remove_fast( +/* Internal function, please use rhashtable_remove_fast() instead */ +static inline int __rhashtable_remove_fast( struct rhashtable *ht, struct rhash_head *obj, - const struct rhashtable_params params) + const struct rhashtable_params params, bool rhlist) { struct bucket_table *tbl; int err; @@ -850,24 +1030,60 @@ static inline int rhashtable_remove_fast( * visible then that guarantees the entry to still be in * the old tbl if it exists. */ - while ((err = __rhashtable_remove_fast(ht, tbl, obj, params)) && + while ((err = __rhashtable_remove_fast_one(ht, tbl, obj, params, + rhlist)) && (tbl = rht_dereference_rcu(tbl->future_tbl, ht))) ; - if (err) - goto out; - - atomic_dec(&ht->nelems); - if (unlikely(ht->p.automatic_shrinking && - rht_shrink_below_30(ht, tbl))) - schedule_work(&ht->run_work); - -out: rcu_read_unlock(); return err; } +/** + * rhashtable_remove_fast - remove object from hash table + * @ht: hash table + * @obj: pointer to hash head inside object + * @params: hash table parameters + * + * Since the hash chain is single linked, the removal operation needs to + * walk the bucket chain upon removal. The removal operation is thus + * considerable slow if the hash table is not correctly sized. + * + * Will automatically shrink the table via rhashtable_expand() if the + * shrink_decision function specified at rhashtable_init() returns true. + * + * Returns zero on success, -ENOENT if the entry could not be found. + */ +static inline int rhashtable_remove_fast( + struct rhashtable *ht, struct rhash_head *obj, + const struct rhashtable_params params) +{ + return __rhashtable_remove_fast(ht, obj, params, false); +} + +/** + * rhltable_remove - remove object from hash list table + * @hlt: hash list table + * @list: pointer to hash list head inside object + * @params: hash table parameters + * + * Since the hash chain is single linked, the removal operation needs to + * walk the bucket chain upon removal. The removal operation is thus + * considerable slow if the hash table is not correctly sized. + * + * Will automatically shrink the table via rhashtable_expand() if the + * shrink_decision function specified at rhashtable_init() returns true. + * + * Returns zero on success, -ENOENT if the entry could not be found. + */ +static inline int rhltable_remove( + struct rhltable *hlt, struct rhlist_head *list, + const struct rhashtable_params params) +{ + return __rhashtable_remove_fast(&hlt->ht, &list->rhead, params, true); +} + /* Internal function, please use rhashtable_replace_fast() instead */ static inline int __rhashtable_replace_fast( struct rhashtable *ht, struct bucket_table *tbl, @@ -958,4 +1174,51 @@ static inline int rhashtable_walk_init(struct rhashtable *ht, return 0; } +/** + * rhltable_walk_enter - Initialise an iterator + * @hlt: Table to walk over + * @iter: Hash table Iterator + * + * This function prepares a hash table walk. + * + * Note that if you restart a walk after rhashtable_walk_stop you + * may see the same object twice. Also, you may miss objects if + * there are removals in between rhashtable_walk_stop and the next + * call to rhashtable_walk_start. + * + * For a completely stable walk you should construct your own data + * structure outside the hash table. + * + * This function may sleep so you must not call it from interrupt + * context or with spin locks held. + * + * You must call rhashtable_walk_exit after this function returns. + */ +static inline void rhltable_walk_enter(struct rhltable *hlt, + struct rhashtable_iter *iter) +{ + return rhashtable_walk_enter(&hlt->ht, iter); +} + +/** + * rhltable_free_and_destroy - free elements and destroy hash list table + * @hlt: the hash list table to destroy + * @free_fn: callback to release resources of element + * @arg: pointer passed to free_fn + * + * See documentation for rhashtable_free_and_destroy. + */ +static inline void rhltable_free_and_destroy(struct rhltable *hlt, + void (*free_fn)(void *ptr, + void *arg), + void *arg) +{ + return rhashtable_free_and_destroy(&hlt->ht, free_fn, arg); +} + +static inline void rhltable_destroy(struct rhltable *hlt) +{ + return rhltable_free_and_destroy(hlt, NULL, NULL); +} + #endif /* _LINUX_RHASHTABLE_H */ diff --git a/lib/rhashtable.c b/lib/rhashtable.c index 06c28728bb53..32d0ad058380 100644 --- a/lib/rhashtable.c +++ b/lib/rhashtable.c @@ -378,22 +378,8 @@ static void rht_deferred_worker(struct work_struct *work) schedule_work(&ht->run_work); } -static bool rhashtable_check_elasticity(struct rhashtable *ht, - struct bucket_table *tbl, - unsigned int hash) -{ - unsigned int elasticity = ht->elasticity; - struct rhash_head *head; - - rht_for_each(head, tbl, hash) - if (!--elasticity) - return true; - - return false; -} - -int rhashtable_insert_rehash(struct rhashtable *ht, - struct bucket_table *tbl) +static int rhashtable_insert_rehash(struct rhashtable *ht, + struct bucket_table *tbl) { struct bucket_table *old_tbl; struct bucket_table *new_tbl; @@ -439,57 +425,165 @@ fail: return err; } -EXPORT_SYMBOL_GPL(rhashtable_insert_rehash); -struct bucket_table *rhashtable_insert_slow(struct rhashtable *ht, - const void *key, - struct rhash_head *obj, - struct bucket_table *tbl, - void **data) +static void *rhashtable_lookup_one(struct rhashtable *ht, + struct bucket_table *tbl, unsigned int hash, + const void *key, struct rhash_head *obj) { + struct rhashtable_compare_arg arg = { + .ht = ht, + .key = key, + }; + struct rhash_head __rcu **pprev; struct rhash_head *head; - unsigned int hash; - int err; + int elasticity; - tbl = rhashtable_last_table(ht, tbl); - hash = head_hashfn(ht, tbl, obj); - spin_lock_nested(rht_bucket_lock(tbl, hash), SINGLE_DEPTH_NESTING); - - err = -EEXIST; - if (key) { - *data = rhashtable_lookup_fast(ht, key, ht->p); - if (*data) - goto exit; + elasticity = ht->elasticity; + pprev = &tbl->buckets[hash]; + rht_for_each(head, tbl, hash) { + struct rhlist_head *list; + struct rhlist_head *plist; + + elasticity--; + if (!key || + (ht->p.obj_cmpfn ? + ht->p.obj_cmpfn(&arg, rht_obj(ht, head)) : + rhashtable_compare(&arg, rht_obj(ht, head)))) + continue; + + if (!ht->rhlist) + return rht_obj(ht, head); + + list = container_of(obj, struct rhlist_head, rhead); + plist = container_of(head, struct rhlist_head, rhead); + + RCU_INIT_POINTER(list->next, plist); + head = rht_dereference_bucket(head->next, tbl, hash); + RCU_INIT_POINTER(list->rhead.next, head); + rcu_assign_pointer(*pprev, obj); + + return NULL; } - err = -E2BIG; - if (unlikely(rht_grow_above_max(ht, tbl))) - goto exit; + if (elasticity <= 0) + return ERR_PTR(-EAGAIN); + + return ERR_PTR(-ENOENT); +} + +static struct bucket_table *rhashtable_insert_one(struct rhashtable *ht, + struct bucket_table *tbl, + unsigned int hash, + struct rhash_head *obj, + void *data) +{ + struct bucket_table *new_tbl; + struct rhash_head *head; + + if (!IS_ERR_OR_NULL(data)) + return ERR_PTR(-EEXIST); - err = -EAGAIN; - if (rhashtable_check_elasticity(ht, tbl, hash) || - rht_grow_above_100(ht, tbl)) - goto exit; + if (PTR_ERR(data) != -EAGAIN && PTR_ERR(data) != -ENOENT) + return ERR_CAST(data); - err = 0; + new_tbl = rcu_dereference(tbl->future_tbl); + if (new_tbl) + return new_tbl; + + if (PTR_ERR(data) != -ENOENT) + return ERR_CAST(data); + + if (unlikely(rht_grow_above_max(ht, tbl))) + return ERR_PTR(-E2BIG); + + if (unlikely(rht_grow_above_100(ht, tbl))) + return ERR_PTR(-EAGAIN); head = rht_dereference_bucket(tbl->buckets[hash], tbl, hash); RCU_INIT_POINTER(obj->next, head); + if (ht->rhlist) { + struct rhlist_head *list; + + list = container_of(obj, struct rhlist_head, rhead); + RCU_INIT_POINTER(list->next, NULL); + } rcu_assign_pointer(tbl->buckets[hash], obj); atomic_inc(&ht->nelems); + if (rht_grow_above_75(ht, tbl)) + schedule_work(&ht->run_work); -exit: - spin_unlock(rht_bucket_lock(tbl, hash)); + return NULL; +} - if (err == 0) - return NULL; - else if (err == -EAGAIN) - return tbl; - else - return ERR_PTR(err); +static void *rhashtable_try_insert(struct rhashtable *ht, const void *key, + struct rhash_head *obj) +{ + struct bucket_table *new_tbl; + struct bucket_table *tbl; + unsigned int hash; + spinlock_t *lock; + void *data; + + tbl = rcu_dereference(ht->tbl); + + /* All insertions must grab the oldest table containing + * the hashed bucket that is yet to be rehashed. + */ + for (;;) { + hash = rht_head_hashfn(ht, tbl, obj, ht->p); + lock = rht_bucket_lock(tbl, hash); + spin_lock_bh(lock); + + if (tbl->rehash <= hash) + break; + + spin_unlock_bh(lock); + tbl = rcu_dereference(tbl->future_tbl); + } + + data = rhashtable_lookup_one(ht, tbl, hash, key, obj); + new_tbl = rhashtable_insert_one(ht, tbl, hash, obj, data); + if (PTR_ERR(new_tbl) != -EEXIST) + data = ERR_CAST(new_tbl); + + while (!IS_ERR_OR_NULL(new_tbl)) { + tbl = new_tbl; + hash = rht_head_hashfn(ht, tbl, obj, ht->p); + spin_lock_nested(rht_bucket_lock(tbl, hash), + SINGLE_DEPTH_NESTING); + + data = rhashtable_lookup_one(ht, tbl, hash, key, obj); + new_tbl = rhashtable_insert_one(ht, tbl, hash, obj, data); + if (PTR_ERR(new_tbl) != -EEXIST) + data = ERR_CAST(new_tbl); + + spin_unlock(rht_bucket_lock(tbl, hash)); + } + + spin_unlock_bh(lock); + + if (PTR_ERR(data) == -EAGAIN) + data = ERR_PTR(rhashtable_insert_rehash(ht, tbl) ?: + -EAGAIN); + + return data; +} + +void *rhashtable_insert_slow(struct rhashtable *ht, const void *key, + struct rhash_head *obj) +{ + void *data; + + do { + rcu_read_lock(); + data = rhashtable_try_insert(ht, key, obj); + rcu_read_unlock(); + } while (PTR_ERR(data) == -EAGAIN); + + return data; } EXPORT_SYMBOL_GPL(rhashtable_insert_slow); @@ -593,11 +687,16 @@ EXPORT_SYMBOL_GPL(rhashtable_walk_start); void *rhashtable_walk_next(struct rhashtable_iter *iter) { struct bucket_table *tbl = iter->walker.tbl; + struct rhlist_head *list = iter->list; struct rhashtable *ht = iter->ht; struct rhash_head *p = iter->p; + bool rhlist = ht->rhlist; if (p) { - p = rht_dereference_bucket_rcu(p->next, tbl, iter->slot); + if (!rhlist || !(list = rcu_dereference(list->next))) { + p = rcu_dereference(p->next); + list = container_of(p, struct rhlist_head, rhead); + } goto next; } @@ -605,6 +704,18 @@ void *rhashtable_walk_next(struct rhashtable_iter *iter) int skip = iter->skip; rht_for_each_rcu(p, tbl, iter->slot) { + if (rhlist) { + list = container_of(p, struct rhlist_head, + rhead); + do { + if (!skip) + goto next; + skip--; + list = rcu_dereference(list->next); + } while (list); + + continue; + } if (!skip) break; skip--; @@ -614,7 +725,8 @@ next: if (!rht_is_a_nulls(p)) { iter->skip++; iter->p = p; - return rht_obj(ht, p); + iter->list = list; + return rht_obj(ht, rhlist ? &list->rhead : p); } iter->skip = 0; @@ -802,6 +914,48 @@ int rhashtable_init(struct rhashtable *ht, } EXPORT_SYMBOL_GPL(rhashtable_init); +/** + * rhltable_init - initialize a new hash list table + * @hlt: hash list table to be initialized + * @params: configuration parameters + * + * Initializes a new hash list table. + * + * See documentation for rhashtable_init. + */ +int rhltable_init(struct rhltable *hlt, const struct rhashtable_params *params) +{ + int err; + + /* No rhlist NULLs marking for now. */ + if (params->nulls_base) + return -EINVAL; + + err = rhashtable_init(&hlt->ht, params); + hlt->ht.rhlist = true; + return err; +} +EXPORT_SYMBOL_GPL(rhltable_init); + +static void rhashtable_free_one(struct rhashtable *ht, struct rhash_head *obj, + void (*free_fn)(void *ptr, void *arg), + void *arg) +{ + struct rhlist_head *list; + + if (!ht->rhlist) { + free_fn(rht_obj(ht, obj), arg); + return; + } + + list = container_of(obj, struct rhlist_head, rhead); + do { + obj = &list->rhead; + list = rht_dereference(list->next, ht); + free_fn(rht_obj(ht, obj), arg); + } while (list); +} + /** * rhashtable_free_and_destroy - free elements and destroy hash table * @ht: the hash table to destroy @@ -839,7 +993,7 @@ void rhashtable_free_and_destroy(struct rhashtable *ht, pos = next, next = !rht_is_a_nulls(pos) ? rht_dereference(pos->next, ht) : NULL) - free_fn(rht_obj(ht, pos), arg); + rhashtable_free_one(ht, pos, free_fn, arg); } } -- cgit v1.2.3-59-g8ed1b From a4f1f9ac8153e22869b6408832b5a9bb9c762bf6 Mon Sep 17 00:00:00 2001 From: Neal Cardwell Date: Mon, 19 Sep 2016 23:39:09 -0400 Subject: lib/win_minmax: windowed min or max estimator This commit introduces a generic library to estimate either the min or max value of a time-varying variable over a recent time window. This is code originally from Kathleen Nichols. The current form of the code is from Van Jacobson. A single struct minmax_sample will track the estimated windowed-max value of the series if you call minmax_running_max() or the estimated windowed-min value of the series if you call minmax_running_min(). Nearly equivalent code is already in place for minimum RTT estimation in the TCP stack. This commit extracts that code and generalizes it to handle both min and max. Moving the code here reduces the footprint and complexity of the TCP code base and makes the filter generally available for other parts of the codebase, including an upcoming TCP congestion control module. This library works well for time series where the measurements are smoothly increasing or decreasing. Signed-off-by: Van Jacobson Signed-off-by: Neal Cardwell Signed-off-by: Yuchung Cheng Signed-off-by: Nandita Dukkipati Signed-off-by: Eric Dumazet Signed-off-by: Soheil Hassas Yeganeh Signed-off-by: David S. Miller --- include/linux/win_minmax.h | 37 +++++++++++++++++ lib/Makefile | 2 +- lib/win_minmax.c | 98 ++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 136 insertions(+), 1 deletion(-) create mode 100644 include/linux/win_minmax.h create mode 100644 lib/win_minmax.c (limited to 'lib') diff --git a/include/linux/win_minmax.h b/include/linux/win_minmax.h new file mode 100644 index 000000000000..56569604278f --- /dev/null +++ b/include/linux/win_minmax.h @@ -0,0 +1,37 @@ +/** + * lib/minmax.c: windowed min/max tracker by Kathleen Nichols. + * + */ +#ifndef MINMAX_H +#define MINMAX_H + +#include + +/* A single data point for our parameterized min-max tracker */ +struct minmax_sample { + u32 t; /* time measurement was taken */ + u32 v; /* value measured */ +}; + +/* State for the parameterized min-max tracker */ +struct minmax { + struct minmax_sample s[3]; +}; + +static inline u32 minmax_get(const struct minmax *m) +{ + return m->s[0].v; +} + +static inline u32 minmax_reset(struct minmax *m, u32 t, u32 meas) +{ + struct minmax_sample val = { .t = t, .v = meas }; + + m->s[2] = m->s[1] = m->s[0] = val; + return m->s[0].v; +} + +u32 minmax_running_max(struct minmax *m, u32 win, u32 t, u32 meas); +u32 minmax_running_min(struct minmax *m, u32 win, u32 t, u32 meas); + +#endif diff --git a/lib/Makefile b/lib/Makefile index 5dc77a8ec297..df747e5eeb7a 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -22,7 +22,7 @@ lib-y := ctype.o string.o vsprintf.o cmdline.o \ sha1.o chacha20.o md5.o irq_regs.o argv_split.o \ flex_proportions.o ratelimit.o show_mem.o \ is_single_threaded.o plist.o decompress.o kobject_uevent.o \ - earlycpio.o seq_buf.o nmi_backtrace.o nodemask.o + earlycpio.o seq_buf.o nmi_backtrace.o nodemask.o win_minmax.o lib-$(CONFIG_MMU) += ioremap.o lib-$(CONFIG_SMP) += cpumask.o diff --git a/lib/win_minmax.c b/lib/win_minmax.c new file mode 100644 index 000000000000..c8420d404926 --- /dev/null +++ b/lib/win_minmax.c @@ -0,0 +1,98 @@ +/** + * lib/minmax.c: windowed min/max tracker + * + * Kathleen Nichols' algorithm for tracking the minimum (or maximum) + * value of a data stream over some fixed time interval. (E.g., + * the minimum RTT over the past five minutes.) It uses constant + * space and constant time per update yet almost always delivers + * the same minimum as an implementation that has to keep all the + * data in the window. + * + * The algorithm