Differential D9770 Diff 29395 head/sys/kern/uipc_socket.c

Changeset View

Standalone View

head/sys/kern/uipc_socket.c

Show First 20 Lines • Show All 100 Lines • ▼ Show 20 Lines
*/		*/

#include <sys/cdefs.h>		#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");		__FBSDID("$FreeBSD$");

#include "opt_inet.h"		#include "opt_inet.h"
#include "opt_inet6.h"		#include "opt_inet6.h"
#include "opt_compat.h"		#include "opt_compat.h"
		#include "opt_sctp.h"

#include <sys/param.h>		#include <sys/param.h>
#include <sys/systm.h>		#include <sys/systm.h>
#include <sys/fcntl.h>		#include <sys/fcntl.h>
#include <sys/limits.h>		#include <sys/limits.h>
#include <sys/lock.h>		#include <sys/lock.h>
#include <sys/mac.h>		#include <sys/mac.h>
#include <sys/malloc.h>		#include <sys/malloc.h>
Show All 32 Lines
#ifdef COMPAT_FREEBSD32		#ifdef COMPAT_FREEBSD32
#include <sys/mount.h>		#include <sys/mount.h>
#include <sys/sysent.h>		#include <sys/sysent.h>
#include <compat/freebsd32/freebsd32.h>		#include <compat/freebsd32/freebsd32.h>
#endif		#endif

static int soreceive_rcvoob(struct socket so, struct uio uio,		static int soreceive_rcvoob(struct socket so, struct uio uio,
int flags);		int flags);
		static void so_rdknl_lock(void *);
		static void so_rdknl_unlock(void *);
		static void so_rdknl_assert_locked(void *);
		static void so_rdknl_assert_unlocked(void *);
		static void so_wrknl_lock(void *);
		static void so_wrknl_unlock(void *);
		static void so_wrknl_assert_locked(void *);
		static void so_wrknl_assert_unlocked(void *);

		bzUnsubmitted Not Done Inline Actions Can you get the "knl" which I suppose stands for "kernel" out of that name? I've never seen us do that before anywhere in native FreeBSD code. so_rlock() so_unlock() so_... so_wlock() so_.. You get the idea? Is there any reason we don't follow that? Given they are static I keep wondering why they need to be functions (despite the compiler going to inline them). Normally we do the "function wrappers" in order to export a stable KPI/KBI to modules and 3rd party code, which doesn't seem to be the need for with these? bz: Can you get the "knl" which I suppose stands for "kernel" out of that name? I've never seen us…
static void filt_sordetach(struct knote *kn);		static void filt_sordetach(struct knote *kn);
static int filt_soread(struct knote *kn, long hint);		static int filt_soread(struct knote *kn, long hint);
static void filt_sowdetach(struct knote *kn);		static void filt_sowdetach(struct knote *kn);
static int filt_sowrite(struct knote *kn, long hint);		static int filt_sowrite(struct knote *kn, long hint);
static int inline hhook_run_socket(struct socket so, void hctx, int32_t h_id);
static int filt_soempty(struct knote *kn, long hint);		static int filt_soempty(struct knote *kn, long hint);
		static int inline hhook_run_socket(struct socket so, void hctx, int32_t h_id);
fo_kqfilter_t soo_kqfilter;		fo_kqfilter_t soo_kqfilter;

static struct filterops soread_filtops = {		static struct filterops soread_filtops = {
.f_isfd = 1,		.f_isfd = 1,
.f_detach = filt_sordetach,		.f_detach = filt_sordetach,
.f_event = filt_soread,		.f_event = filt_soread,
};		};
static struct filterops sowrite_filtops = {		static struct filterops sowrite_filtops = {
▲ Show 20 Lines • Show All 216 Lines • ▼ Show 20 Lines	if (mac_socket_init(so, M_NOWAIT) != 0) {
return (NULL);		return (NULL);
}		}
#endif		#endif
if (khelp_init_osd(HELPER_CLASS_SOCKET, &so->osd)) {		if (khelp_init_osd(HELPER_CLASS_SOCKET, &so->osd)) {
uma_zfree(socket_zone, so);		uma_zfree(socket_zone, so);
return (NULL);		return (NULL);
}		}

		/*
		* The socket locking protocol allows to lock 2 sockets at a time,
		* however, the first one must be a listening socket. WITNESS lacks
		* a feature to change class of an existing lock, so we use DUPOK.
		*/
		mtx_init(&so->so_lock, "socket", NULL, MTX_DEF \| MTX_DUPOK);
SOCKBUF_LOCK_INIT(&so->so_snd, "so_snd");		SOCKBUF_LOCK_INIT(&so->so_snd, "so_snd");
SOCKBUF_LOCK_INIT(&so->so_rcv, "so_rcv");		SOCKBUF_LOCK_INIT(&so->so_rcv, "so_rcv");
		so->so_rcv.sb_sel = &so->so_rdsel;
		so->so_snd.sb_sel = &so->so_wrsel;
sx_init(&so->so_snd.sb_sx, "so_snd_sx");		sx_init(&so->so_snd.sb_sx, "so_snd_sx");
sx_init(&so->so_rcv.sb_sx, "so_rcv_sx");		sx_init(&so->so_rcv.sb_sx, "so_rcv_sx");
TAILQ_INIT(&so->so_snd.sb_aiojobq);		TAILQ_INIT(&so->so_snd.sb_aiojobq);
TAILQ_INIT(&so->so_rcv.sb_aiojobq);		TAILQ_INIT(&so->so_rcv.sb_aiojobq);
TASK_INIT(&so->so_snd.sb_aiotask, 0, soaio_snd, so);		TASK_INIT(&so->so_snd.sb_aiotask, 0, soaio_snd, so);
TASK_INIT(&so->so_rcv.sb_aiotask, 0, soaio_rcv, so);		TASK_INIT(&so->so_rcv.sb_aiotask, 0, soaio_rcv, so);
#ifdef VIMAGE		#ifdef VIMAGE
VNET_ASSERT(vnet != NULL, ("%s:%d vnet is NULL, so=%p",		VNET_ASSERT(vnet != NULL, ("%s:%d vnet is NULL, so=%p",
Show All 39 Lines
#endif		#endif
mtx_unlock(&so_global_mtx);		mtx_unlock(&so_global_mtx);
if (so->so_rcv.sb_hiwat)		if (so->so_rcv.sb_hiwat)
(void)chgsbsize(so->so_cred->cr_uidinfo,		(void)chgsbsize(so->so_cred->cr_uidinfo,
&so->so_rcv.sb_hiwat, 0, RLIM_INFINITY);		&so->so_rcv.sb_hiwat, 0, RLIM_INFINITY);
if (so->so_snd.sb_hiwat)		if (so->so_snd.sb_hiwat)
(void)chgsbsize(so->so_cred->cr_uidinfo,		(void)chgsbsize(so->so_cred->cr_uidinfo,
&so->so_snd.sb_hiwat, 0, RLIM_INFINITY);		&so->so_snd.sb_hiwat, 0, RLIM_INFINITY);
/* remove accept filter if one is present. */
if (so->so_accf != NULL)
accept_filt_setopt(so, NULL);
#ifdef MAC		#ifdef MAC
mac_socket_destroy(so);		mac_socket_destroy(so);
#endif		#endif
hhook_run_socket(so, NULL, HHOOK_SOCKET_CLOSE);		hhook_run_socket(so, NULL, HHOOK_SOCKET_CLOSE);

crfree(so->so_cred);		crfree(so->so_cred);
khelp_destroy_osd(&so->osd);		khelp_destroy_osd(&so->osd);
		if (SOLISTENING(so)) {
		if (so->sol_accept_filter != NULL)
		accept_filt_setopt(so, NULL);
		} else {
sx_destroy(&so->so_snd.sb_sx);		sx_destroy(&so->so_snd.sb_sx);
sx_destroy(&so->so_rcv.sb_sx);		sx_destroy(&so->so_rcv.sb_sx);
SOCKBUF_LOCK_DESTROY(&so->so_snd);		SOCKBUF_LOCK_DESTROY(&so->so_snd);
SOCKBUF_LOCK_DESTROY(&so->so_rcv);		SOCKBUF_LOCK_DESTROY(&so->so_rcv);
		}
		mtx_destroy(&so->so_lock);
uma_zfree(socket_zone, so);		uma_zfree(socket_zone, so);
}		}

