freebsd-dev/sys/net/if_llatbl.h
Alexander V. Chernikov df629abf3e Rework LLE code locking:
* struct llentry is now basically split into 2 pieces:
  all fields within 64 bytes (amd64) are now protected by both
  ifdata lock AND lle lock, e.g. you require both locks to be held
  exclusively for modification. All data necessary for fast path
  operations is kept here. Some fields were added:
  - r_l3addr - makes lookup key liev within first 64 bytes.
  - r_flags - flags, containing pre-compiled decision whether given
    lle contains usable data or not. Current the only flag is RLLE_VALID.
  - r_len - prepend data len, currently unused
  - r_kick - used to provide feedback to control plane (see below).
  All other fields are protected by lle lock.
* Add simple state machine for ARP to handle "about to expire" case:
  Current model (for the fast path) is the following:
  - rlock afdata
  - find / rlock rte
  - runlock afdata
  - see if "expire time" is approaching
    (time_uptime + la->la_preempt > la->la_expire)
  - if true, call arprequest() and decrease la_preempt
  - store MAC and runlock rte
  New model (data plane):
  - rlock afdata
  - find rte
  - check if it can be used using r_* fields only
  - if true, store MAC
  - if r_kick field != 0 set it to 0.
  - runlock afdata
  New mode (control plane):
  - schedule arptimer to be called in (V_arpt_keep - V_arp_maxtries)
    seconds instead of V_arpt_keep.
  - on first timer invocation change state from ARP_LLINFO_REACHABLE
    to ARP_LLINFO_VERIFY, sets r_kick to 1 and shedules next call in
    V_arpt_rexmit (default to 1 sec).
  - on subsequent timer invocations in ARP_LLINFO_VERIFY state, checks
    for r_kick value: reschedule if not changed, and send arprequest()
    if set to zero (e.g. entry was used).
* Convert IPv4 path to use new single-lock approach. IPv6 bits to follow.
* Slow down in_arpinput(): now valid reply will (in most cases) require
  acquiring afdata WLOCK twice. This is requirement for storing changed
  lle data. This change will be slightly optimized in future.
* Provide explicit hash link/unlink functions for both ipv4/ipv6 code.
  This will probably be moved to generic lle code once we have per-AF
  hashing callback inside lltable.
* Perform lle unlink on deletion immediately instead of delaying it to
  the timer routine.
* Make r244183 more explicit: use new LLE_CALLOUTREF flag to indicate the
  presence of lle reference used for safe callout calls.
2014-11-16 20:12:49 +00:00

248 lines
7.5 KiB
C

/*
* Copyright (c) 2004 Luigi Rizzo, Alessandro Cerri. All rights reserved.
* Copyright (c) 2004-2008 Qing Li. All rights reserved.
* Copyright (c) 2008 Kip Macy. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#ifndef _NET_IF_LLATBL_H_
#define _NET_IF_LLATBL_H_
#include <sys/_rwlock.h>
#include <netinet/in.h>
struct ifnet;
struct sysctl_req;
struct rt_msghdr;
struct rt_addrinfo;
struct llentry;
LIST_HEAD(llentries, llentry);
extern struct rwlock lltable_rwlock;
#define LLTABLE_RLOCK() rw_rlock(&lltable_rwlock)
#define LLTABLE_RUNLOCK() rw_runlock(&lltable_rwlock)
#define LLTABLE_WLOCK() rw_wlock(&lltable_rwlock)
#define LLTABLE_WUNLOCK() rw_wunlock(&lltable_rwlock)
#define LLTABLE_LOCK_ASSERT() rw_assert(&lltable_rwlock, RA_LOCKED)
struct llentry {
/* FIELDS PROTECTED BY IFDATA LOCK */
LIST_ENTRY(llentry) lle_next;
union {
struct in_addr addr4;
struct in6_addr addr6;
} r_l3addr;
union {
uint64_t mac_aligned;
uint16_t mac16[3];
uint8_t mac8[20]; /* IB needs 20 bytes. */
char ll_prepend[20]; /* L2 data to prepend */
} ll_addr;
uint16_t r_flags; /* runtime flags */
uint16_t r_len; /* length of prepend data */
uint64_t r_kick; /* for unused lle detection */
/* FIELDS PROTECTED BY LLE rwlock */
struct lltable *lle_tbl;
struct llentries *lle_head;
void (*lle_free)(struct lltable *, struct llentry *);
struct mbuf *la_hold;
int la_numheld; /* # of packets currently held */
time_t la_expire;
uint16_t la_flags;
uint16_t la_asked;
uint16_t la_preempt;
uint16_t ln_byhint;
int16_t ln_state; /* IPv6 has ND6_LLINFO_NOSTATE == -2 */
uint16_t ln_router;
time_t ln_ntick;
int lle_refcnt;
struct rwlock lle_lock;
/* XXX af-private? */
union {
struct callout ln_timer_ch;
struct callout la_timer;
} lle_timer;
/* NB: struct sockaddr must immediately follow */