keeps track of the best, 2nd best & 3rd best min + * values, maintaining an invariant that the measurement time of + * the n'th best >= n-1'th best. It also makes sure that the three + * values are widely separated in the time window since that bounds + * the worse case error when that data is monotonically increasing + * over the window. + * + * Upon getting a new min, we can forget everything earlier because + * it has no value - the new min is <= everything else in the window + * by definition and it's the most recent. So we restart fresh on + * every new min and overwrites 2nd & 3rd choices. The same property + * holds for 2nd & 3rd best. + */ +#include +#include + +/* As time advances, update the 1st, 2nd, and 3rd choices. */ +static u32 minmax_subwin_update(struct minmax *m, u32 win, + const struct minmax_sample *val) +{ + u32 dt = val->t - m->s[0].t; + + if (unlikely(dt > win)) { + /* + * Passed entire window without a new val so make 2nd + * choice the new val & 3rd choice the new 2nd choice. + * we may have to iterate this since our 2nd choice + * may also be outside the window (we checked on entry + * that the third choice was in the window). + */ + m->s[0] = m->s[1]; + m->s[1] = m->s[2]; + m->s[2] = *val; + if (unlikely(val->t - m->s[0].t > win)) { + m->s[0] = m->s[1]; + m->s[1] = m->s[2]; + m->s[2] = *val; + } + } else if (unlikely(m->s[1].t == m->s[0].t) && dt > win/4) { + /* + * We've passed a quarter of the window without a new val + * so take a 2nd choice from the 2nd quarter of the window. + */ + m->s[2] = m->s[1] = *val; + } else if (unlikely(m->s[2].t == m->s[1].t) && dt > win/2) { + /* + * We've passed half the window without finding a new val + * so take a 3rd choice from the last half of the window + */ + m->s[2] = *val; + } + return m->s[0].v; +} + +/* Check if new measurement updates the 1st, 2nd or 3rd choice max. */ +u32 minmax_running_max(struct minmax *m, u32 win, u32 t, u32 meas) +{ + struct minmax_sample val = { .t = t, .v = meas }; + + if (unlikely(val.v >= m->s[0].v) || /* found new max? */ + unlikely(val.t - m->s[2].t > win)) /* nothing left in window? */ + return minmax_reset(m, t, meas); /* forget earlier samples */ + + if (unlikely(val.v >= m->s[1].v)) + m->s[2] = m->s[1] = val; + else if (unlikely(val.v >= m->s[2].v)) + m->s[2] = val; + + return minmax_subwin_update(m, win, &val); +} +EXPORT_SYMBOL(minmax_running_max); + +/* Check if new measurement updates the 1st, 2nd or 3rd choice min. */ +u32 minmax_running_min(struct minmax *m, u32 win, u32 t, u32 meas) +{ + struct minmax_sample val = { .t = t, .v = meas }; + + if (unlikely(val.v <= m->s[0].v) || /* found new min? */ + unlikely(val.t - m->s[2].t > win)) /* nothing left in window? */ + return minmax_reset(m, t, meas); /* forget earlier samples */ + + if (unlikely(val.v <= m->s[1].v)) + m->s[2] = m->s[1] = val; + else if (unlikely(val.v <= m->s[2].v)) + m->s[2] = val; + + return minmax_subwin_update(m, win, &val); +} -- cgit v1.2.3-59-g8ed1b From 3796c3cbfb91ad1e3269aa0d15060f888063924b Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Tue, 27 Sep 2016 08:42:42 -0700 Subject: lib: clean up put_cpu_var usage put_cpu_var takes the percpu data, not the data returned from get_cpu_var. This doesn't change the behavior. Cc: Tejun Heo Signed-off-by: Shaohua Li Acked-by: Tejun Heo Signed-off-by: David S. Miller --- lib/random32.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/random32.c b/lib/random32.c index 69ed593aab07..915982b304bb 100644 --- a/lib/random32.c +++ b/lib/random32.c @@ -81,7 +81,7 @@ u32 prandom_u32(void) u32 res; res = prandom_u32_state(state); - put_cpu_var(state); + put_cpu_var(net_rand_state); return res; } @@ -128,7 +128,7 @@ void prandom_bytes(void *buf, size_t bytes) struct rnd_state *state = &get_cpu_var(net_rand_state); prandom_bytes_state(state, buf, bytes); - put_cpu_var(state); + put_cpu_var(net_rand_state); } EXPORT_SYMBOL(prandom_bytes); -- cgit v1.2.3-59-g8ed1b