/*		/*
* socreate returns a socket with a ref count of 1. The socket should be		* socreate returns a socket with a ref count of 1. The socket should be
* closed with soclose().		* closed with soclose().
*/		*/
int		int
Show All 26 Lines	if (prison_check_af(cred, prp->pr_domain->dom_family) != 0)
return (EPROTONOSUPPORT);		return (EPROTONOSUPPORT);

if (prp->pr_type != type)		if (prp->pr_type != type)
return (EPROTOTYPE);		return (EPROTOTYPE);
so = soalloc(CRED_TO_VNET(cred));		so = soalloc(CRED_TO_VNET(cred));
if (so == NULL)		if (so == NULL)
return (ENOBUFS);		return (ENOBUFS);

TAILQ_INIT(&so->so_incomp);
TAILQ_INIT(&so->so_comp);
so->so_type = type;		so->so_type = type;
so->so_cred = crhold(cred);		so->so_cred = crhold(cred);
if ((prp->pr_domain->dom_family == PF_INET) \|\|		if ((prp->pr_domain->dom_family == PF_INET) \|\|
(prp->pr_domain->dom_family == PF_INET6) \|\|		(prp->pr_domain->dom_family == PF_INET6) \|\|
(prp->pr_domain->dom_family == PF_ROUTE))		(prp->pr_domain->dom_family == PF_ROUTE))
so->so_fibnum = td->td_proc->p_fibnum;		so->so_fibnum = td->td_proc->p_fibnum;
else		else
so->so_fibnum = 0;		so->so_fibnum = 0;
so->so_proto = prp;		so->so_proto = prp;
#ifdef MAC		#ifdef MAC
mac_socket_create(cred, so);		mac_socket_create(cred, so);
#endif		#endif
knlist_init_mtx(&so->so_rcv.sb_sel.si_note, SOCKBUF_MTX(&so->so_rcv));		knlist_init(&so->so_rdsel.si_note, so, so_rdknl_lock, so_rdknl_unlock,
knlist_init_mtx(&so->so_snd.sb_sel.si_note, SOCKBUF_MTX(&so->so_snd));		so_rdknl_assert_locked, so_rdknl_assert_unlocked);
so->so_count = 1;		knlist_init(&so->so_wrsel.si_note, so, so_wrknl_lock, so_wrknl_unlock,
		so_wrknl_assert_locked, so_wrknl_assert_unlocked);
/*		/*
* Auto-sizing of socket buffers is managed by the protocols and		* Auto-sizing of socket buffers is managed by the protocols and
* the appropriate flags must be set in the pru_attach function.		* the appropriate flags must be set in the pru_attach function.
*/		*/
CURVNET_SET(so->so_vnet);		CURVNET_SET(so->so_vnet);
error = (*prp->pr_usrreqs->pru_attach)(so, proto, td);		error = (*prp->pr_usrreqs->pru_attach)(so, proto, td);
CURVNET_RESTORE();		CURVNET_RESTORE();
if (error) {		if (error) {
KASSERT(so->so_count == 1, ("socreate: so_count %d",
so->so_count));
so->so_count = 0;
sodealloc(so);		sodealloc(so);
return (error);		return (error);
}		}
		soref(so);
*aso = so;		*aso = so;
return (0);		return (0);
}		}

#ifdef REGRESSION		#ifdef REGRESSION
static int regression_sonewconn_earlytest = 1;		static int regression_sonewconn_earlytest = 1;
SYSCTL_INT(_regression, OID_AUTO, sonewconn_earlytest, CTLFLAG_RW,		SYSCTL_INT(_regression, OID_AUTO, sonewconn_earlytest, CTLFLAG_RW,
&regression_sonewconn_earlytest, 0, "Perform early sonewconn limit test");		&regression_sonewconn_earlytest, 0, "Perform early sonewconn limit test");
Show All 11 Lines
struct socket *		struct socket *
sonewconn(struct socket *head, int connstatus)		sonewconn(struct socket *head, int connstatus)
{		{
static struct timeval lastover;		static struct timeval lastover;
static struct timeval overinterval = { 60, 0 };		static struct timeval overinterval = { 60, 0 };
static int overcount;		static int overcount;

struct socket *so;		struct socket *so;
int over;		u_int over;

ACCEPT_LOCK();		SOLISTEN_LOCK(head);
over = (head->so_qlen > 3 * head->so_qlimit / 2);		over = (head->sol_qlen > 3 * head->sol_qlimit / 2);
ACCEPT_UNLOCK();		SOLISTEN_UNLOCK(head);
#ifdef REGRESSION		#ifdef REGRESSION
if (regression_sonewconn_earlytest && over) {		if (regression_sonewconn_earlytest && over) {
#else		#else
if (over) {		if (over) {
#endif		#endif
overcount++;		overcount++;

if (ratecheck(&lastover, &overinterval)) {		if (ratecheck(&lastover, &overinterval)) {
log(LOG_DEBUG, "%s: pcb %p: Listen queue overflow: "		log(LOG_DEBUG, "%s: pcb %p: Listen queue overflow: "
"%i already in queue awaiting acceptance "		"%i already in queue awaiting acceptance "
"(%d occurrences)\n",		"(%d occurrences)\n",
__func__, head->so_pcb, head->so_qlen, overcount);		__func__, head->so_pcb, head->sol_qlen, overcount);

overcount = 0;		overcount = 0;
}		}

return (NULL);		return (NULL);
}		}
VNET_ASSERT(head->so_vnet != NULL, ("%s:%d so_vnet is NULL, head=%p",		VNET_ASSERT(head->so_vnet != NULL, ("%s: so %p vnet is NULL",
__func__, __LINE__, head));		__func__, head));
so = soalloc(head->so_vnet);		so = soalloc(head->so_vnet);
if (so == NULL) {		if (so == NULL) {
log(LOG_DEBUG, "%s: pcb %p: New socket allocation failure: "		log(LOG_DEBUG, "%s: pcb %p: New socket allocation failure: "
"limit reached or out of memory\n",		"limit reached or out of memory\n",
__func__, head->so_pcb);		__func__, head->so_pcb);
return (NULL);		return (NULL);
}		}
if ((head->so_options & SO_ACCEPTFILTER) != 0)		so->so_listen = head;
connstatus = 0;
so->so_head = head;
so->so_type = head->so_type;		so->so_type = head->so_type;
so->so_options = head->so_options &~ SO_ACCEPTCONN;
so->so_linger = head->so_linger;		so->so_linger = head->so_linger;
so->so_state = head->so_state \| SS_NOFDREF;		so->so_state = head->so_state \| SS_NOFDREF;
so->so_fibnum = head->so_fibnum;		so->so_fibnum = head->so_fibnum;
so->so_proto = head->so_proto;		so->so_proto = head->so_proto;
so->so_cred = crhold(head->so_cred);		so->so_cred = crhold(head->so_cred);
#ifdef MAC		#ifdef MAC
mac_socket_newconn(head, so);		mac_socket_newconn(head, so);
#endif		#endif
knlist_init_mtx(&so->so_rcv.sb_sel.si_note, SOCKBUF_MTX(&so->so_rcv));		knlist_init(&so->so_rdsel.si_note, so, so_rdknl_lock, so_rdknl_unlock,
knlist_init_mtx(&so->so_snd.sb_sel.si_note, SOCKBUF_MTX(&so->so_snd));		so_rdknl_assert_locked, so_rdknl_assert_unlocked);
		knlist_init(&so->so_wrsel.si_note, so, so_wrknl_lock, so_wrknl_unlock,
		so_wrknl_assert_locked, so_wrknl_assert_unlocked);
VNET_SO_ASSERT(head);		VNET_SO_ASSERT(head);
if (soreserve(so, head->so_snd.sb_hiwat, head->so_rcv.sb_hiwat)) {		if (soreserve(so, head->sol_sbsnd_hiwat, head->sol_sbrcv_hiwat)) {
sodealloc(so);		sodealloc(so);
log(LOG_DEBUG, "%s: pcb %p: soreserve() failed\n",		log(LOG_DEBUG, "%s: pcb %p: soreserve() failed\n",
__func__, head->so_pcb);		__func__, head->so_pcb);
return (NULL);		return (NULL);
}		}
if ((*so->so_proto->pr_usrreqs->pru_attach)(so, 0, NULL)) {		if ((*so->so_proto->pr_usrreqs->pru_attach)(so, 0, NULL)) {
sodealloc(so);		sodealloc(so);
log(LOG_DEBUG, "%s: pcb %p: pru_attach() failed\n",		log(LOG_DEBUG, "%s: pcb %p: pru_attach() failed\n",
__func__, head->so_pcb);		__func__, head->so_pcb);
return (NULL);		return (NULL);
}		}
so->so_rcv.sb_lowat = head->so_rcv.sb_lowat;		so->so_rcv.sb_lowat = head->sol_sbrcv_lowat;
so->so_snd.sb_lowat = head->so_snd.sb_lowat;		so->so_snd.sb_lowat = head->sol_sbsnd_lowat;
so->so_rcv.sb_timeo = head->so_rcv.sb_timeo;		so->so_rcv.sb_timeo = head->sol_sbrcv_timeo;
so->so_snd.sb_timeo = head->so_snd.sb_timeo;		so->so_snd.sb_timeo = head->sol_sbsnd_timeo;
so->so_rcv.sb_flags \|= head->so_rcv.sb_flags & SB_AUTOSIZE;		so->so_rcv.sb_flags \|= head->sol_sbrcv_flags & SB_AUTOSIZE;
so->so_snd.sb_flags \|= head->so_snd.sb_flags & SB_AUTOSIZE;		so->so_snd.sb_flags \|= head->sol_sbsnd_flags & SB_AUTOSIZE;