};
#define LLE_WLOCK(lle) rw_wlock(&(lle)->lle_lock)
#define LLE_RLOCK(lle) rw_rlock(&(lle)->lle_lock)
#define LLE_WUNLOCK(lle) rw_wunlock(&(lle)->lle_lock)
#define LLE_RUNLOCK(lle) rw_runlock(&(lle)->lle_lock)
#define LLE_DOWNGRADE(lle) rw_downgrade(&(lle)->lle_lock)
#define LLE_TRY_UPGRADE(lle) rw_try_upgrade(&(lle)->lle_lock)
#define LLE_LOCK_INIT(lle) rw_init_flags(&(lle)->lle_lock, "lle", RW_DUPOK)
#define LLE_LOCK_DESTROY(lle) rw_destroy(&(lle)->lle_lock)
#define LLE_WLOCK_ASSERT(lle) rw_assert(&(lle)->lle_lock, RA_WLOCKED)
#define LLE_ADDREF(lle) do { \
LLE_WLOCK_ASSERT(lle); \
KASSERT((lle)->lle_refcnt >= 0, \
("negative refcnt %d on lle %p", \
(lle)->lle_refcnt, (lle))); \
(lle)->lle_refcnt++; \
} while (0)
#define LLE_REMREF(lle) do { \
LLE_WLOCK_ASSERT(lle); \
KASSERT((lle)->lle_refcnt > 0, \
("bogus refcnt %d on lle %p", \
(lle)->lle_refcnt, (lle))); \
(lle)->lle_refcnt--; \
} while (0)
#define LLE_FREE_LOCKED(lle) do { \
if ((lle)->lle_refcnt == 1) \
(lle)->lle_free((lle)->lle_tbl, (lle)); \
else { \
LLE_REMREF(lle); \
LLE_WUNLOCK(lle); \
} \
/* guard against invalid refs */ \
(lle) = NULL; \
} while (0)
#define LLE_FREE(lle) do { \
LLE_WLOCK(lle); \
LLE_FREE_LOCKED(lle); \
} while (0)
#define ln_timer_ch lle_timer.ln_timer_ch
#define la_timer lle_timer.la_timer
/* XXX bad name */
#define L3_ADDR(lle) ((struct sockaddr *)(&lle[1]))
#define L3_ADDR_LEN(lle) (((struct sockaddr *)(&lle[1]))->sa_len)
#ifndef LLTBL_HASHTBL_SIZE
#define LLTBL_HASHTBL_SIZE 32 /* default 32 ? */
#endif
#ifndef LLTBL_HASHMASK
#define LLTBL_HASHMASK (LLTBL_HASHTBL_SIZE - 1)
#endif
typedef struct llentry *(llt_lookup_t)(struct lltable *, u_int flags,
const struct sockaddr *l3addr);
typedef struct llentry *(llt_create_t)(struct lltable *, u_int flags,
const struct sockaddr *l3addr);
typedef int (llt_delete_t)(struct lltable *, u_int flags,
const struct sockaddr *l3addr);
typedef void (llt_prefix_free_t)(struct lltable *,
const struct sockaddr *prefix, const struct sockaddr *mask, u_int flags);
typedef int (llt_dump_t)(struct lltable *, struct sysctl_req *);
struct lltable {
SLIST_ENTRY(lltable) llt_link;
struct llentries lle_head[LLTBL_HASHTBL_SIZE];
int llt_af;
struct ifnet *llt_ifp;
llt_lookup_t *llt_lookup;
llt_create_t *llt_create;
llt_delete_t *llt_delete;
llt_prefix_free_t *llt_prefix_free;
llt_dump_t *llt_dump;
};
MALLOC_DECLARE(M_LLTABLE);
/*
* LLE flags used by fast path code
*/
#define RLLE_VALID 0x0001 /* ll_addr can be used */
/*
* Various LLE flags
*/
#define LLE_DELETED 0x0001 /* entry must be deleted */
#define LLE_STATIC 0x0002 /* entry is static */
#define LLE_IFADDR 0x0004 /* entry is interface addr */
#define LLE_VALID 0x0008 /* ll_addr is valid */
#define LLE_PUB 0x0020 /* publish entry ??? */
#define LLE_LINKED 0x0040 /* linked to lookup structure */
#define LLE_CALLOUTREF 0x0080 /* callout set */
/* LLE request flags */
#define LLE_UNLOCKED 0x0100 /* return lle unlocked */
#define LLE_EXCLUSIVE 0x0200 /* return lle wlocked */
#define LLATBL_HASH(key, mask) \
(((((((key >> 8) ^ key) >> 8) ^ key) >> 8) ^ key) & mask)
struct lltable *lltable_init(struct ifnet *, int);
void lltable_free(struct lltable *);
void lltable_prefix_free(int, struct sockaddr *,
struct sockaddr *, u_int);
#if 0
void lltable_drain(int);
#endif
int lltable_sysctl_dumparp(int, struct sysctl_req *);
size_t llentry_free(struct llentry *);
struct llentry *llentry_alloc(struct ifnet *, struct lltable *,
struct sockaddr_storage *);
/*
* Generic link layer address lookup function.
*/
static __inline struct llentry *
lla_lookup(struct lltable *llt, u_int flags, const struct sockaddr *l3addr)
{
return llt->llt_lookup(llt, flags, l3addr);
}
static __inline struct llentry *
lla_create(struct lltable *llt, u_int flags, const struct sockaddr *l3addr)
{
return llt->llt_create(llt, flags, l3addr);
}
static __inline int
lla_delete(struct lltable *llt, u_int flags, const struct sockaddr *l3addr)
{
return llt->llt_delete(llt, flags, l3addr);
}
int lla_rt_output(struct rt_msghdr *, struct rt_addrinfo *);
#include <sys/eventhandler.h>
enum {
LLENTRY_RESOLVED,
LLENTRY_TIMEDOUT,
LLENTRY_DELETED,
LLENTRY_EXPIRED,
};
typedef void (*lle_event_fn)(void *, struct llentry *, int);
EVENTHANDLER_DECLARE(lle_event, lle_event_fn);
#endif /* _NET_IF_LLATBL_H_ */