		SOLISTEN_LOCK(head);
		if (head->sol_accept_filter != NULL)
		connstatus = 0;
so->so_state \|= connstatus;		so->so_state \|= connstatus;
ACCEPT_LOCK();		so->so_options = head->so_options & ~SO_ACCEPTCONN;
/*		soref(head); /* A socket on (in)complete queue refs head. */
* The accept socket may be tearing down but we just
* won a race on the ACCEPT_LOCK.
* However, if sctp_peeloff() is called on a 1-to-many
* style socket, the SO_ACCEPTCONN doesn't need to be set.
*/
if (!(head->so_options & SO_ACCEPTCONN) &&
((head->so_proto->pr_protocol != IPPROTO_SCTP) \|\|
(head->so_type != SOCK_SEQPACKET))) {
SOCK_LOCK(so);
so->so_head = NULL;
sofree(so); /* NB: returns ACCEPT_UNLOCK'ed. */
return (NULL);
}
if (connstatus) {		if (connstatus) {
TAILQ_INSERT_TAIL(&head->so_comp, so, so_list);		TAILQ_INSERT_TAIL(&head->sol_comp, so, so_list);
so->so_qstate \|= SQ_COMP;		so->so_qstate = SQ_COMP;
head->so_qlen++;		head->sol_qlen++;
		solisten_wakeup(head); /* unlocks */
} else {		} else {
/*		/*
* Keep removing sockets from the head until there's room for		* Keep removing sockets from the head until there's room for
* us to insert on the tail. In pre-locking revisions, this		* us to insert on the tail. In pre-locking revisions, this
* was a simple if(), but as we could be racing with other		* was a simple if(), but as we could be racing with other
* threads and soabort() requires dropping locks, we must		* threads and soabort() requires dropping locks, we must
* loop waiting for the condition to be true.		* loop waiting for the condition to be true.
*/		*/
while (head->so_incqlen > head->so_qlimit) {		while (head->sol_incqlen > head->sol_qlimit) {
struct socket *sp;		struct socket *sp;
sp = TAILQ_FIRST(&head->so_incomp);
TAILQ_REMOVE(&head->so_incomp, sp, so_list);		sp = TAILQ_FIRST(&head->sol_incomp);
head->so_incqlen--;		TAILQ_REMOVE(&head->sol_incomp, sp, so_list);
sp->so_qstate &= ~SQ_INCOMP;		head->sol_incqlen--;
sp->so_head = NULL;		SOCK_LOCK(sp);
ACCEPT_UNLOCK();		sp->so_qstate = SQ_NONE;
		sp->so_listen = NULL;
		SOCK_UNLOCK(sp);
		sorele(head); /* does SOLISTEN_UNLOCK, head stays */
soabort(sp);		soabort(sp);
ACCEPT_LOCK();		SOLISTEN_LOCK(head);
}		}
TAILQ_INSERT_TAIL(&head->so_incomp, so, so_list);		TAILQ_INSERT_TAIL(&head->sol_incomp, so, so_list);
so->so_qstate \|= SQ_INCOMP;		so->so_qstate = SQ_INCOMP;
head->so_incqlen++;		head->sol_incqlen++;
		SOLISTEN_UNLOCK(head);
}		}
ACCEPT_UNLOCK();		return (so);
if (connstatus) {
sorwakeup(head);
wakeup_one(&head->so_timeo);
}		}

		#ifdef SCTP
		/*
		* Socket part of sctp_peeloff(). Detach a new socket from an
		* association. The new socket is returned with a reference.
		*/
		struct socket *
		sopeeloff(struct socket *head)
		{
		struct socket *so;

		VNET_ASSERT(head->so_vnet != NULL, ("%s:%d so_vnet is NULL, head=%p",
		__func__, __LINE__, head));
		so = soalloc(head->so_vnet);
		if (so == NULL) {
		log(LOG_DEBUG, "%s: pcb %p: New socket allocation failure: "
		"limit reached or out of memory\n",
		__func__, head->so_pcb);
		return (NULL);
		}
		so->so_type = head->so_type;
		so->so_options = head->so_options;
		so->so_linger = head->so_linger;
		so->so_state = (head->so_state & SS_NBIO) \| SS_ISCONNECTED;
		so->so_fibnum = head->so_fibnum;
		so->so_proto = head->so_proto;
		so->so_cred = crhold(head->so_cred);
		#ifdef MAC
		mac_socket_newconn(head, so);
		#endif
		knlist_init(&so->so_rdsel.si_note, so, so_rdknl_lock, so_rdknl_unlock,
		so_rdknl_assert_locked, so_rdknl_assert_unlocked);
		knlist_init(&so->so_wrsel.si_note, so, so_wrknl_lock, so_wrknl_unlock,
		so_wrknl_assert_locked, so_wrknl_assert_unlocked);
		VNET_SO_ASSERT(head);
		if (soreserve(so, head->so_snd.sb_hiwat, head->so_rcv.sb_hiwat)) {
		sodealloc(so);
		log(LOG_DEBUG, "%s: pcb %p: soreserve() failed\n",
		__func__, head->so_pcb);
		return (NULL);
		}
		if ((*so->so_proto->pr_usrreqs->pru_attach)(so, 0, NULL)) {
		sodealloc(so);
		log(LOG_DEBUG, "%s: pcb %p: pru_attach() failed\n",
		__func__, head->so_pcb);
		return (NULL);
		}
		so->so_rcv.sb_lowat = head->so_rcv.sb_lowat;
		so->so_snd.sb_lowat = head->so_snd.sb_lowat;
		so->so_rcv.sb_timeo = head->so_rcv.sb_timeo;
		so->so_snd.sb_timeo = head->so_snd.sb_timeo;
		so->so_rcv.sb_flags \|= head->so_rcv.sb_flags & SB_AUTOSIZE;
		so->so_snd.sb_flags \|= head->so_snd.sb_flags & SB_AUTOSIZE;

		soref(so);

return (so);		return (so);
}		}
		#endif /* SCTP */

int		int
sobind(struct socket so, struct sockaddr nam, struct thread *td)		sobind(struct socket so, struct sockaddr nam, struct thread *td)
{		{
int error;		int error;

CURVNET_SET(so->so_vnet);		CURVNET_SET(so->so_vnet);
error = (*so->so_proto->pr_usrreqs->pru_bind)(so, nam, td);		error = (*so->so_proto->pr_usrreqs->pru_bind)(so, nam, td);
▲ Show 20 Lines • Show All 45 Lines • ▼ Show 20 Lines	if (so->so_state & (SS_ISCONNECTED \| SS_ISCONNECTING \|
SS_ISDISCONNECTING))		SS_ISDISCONNECTING))
return (EINVAL);		return (EINVAL);
return (0);		return (0);
}		}

void		void
solisten_proto(struct socket *so, int backlog)		solisten_proto(struct socket *so, int backlog)
{		{
		int sbrcv_lowat, sbsnd_lowat;
		u_int sbrcv_hiwat, sbsnd_hiwat;
		short sbrcv_flags, sbsnd_flags;
		sbintime_t sbrcv_timeo, sbsnd_timeo;

SOCK_LOCK_ASSERT(so);		SOCK_LOCK_ASSERT(so);

		if (SOLISTENING(so))
		goto listening;

		/*
		* Change this socket to listening state.
		*/
		sbrcv_lowat = so->so_rcv.sb_lowat;
		sbsnd_lowat = so->so_snd.sb_lowat;
		sbrcv_hiwat = so->so_rcv.sb_hiwat;
		sbsnd_hiwat = so->so_snd.sb_hiwat;
		sbrcv_flags = so->so_rcv.sb_flags;
		sbsnd_flags = so->so_snd.sb_flags;
		sbrcv_timeo = so->so_rcv.sb_timeo;
		sbsnd_timeo = so->so_snd.sb_timeo;

		sbdestroy(&so->so_snd, so);
		sbdestroy(&so->so_rcv, so);
		sx_destroy(&so->so_snd.sb_sx);
		sx_destroy(&so->so_rcv.sb_sx);
		SOCKBUF_LOCK_DESTROY(&so->so_snd);
		SOCKBUF_LOCK_DESTROY(&so->so_rcv);

		#ifdef INVARIANTS
		bzero(&so->so_rcv,
		sizeof(struct socket) - offsetof(struct socket, so_rcv));
		#endif

		so->sol_sbrcv_lowat = sbrcv_lowat;
		so->sol_sbsnd_lowat = sbsnd_lowat;
		so->sol_sbrcv_hiwat = sbrcv_hiwat;
		so->sol_sbsnd_hiwat = sbsnd_hiwat;
		so->sol_sbrcv_flags = sbrcv_flags;
		so->sol_sbsnd_flags = sbsnd_flags;
		so->sol_sbrcv_timeo = sbrcv_timeo;
		so->sol_sbsnd_timeo = sbsnd_timeo;

		so->sol_qlen = so->sol_incqlen = 0;
		TAILQ_INIT(&so->sol_incomp);
		TAILQ_INIT(&so->sol_comp);

		so->sol_accept_filter = NULL;
		so->sol_accept_filter_arg = NULL;
		so->sol_accept_filter_str = NULL;

		so->so_options \|= SO_ACCEPTCONN;

		listening:
if (backlog < 0 \|\| backlog > somaxconn)		if (backlog < 0 \|\| backlog > somaxconn)
backlog = somaxconn;		backlog = somaxconn;
so->so_qlimit = backlog;		so->sol_qlimit = backlog;
so->so_options \|= SO_ACCEPTCONN;
}		}

/*		/*
		* Wakeup listeners/subsystems once we have a complete connection.
		* Enters with lock, returns unlocked.
		*/
		void
		solisten_wakeup(struct socket *sol)
		{

		if (sol->sol_upcall != NULL)
		(void )sol->sol_upcall(sol, sol->sol_upcallarg, M_NOWAIT);
		else {
		selwakeuppri(&sol->so_rdsel, PSOCK);
		KNOTE_LOCKED(&sol->so_rdsel.si_note, 0);
		}
		SOLISTEN_UNLOCK(sol);
		wakeup_one(&sol->sol_comp);
		}

		/*
		* Return single connection off a listening socket queue. Main consumer of
		* the function is kern_accept4(). Some modules, that do their own accept
		* management also use the function.
		*
		* Listening socket must be locked on entry and is returned unlocked on
		* return.
		* The flags argument is set of accept4(2) flags and ACCEPT4_INHERIT.
		*/
		int
		solisten_dequeue(struct socket head, struct socket *ret, int flags)
		{
		struct socket *so;
		int error;

		SOLISTEN_LOCK_ASSERT(head);

		while (!(head->so_state & SS_NBIO) && TAILQ_EMPTY(&head->sol_comp) &&
		head->so_error == 0) {
		error = msleep(&head->sol_comp, &head->so_lock, PSOCK \| PCATCH,
		"accept", 0);
		if (error != 0) {
		SOLISTEN_UNLOCK(head);
		return (error);
		}
		}
		if (head->so_error) {
		error = head->so_error;
		head->so_error = 0;
		SOLISTEN_UNLOCK(head);
		return (error);
		}
		if ((head->so_state & SS_NBIO) && TAILQ_EMPTY(&head->sol_comp)) {
		SOLISTEN_UNLOCK(head);
		return (EWOULDBLOCK);
		}
		so = TAILQ_FIRST(&head->sol_comp);
		SOCK_LOCK(so);
		KASSERT(so->so_qstate == SQ_COMP,
		("%s: so %p not SQ_COMP", __func__, so));
		soref(so);
		head->sol_qlen--;
		so->so_qstate = SQ_NONE;
		so->so_listen = NULL;
		TAILQ_REMOVE(&head->sol_comp, so, so_list);
		if (flags & ACCEPT4_INHERIT)
		so->so_state \|= (head->so_state & SS_NBIO);
		else
		so->so_state \|= (flags & SOCK_NONBLOCK) ? SS_NBIO : 0;
		SOCK_UNLOCK(so);
		sorele(head);

		*ret = so;
		return (0);
		}

		/*
* Evaluate the reference count and named references on a socket; if no		* Evaluate the reference count and named references on a socket; if no
* references remain, free it. This should be called whenever a reference is		* references remain, free it. This should be called whenever a reference is
* released, such as in sorele(), but also when named reference flags are		* released, such as in sorele(), but also when named reference flags are
* cleared in socket or protocol code.		* cleared in socket or protocol code.
*		*
* sofree() will free the socket if:		* sofree() will free the socket if:
*		*
* - There are no outstanding file descriptor references or related consumers		* - There are no outstanding file descriptor references or related consumers
* (so_count == 0).		* (so_count == 0).
*		*
* - The socket has been closed by user space, if ever open (SS_NOFDREF).		* - The socket has been closed by user space, if ever open (SS_NOFDREF).
*		*
* - The protocol does not have an outstanding strong reference on the socket		* - The protocol does not have an outstanding strong reference on the socket
* (SS_PROTOREF).		* (SS_PROTOREF).
*		*
* - The socket is not in a completed connection queue, so a process has been		* - The socket is not in a completed connection queue, so a process has been
* notified that it is present. If it is removed, the user process may		* notified that it is present. If it is removed, the user process may
* block in accept() despite select() saying the socket was ready.		* block in accept() despite select() saying the socket was ready.
*/		*/
void		void
sofree(struct socket *so)		sofree(struct socket *so)
{		{
struct protosw *pr = so->so_proto;		struct protosw *pr = so->so_proto;
struct socket *head;

ACCEPT_LOCK_ASSERT();
SOCK_LOCK_ASSERT(so);		SOCK_LOCK_ASSERT(so);

if ((so->so_state & SS_NOFDREF) == 0 \|\| so->so_count != 0 \|\|		if ((so->so_state & SS_NOFDREF) == 0 \|\| so->so_count != 0 \|\|
(so->so_state & SS_PROTOREF) \|\| (so->so_qstate & SQ_COMP)) {		(so->so_state & SS_PROTOREF) \|\| (so->so_qstate == SQ_COMP)) {
SOCK_UNLOCK(so);		SOCK_UNLOCK(so);
ACCEPT_UNLOCK();
return;		return;
}		}

head = so->so_head;		if (!SOLISTENING(so) && so->so_qstate == SQ_INCOMP) {
if (head != NULL) {		struct socket *sol;
KASSERT((so->so_qstate & SQ_COMP) != 0 \|\|
(so->so_qstate & SQ_INCOMP) != 0,		sol = so->so_listen;
("sofree: so_head != NULL, but neither SQ_COMP nor "		KASSERT(sol, ("%s: so %p on incomp of NULL", __func__, so));
"SQ_INCOMP"));
KASSERT((so->so_qstate & SQ_COMP) == 0 \|\|		/*
(so->so_qstate & SQ_INCOMP) == 0,		* To solve race between close of a listening socket and
("sofree: so->so_qstate is SQ_COMP and also SQ_INCOMP"));		* a socket on its incomplete queue, we need to lock both.
TAILQ_REMOVE(&head->so_incomp, so, so_list);		* The order is first listening socket, then regular.
head->so_incqlen--;		* Since we don't have SS_NOFDREF neither SS_PROTOREF, this
so->so_qstate &= ~SQ_INCOMP;		* function and the listening socket are the only pointers
so->so_head = NULL;		* to so. To preserve so and sol, we reference both and then
		* relock.
		* After relock the socket may not move to so_comp since it
		* doesn't have PCB already, but it may be removed from
		* so_incomp. If that happens, we share responsiblity on
		* freeing the socket, but soclose() has already removed
		* it from queue.
		*/
		soref(sol);
		soref(so);
		SOCK_UNLOCK(so);
		SOLISTEN_LOCK(sol);
		SOCK_LOCK(so);
		if (so->so_qstate == SQ_INCOMP) {
		KASSERT(so->so_listen == sol,
		("%s: so %p migrated out of sol %p",
		__func__, so, sol));
		TAILQ_REMOVE(&sol->sol_incomp, so, so_list);
		sol->sol_incqlen--;
		/* This is guarenteed not to be the last. */
		refcount_release(&sol->so_count);
		so->so_qstate = SQ_NONE;
		so->so_listen = NULL;
		} else
		KASSERT(so->so_listen == NULL,
		("%s: so %p not on (in)comp with so_listen",
		__func__, so));
		sorele(sol);
		KASSERT(so->so_count == 1,
		("%s: so %p count %u", __func__, so, so->so_count));
		so->so_count = 0;
}		}
KASSERT((so->so_qstate & SQ_COMP) == 0 &&		if (SOLISTENING(so))
(so->so_qstate & SQ_INCOMP) == 0,		so->so_error = ECONNABORTED;
("sofree: so_head == NULL, but still SQ_COMP(%d) or SQ_INCOMP(%d)",
so->so_qstate & SQ_COMP, so->so_qstate & SQ_INCOMP));
if (so->so_options & SO_ACCEPTCONN) {
KASSERT((TAILQ_EMPTY(&so->so_comp)),
("sofree: so_comp populated"));
KASSERT((TAILQ_EMPTY(&so->so_incomp)),
("sofree: so_incomp populated"));
}
SOCK_UNLOCK(so);		SOCK_UNLOCK(so);
ACCEPT_UNLOCK();

VNET_SO_ASSERT(so);		VNET_SO_ASSERT(so);
if (pr->pr_flags & PR_RIGHTS && pr->pr_domain->dom_dispose != NULL)		if (pr->pr_flags & PR_RIGHTS && pr->pr_domain->dom_dispose != NULL)
(*pr->pr_domain->dom_dispose)(so);		(*pr->pr_domain->dom_dispose)(so);
if (pr->pr_usrreqs->pru_detach != NULL)		if (pr->pr_usrreqs->pru_detach != NULL)
(*pr->pr_usrreqs->pru_detach)(so);		(*pr->pr_usrreqs->pru_detach)(so);

/*		/*
* From this point on, we assume that no other references to this		* From this point on, we assume that no other references to this
* socket exist anywhere else in the stack. Therefore, no locks need		* socket exist anywhere else in the stack. Therefore, no locks need
* to be acquired or held.		* to be acquired or held.
*		*
* We used to do a lot of socket buffer and socket locking here, as		* We used to do a lot of socket buffer and socket locking here, as
* well as invoke sorflush() and perform wakeups. The direct call to		* well as invoke sorflush() and perform wakeups. The direct call to
* dom_dispose() and sbrelease_internal() are an inlining of what was		* dom_dispose() and sbrelease_internal() are an inlining of what was
* necessary from sorflush().		* necessary from sorflush().
*		*
* Notice that the socket buffer and kqueue state are torn down		* Notice that the socket buffer and kqueue state are torn down
* before calling pru_detach. This means that protocols shold not		* before calling pru_detach. This means that protocols shold not
* assume they can perform socket wakeups, etc, in their detach code.		* assume they can perform socket wakeups, etc, in their detach code.
*/		*/
		if (!SOLISTENING(so)) {
sbdestroy(&so->so_snd, so);		sbdestroy(&so->so_snd, so);
sbdestroy(&so->so_rcv, so);		sbdestroy(&so->so_rcv, so);
seldrain(&so->so_snd.sb_sel);		}
seldrain(&so->so_rcv.sb_sel);		seldrain(&so->so_rdsel);
knlist_destroy(&so->so_rcv.sb_sel.si_note);		seldrain(&so->so_wrsel);
knlist_destroy(&so->so_snd.sb_sel.si_note);		knlist_destroy(&so->so_rdsel.si_note);
		knlist_destroy(&so->so_wrsel.si_note);
sodealloc(so);		sodealloc(so);
}		}

/*		/*
* Close a socket on last file table reference removal. Initiate disconnect		* Close a socket on last file table reference removal. Initiate disconnect
* if connected. Free socket when disconnect complete.		* if connected. Free socket when disconnect complete.
*		*
* This function will sorele() the socket. Note that soclose() may be called		* This function will sorele() the socket. Note that soclose() may be called
* prior to the ref count reaching zero. The actual socket structure will		* prior to the ref count reaching zero. The actual socket structure will
* not be freed until the ref count reaches zero.		* not be freed until the ref count reaches zero.
*/		*/
int		int
soclose(struct socket *so)		soclose(struct socket *so)
{		{
		struct accept_queue lqueue;
		bool listening;
int error = 0;		int error = 0;

KASSERT(!(so->so_state & SS_NOFDREF), ("soclose: SS_NOFDREF on enter"));		KASSERT(!(so->so_state & SS_NOFDREF), ("soclose: SS_NOFDREF on enter"));

CURVNET_SET(so->so_vnet);		CURVNET_SET(so->so_vnet);
funsetown(&so->so_sigio);		funsetown(&so->so_sigio);
if (so->so_state & SS_ISCONNECTED) {		if (so->so_state & SS_ISCONNECTED) {
if ((so->so_state & SS_ISDISCONNECTING) == 0) {		if ((so->so_state & SS_ISDISCONNECTING) == 0) {
Show All 16 Lines	if (so->so_options & SO_LINGER) {
break;		break;
}		}
}		}
}		}

drop:		drop:
if (so->so_proto->pr_usrreqs->pru_close != NULL)		if (so->so_proto->pr_usrreqs->pru_close != NULL)
(*so->so_proto->pr_usrreqs->pru_close)(so);		(*so->so_proto->pr_usrreqs->pru_close)(so);
ACCEPT_LOCK();
if (so->so_options & SO_ACCEPTCONN) {		SOCK_LOCK(so);
		if ((listening = (so->so_options & SO_ACCEPTCONN))) {
struct socket *sp;		struct socket *sp;
/*
* Prevent new additions to the accept queues due		TAILQ_INIT(&lqueue);
* to ACCEPT_LOCK races while we are draining them.		TAILQ_SWAP(&lqueue, &so->sol_incomp, socket, so_list);
*/		TAILQ_CONCAT(&lqueue, &so->sol_comp, so_list);
so->so_options &= ~SO_ACCEPTCONN;
while ((sp = TAILQ_FIRST(&so->so_incomp)) != NULL) {		so->sol_qlen = so->sol_incqlen = 0;
TAILQ_REMOVE(&so->so_incomp, sp, so_list);
so->so_incqlen--;		TAILQ_FOREACH(sp, &lqueue, so_list) {
sp->so_qstate &= ~SQ_INCOMP;		SOCK_LOCK(sp);
sp->so_head = NULL;		sp->so_qstate = SQ_NONE;
ACCEPT_UNLOCK();		sp->so_listen = NULL;
soabort(sp);		SOCK_UNLOCK(sp);
ACCEPT_LOCK();		/* Guaranteed not to be the last. */
		refcount_release(&so->so_count);
}		}
while ((sp = TAILQ_FIRST(&so->so_comp)) != NULL) {
TAILQ_REMOVE(&so->so_comp, sp, so_list);
so->so_qlen--;
sp->so_qstate &= ~SQ_COMP;
sp->so_head = NULL;
ACCEPT_UNLOCK();
soabort(sp);
ACCEPT_LOCK();
}		}
KASSERT((TAILQ_EMPTY(&so->so_comp)),
("%s: so_comp populated", __func__));
KASSERT((TAILQ_EMPTY(&so->so_incomp)),
("%s: so_incomp populated", __func__));
}
SOCK_LOCK(so);
KASSERT((so->so_state & SS_NOFDREF) == 0, ("soclose: NOFDREF"));		KASSERT((so->so_state & SS_NOFDREF) == 0, ("soclose: NOFDREF"));
so->so_state \|= SS_NOFDREF;		so->so_state \|= SS_NOFDREF;
sorele(so); /* NB: Returns with ACCEPT_UNLOCK(). */		sorele(so);
		if (listening) {
		struct socket *sp;

		TAILQ_FOREACH(sp, &lqueue, so_list) {
		SOCK_LOCK(sp);
		if (sp->so_count == 0) {
		SOCK_UNLOCK(sp);
		soabort(sp);
		} else
		/* sp is now in sofree() */
		SOCK_UNLOCK(sp);
		}
		}
CURVNET_RESTORE();		CURVNET_RESTORE();
return (error);		return (error);
}		}

/*		/*
* soabort() is used to abruptly tear down a connection, such as when a		* soabort() is used to abruptly tear down a connection, such as when a
* resource limit is reached (listen queue depth exceeded), or if a listen		* resource limit is reached (listen queue depth exceeded), or if a listen
* socket is closed while there are sockets waiting to be accepted.		* socket is closed while there are sockets waiting to be accepted.
Show All 15 Lines	soabort(struct socket *so)
* In as much as is possible, assert that no references to this		* In as much as is possible, assert that no references to this
* socket are held. This is not quite the same as asserting that the		* socket are held. This is not quite the same as asserting that the
* current thread is responsible for arranging for no references, but		* current thread is responsible for arranging for no references, but
* is as close as we can get for now.		* is as close as we can get for now.
*/		*/
KASSERT(so->so_count == 0, ("soabort: so_count"));		KASSERT(so->so_count == 0, ("soabort: so_count"));
KASSERT((so->so_state & SS_PROTOREF) == 0, ("soabort: SS_PROTOREF"));		KASSERT((so->so_state & SS_PROTOREF) == 0, ("soabort: SS_PROTOREF"));
KASSERT(so->so_state & SS_NOFDREF, ("soabort: !SS_NOFDREF"));		KASSERT(so->so_state & SS_NOFDREF, ("soabort: !SS_NOFDREF"));
KASSERT((so->so_state & SQ_COMP) == 0, ("soabort: SQ_COMP"));		KASSERT(so->so_qstate == SQ_NONE, ("soabort: !SQ_NONE"));
KASSERT((so->so_state & SQ_INCOMP) == 0, ("soabort: SQ_INCOMP"));
VNET_SO_ASSERT(so);		VNET_SO_ASSERT(so);

if (so->so_proto->pr_usrreqs->pru_abort != NULL)		if (so->so_proto->pr_usrreqs->pru_abort != NULL)
(*so->so_proto->pr_usrreqs->pru_abort)(so);		(*so->so_proto->pr_usrreqs->pru_abort)(so);
ACCEPT_LOCK();
SOCK_LOCK(so);		SOCK_LOCK(so);
sofree(so);		sofree(so);
}		}

int		int
soaccept(struct socket so, struct sockaddr *nam)		soaccept(struct socket so, struct sockaddr *nam)
{		{
int error;		int error;
▲ Show 20 Lines • Show All 1,918 Lines • ▼ Show 20 Lines	#ifdef MAC
goto bad;		goto bad;
error = sooptcopyout(sopt, &extmac, sizeof extmac);		error = sooptcopyout(sopt, &extmac, sizeof extmac);
#else		#else
error = EOPNOTSUPP;		error = EOPNOTSUPP;
#endif		#endif
break;		break;

case SO_LISTENQLIMIT:		case SO_LISTENQLIMIT:
optval = so->so_qlimit;		optval = SOLISTENING(so) ? so->sol_qlimit : 0;
goto integer;		goto integer;

case SO_LISTENQLEN:		case SO_LISTENQLEN:
optval = so->so_qlen;		optval = SOLISTENING(so) ? so->sol_qlen : 0;
goto integer;		goto integer;

case SO_LISTENINCQLEN:		case SO_LISTENINCQLEN:
optval = so->so_incqlen;		optval = SOLISTENING(so) ? so->sol_incqlen : 0;
goto integer;		goto integer;

case SO_TS_CLOCK:		case SO_TS_CLOCK:
optval = so->so_ts_clock;		optval = so->so_ts_clock;
goto integer;		goto integer;

case SO_MAX_PACING_RATE:		case SO_MAX_PACING_RATE:
optval = so->so_max_pacing_rate;		optval = so->so_max_pacing_rate;
▲ Show 20 Lines • Show All 130 Lines • ▼ Show 20 Lines
* out-of-band data, which will then notify socket consumers.		* out-of-band data, which will then notify socket consumers.
*/		*/
void		void
sohasoutofband(struct socket *so)		sohasoutofband(struct socket *so)
{		{

if (so->so_sigio != NULL)		if (so->so_sigio != NULL)
pgsigio(&so->so_sigio, SIGURG, 0);		pgsigio(&so->so_sigio, SIGURG, 0);
selwakeuppri(&so->so_rcv.sb_sel, PSOCK);		selwakeuppri(&so->so_rdsel, PSOCK);
}		}

int		int
sopoll(struct socket so, int events, struct ucred active_cred,		sopoll(struct socket so, int events, struct ucred active_cred,
struct thread *td)		struct thread *td)
{		{

/*		/*
* We do not need to set or assert curvnet as long as everyone uses		* We do not need to set or assert curvnet as long as everyone uses
* sopoll_generic().		* sopoll_generic().
*/		*/
return (so->so_proto->pr_usrreqs->pru_sopoll(so, events, active_cred,		return (so->so_proto->pr_usrreqs->pru_sopoll(so, events, active_cred,
td));		td));
}		}

int		int
sopoll_generic(struct socket so, int events, struct ucred active_cred,		sopoll_generic(struct socket so, int events, struct ucred active_cred,
struct thread *td)		struct thread *td)
{		{
int revents = 0;		int revents;

		SOCK_LOCK(so);
		if (SOLISTENING(so)) {
		if (!(events & (POLLIN \| POLLRDNORM)))
		revents = 0;
		else if (!TAILQ_EMPTY(&so->sol_comp))
		revents = events & (POLLIN \| POLLRDNORM);
		else {
		selrecord(td, &so->so_rdsel);
		revents = 0;
		}
		} else {
		revents = 0;
SOCKBUF_LOCK(&so->so_snd);		SOCKBUF_LOCK(&so->so_snd);
SOCKBUF_LOCK(&so->so_rcv);		SOCKBUF_LOCK(&so->so_rcv);
if (events & (POLLIN \| POLLRDNORM))		if (events & (POLLIN \| POLLRDNORM))
if (soreadabledata(so))		if (soreadabledata(so))
revents \|= events & (POLLIN \| POLLRDNORM);		revents \|= events & (POLLIN \| POLLRDNORM);

if (events & (POLLOUT \| POLLWRNORM))		if (events & (POLLOUT \| POLLWRNORM))
if (sowriteable(so))		if (sowriteable(so))
revents \|= events & (POLLOUT \| POLLWRNORM);		revents \|= events & (POLLOUT \| POLLWRNORM);

if (events & (POLLPRI \| POLLRDBAND))		if (events & (POLLPRI \| POLLRDBAND))
if (so->so_oobmark \|\| (so->so_rcv.sb_state & SBS_RCVATMARK))		if (so->so_oobmark \|\|
		(so->so_rcv.sb_state & SBS_RCVATMARK))
revents \|= events & (POLLPRI \| POLLRDBAND);		revents \|= events & (POLLPRI \| POLLRDBAND);

if ((events & POLLINIGNEOF) == 0) {		if ((events & POLLINIGNEOF) == 0) {
if (so->so_rcv.sb_state & SBS_CANTRCVMORE) {		if (so->so_rcv.sb_state & SBS_CANTRCVMORE) {
revents \|= events & (POLLIN \| POLLRDNORM);		revents \|= events & (POLLIN \| POLLRDNORM);
if (so->so_snd.sb_state & SBS_CANTSENDMORE)		if (so->so_snd.sb_state & SBS_CANTSENDMORE)
revents \|= POLLHUP;		revents \|= POLLHUP;
}		}
}		}

if (revents == 0) {		if (revents == 0) {
if (events & (POLLIN \| POLLPRI \| POLLRDNORM \| POLLRDBAND)) {		if (events &
selrecord(td, &so->so_rcv.sb_sel);		(POLLIN \| POLLPRI \| POLLRDNORM \| POLLRDBAND)) {
		selrecord(td, &so->so_rdsel);
so->so_rcv.sb_flags \|= SB_SEL;		so->so_rcv.sb_flags \|= SB_SEL;
}		}

if (events & (POLLOUT \| POLLWRNORM)) {		if (events & (POLLOUT \| POLLWRNORM)) {
selrecord(td, &so->so_snd.sb_sel);		selrecord(td, &so->so_wrsel);
so->so_snd.sb_flags \|= SB_SEL;		so->so_snd.sb_flags \|= SB_SEL;
}		}
}		}

SOCKBUF_UNLOCK(&so->so_rcv);		SOCKBUF_UNLOCK(&so->so_rcv);
SOCKBUF_UNLOCK(&so->so_snd);		SOCKBUF_UNLOCK(&so->so_snd);
		}
		SOCK_UNLOCK(so);
return (revents);		return (revents);
}		}

int		int
soo_kqfilter(struct file fp, struct knote kn)		soo_kqfilter(struct file fp, struct knote kn)
{		{
struct socket *so = kn->kn_fp->f_data;		struct socket *so = kn->kn_fp->f_data;
struct sockbuf *sb;		struct sockbuf *sb;
		struct knlist *knl;

switch (kn->kn_filter) {		switch (kn->kn_filter) {
case EVFILT_READ:		case EVFILT_READ:
kn->kn_fop = &soread_filtops;		kn->kn_fop = &soread_filtops;
		knl = &so->so_rdsel.si_note;
sb = &so->so_rcv;		sb = &so->so_rcv;
break;		break;
case EVFILT_WRITE:		case EVFILT_WRITE:
kn->kn_fop = &sowrite_filtops;		kn->kn_fop = &sowrite_filtops;
		knl = &so->so_wrsel.si_note;
sb = &so->so_snd;		sb = &so->so_snd;
break;		break;
case EVFILT_EMPTY:		case EVFILT_EMPTY:
kn->kn_fop = &soempty_filtops;		kn->kn_fop = &soempty_filtops;
		knl = &so->so_wrsel.si_note;
sb = &so->so_snd;		sb = &so->so_snd;
break;		break;
default:		default:
return (EINVAL);		return (EINVAL);
}		}

		SOCK_LOCK(so);
		if (SOLISTENING(so)) {
		knlist_add(knl, kn, 1);
		} else {
SOCKBUF_LOCK(sb);		SOCKBUF_LOCK(sb);
knlist_add(&sb->sb_sel.si_note, kn, 1);		knlist_add(knl, kn, 1);
sb->sb_flags \|= SB_KNOTE;		sb->sb_flags \|= SB_KNOTE;
SOCKBUF_UNLOCK(sb);		SOCKBUF_UNLOCK(sb);
		}
		SOCK_UNLOCK(so);
return (0);		return (0);
}		}

/*		/*
* Some routines that return EOPNOTSUPP for entry points that are not		* Some routines that return EOPNOTSUPP for entry points that are not
* supported by a protocol. Fill in as needed.		* supported by a protocol. Fill in as needed.
*/		*/
int		int
▲ Show 20 Lines • Show All 162 Lines • ▼ Show 20 Lines	pru_sopoll_notsupp(struct socket so, int events, struct ucred cred,
return EOPNOTSUPP;		return EOPNOTSUPP;
}		}

static void		static void
filt_sordetach(struct knote *kn)		filt_sordetach(struct knote *kn)
{		{
struct socket *so = kn->kn_fp->f_data;		struct socket *so = kn->kn_fp->f_data;

SOCKBUF_LOCK(&so->so_rcv);		so_rdknl_lock(so);
knlist_remove(&so->so_rcv.sb_sel.si_note, kn, 1);		knlist_remove(&so->so_rdsel.si_note, kn, 1);
if (knlist_empty(&so->so_rcv.sb_sel.si_note))		if (!SOLISTENING(so) && knlist_empty(&so->so_rdsel.si_note))
so->so_rcv.sb_flags &= ~SB_KNOTE;		so->so_rcv.sb_flags &= ~SB_KNOTE;
SOCKBUF_UNLOCK(&so->so_rcv);		so_rdknl_unlock(so);
}		}

/ARGSUSED/		/ARGSUSED/
static int		static int
filt_soread(struct knote *kn, long hint)		filt_soread(struct knote *kn, long hint)
{		{
struct socket *so;		struct socket *so;

so = kn->kn_fp->f_data;		so = kn->kn_fp->f_data;
if (so->so_options & SO_ACCEPTCONN) {
kn->kn_data = so->so_qlen;
return (!TAILQ_EMPTY(&so->so_comp));

		if (SOLISTENING(so)) {
		SOCK_LOCK_ASSERT(so);
		kn->kn_data = so->sol_qlen;
		return (!TAILQ_EMPTY(&so->sol_comp));
}		}

SOCKBUF_LOCK_ASSERT(&so->so_rcv);		SOCKBUF_LOCK_ASSERT(&so->so_rcv);

kn->kn_data = sbavail(&so->so_rcv) - so->so_rcv.sb_ctl;		kn->kn_data = sbavail(&so->so_rcv) - so->so_rcv.sb_ctl;
if (so->so_rcv.sb_state & SBS_CANTRCVMORE) {		if (so->so_rcv.sb_state & SBS_CANTRCVMORE) {
kn->kn_flags \|= EV_EOF;		kn->kn_flags \|= EV_EOF;
kn->kn_fflags = so->so_error;		kn->kn_fflags = so->so_error;
return (1);		return (1);
} else if (so->so_error) /* temporary udp error */		} else if (so->so_error) /* temporary udp error */
Show All 9 Lines	filt_soread(struct knote *kn, long hint)
return (hhook_run_socket(so, NULL, HHOOK_FILT_SOREAD));		return (hhook_run_socket(so, NULL, HHOOK_FILT_SOREAD));
}		}

static void		static void
filt_sowdetach(struct knote *kn)		filt_sowdetach(struct knote *kn)
{		{
struct socket *so = kn->kn_fp->f_data;		struct socket *so = kn->kn_fp->f_data;

SOCKBUF_LOCK(&so->so_snd);		so_wrknl_lock(so);
knlist_remove(&so->so_snd.sb_sel.si_note, kn, 1);		knlist_remove(&so->so_wrsel.si_note, kn, 1);
if (knlist_empty(&so->so_snd.sb_sel.si_note))		if (!SOLISTENING(so) && knlist_empty(&so->so_wrsel.si_note))
so->so_snd.sb_flags &= ~SB_KNOTE;		so->so_snd.sb_flags &= ~SB_KNOTE;
SOCKBUF_UNLOCK(&so->so_snd);		so_wrknl_unlock(so);
}		}

/ARGSUSED/		/ARGSUSED/
static int		static int
filt_sowrite(struct knote *kn, long hint)		filt_sowrite(struct knote *kn, long hint)
{		{
struct socket *so;		struct socket *so;

so = kn->kn_fp->f_data;		so = kn->kn_fp->f_data;

		if (SOLISTENING(so))
		return (0);

SOCKBUF_LOCK_ASSERT(&so->so_snd);		SOCKBUF_LOCK_ASSERT(&so->so_snd);
kn->kn_data = sbspace(&so->so_snd);		kn->kn_data = sbspace(&so->so_snd);

hhook_run_socket(so, kn, HHOOK_FILT_SOWRITE);		hhook_run_socket(so, kn, HHOOK_FILT_SOWRITE);

if (so->so_snd.sb_state & SBS_CANTSENDMORE) {		if (so->so_snd.sb_state & SBS_CANTSENDMORE) {
kn->kn_flags \|= EV_EOF;		kn->kn_flags \|= EV_EOF;
kn->kn_fflags = so->so_error;		kn->kn_fflags = so->so_error;
Show All 10 Lines
}		}

static int		static int
filt_soempty(struct knote *kn, long hint)		filt_soempty(struct knote *kn, long hint)
{		{
struct socket *so;		struct socket *so;

so = kn->kn_fp->f_data;		so = kn->kn_fp->f_data;

		if (SOLISTENING(so))
		return (1);

SOCKBUF_LOCK_ASSERT(&so->so_snd);		SOCKBUF_LOCK_ASSERT(&so->so_snd);
kn->kn_data = sbused(&so->so_snd);		kn->kn_data = sbused(&so->so_snd);

if (kn->kn_data == 0)		if (kn->kn_data == 0)
return (1);		return (1);
else		else
return (0);		return (0);
}		}
▲ Show 20 Lines • Show All 52 Lines • ▼ Show 20 Lines
}		}

void		void
soisconnected(struct socket *so)		soisconnected(struct socket *so)
{		{
struct socket *head;		struct socket *head;
int ret;		int ret;

		/*
		* XXXGL: this is the only place where we acquire socket locks
		* in reverse order: first child, then listening socket. To
		* avoid possible LOR, use try semantics.
		*/
restart:		restart:
ACCEPT_LOCK();
SOCK_LOCK(so);		SOCK_LOCK(so);
		if ((head = so->so_listen) != NULL &&
		__predict_false(SOLISTEN_TRYLOCK(head) == 0)) {
		SOCK_UNLOCK(so);
		goto restart;
		}
so->so_state &= ~(SS_ISCONNECTING\|SS_ISDISCONNECTING\|SS_ISCONFIRMING);		so->so_state &= ~(SS_ISCONNECTING\|SS_ISDISCONNECTING\|SS_ISCONFIRMING);
so->so_state \|= SS_ISCONNECTED;		so->so_state \|= SS_ISCONNECTED;
head = so->so_head;		if (head != NULL && (so->so_qstate == SQ_INCOMP)) {
if (head != NULL && (so->so_qstate & SQ_INCOMP)) {		again:
if ((so->so_options & SO_ACCEPTFILTER) == 0) {		if ((so->so_options & SO_ACCEPTFILTER) == 0) {
		TAILQ_REMOVE(&head->sol_incomp, so, so_list);
		head->sol_incqlen--;
		TAILQ_INSERT_TAIL(&head->sol_comp, so, so_list);
		head->sol_qlen++;
		so->so_qstate = SQ_COMP;
SOCK_UNLOCK(so);		SOCK_UNLOCK(so);
TAILQ_REMOVE(&head->so_incomp, so, so_list);		solisten_wakeup(head); /* unlocks */
head->so_incqlen--;
so->so_qstate &= ~SQ_INCOMP;
TAILQ_INSERT_TAIL(&head->so_comp, so, so_list);
head->so_qlen++;
so->so_qstate \|= SQ_COMP;
ACCEPT_UNLOCK();
sorwakeup(head);
wakeup_one(&head->so_timeo);
} else {		} else {
ACCEPT_UNLOCK();		SOCKBUF_LOCK(&so->so_rcv);
soupcall_set(so, SO_RCV,		soupcall_set(so, SO_RCV,
head->so_accf->so_accept_filter->accf_callback,		head->sol_accept_filter->accf_callback,
head->so_accf->so_accept_filter_arg);		head->sol_accept_filter_arg);
so->so_options &= ~SO_ACCEPTFILTER;		so->so_options &= ~SO_ACCEPTFILTER;
ret = head->so_accf->so_accept_filter->accf_callback(so,		ret = head->sol_accept_filter->accf_callback(so,
head->so_accf->so_accept_filter_arg, M_NOWAIT);		head->sol_accept_filter_arg, M_NOWAIT);
if (ret == SU_ISCONNECTED)		if (ret == SU_ISCONNECTED) {
soupcall_clear(so, SO_RCV);		soupcall_clear(so, SO_RCV);
		SOCKBUF_UNLOCK(&so->so_rcv);
		goto again;
		}
		SOCKBUF_UNLOCK(&so->so_rcv);
SOCK_UNLOCK(so);		SOCK_UNLOCK(so);
if (ret == SU_ISCONNECTED)		SOLISTEN_UNLOCK(head);
goto restart;
}		}
return;		return;
}		}
		if (head != NULL)
		SOLISTEN_UNLOCK(head);
SOCK_UNLOCK(so);		SOCK_UNLOCK(so);
ACCEPT_UNLOCK();
wakeup(&so->so_timeo);		wakeup(&so->so_timeo);
sorwakeup(so);		sorwakeup(so);
sowwakeup(so);		sowwakeup(so);
}		}

void		void
soisdisconnecting(struct socket *so)		soisdisconnecting(struct socket *so)
{		{

/*		SOCK_LOCK(so);
* Note: This code assumes that SOCK_LOCK(so) and
* SOCKBUF_LOCK(&so->so_rcv) are the same.
*/
SOCKBUF_LOCK(&so->so_rcv);
so->so_state &= ~SS_ISCONNECTING;		so->so_state &= ~SS_ISCONNECTING;
so->so_state \|= SS_ISDISCONNECTING;		so->so_state \|= SS_ISDISCONNECTING;

		if (!SOLISTENING(so)) {
		SOCKBUF_LOCK(&so->so_rcv);
socantrcvmore_locked(so);		socantrcvmore_locked(so);
SOCKBUF_LOCK(&so->so_snd);		SOCKBUF_LOCK(&so->so_snd);
socantsendmore_locked(so);		socantsendmore_locked(so);
		}
		SOCK_UNLOCK(so);
wakeup(&so->so_timeo);		wakeup(&so->so_timeo);
}		}

void		void
soisdisconnected(struct socket *so)		soisdisconnected(struct socket *so)
{		{

/*		SOCK_LOCK(so);
* Note: This code assumes that SOCK_LOCK(so) and
* SOCKBUF_LOCK(&so->so_rcv) are the same.
*/
SOCKBUF_LOCK(&so->so_rcv);
so->so_state &= ~(SS_ISCONNECTING\|SS_ISCONNECTED\|SS_ISDISCONNECTING);		so->so_state &= ~(SS_ISCONNECTING\|SS_ISCONNECTED\|SS_ISDISCONNECTING);
so->so_state \|= SS_ISDISCONNECTED;		so->so_state \|= SS_ISDISCONNECTED;

		if (!SOLISTENING(so)) {
		SOCKBUF_LOCK(&so->so_rcv);
socantrcvmore_locked(so);		socantrcvmore_locked(so);
SOCKBUF_LOCK(&so->so_snd);		SOCKBUF_LOCK(&so->so_snd);
sbdrop_locked(&so->so_snd, sbused(&so->so_snd));		sbdrop_locked(&so->so_snd, sbused(&so->so_snd));
socantsendmore_locked(so);		socantsendmore_locked(so);
		}
		SOCK_UNLOCK(so);
wakeup(&so->so_timeo);		wakeup(&so->so_timeo);
}		}

/*		/*
* Make a copy of a sockaddr in a malloced buffer of type M_SONAME.		* Make a copy of a sockaddr in a malloced buffer of type M_SONAME.
*/		*/
struct sockaddr *		struct sockaddr *
sodupsockaddr(const struct sockaddr *sa, int mflags)		sodupsockaddr(const struct sockaddr *sa, int mflags)
Show All 9 Lines
/*		/*
* Register per-socket buffer upcalls.		* Register per-socket buffer upcalls.
*/		*/
void		void
soupcall_set(struct socket so, int which, so_upcall_t func, void arg)		soupcall_set(struct socket so, int which, so_upcall_t func, void arg)
{		{
struct sockbuf *sb;		struct sockbuf *sb;

		KASSERT(!SOLISTENING(so), ("%s: so %p listening", __func__, so));

switch (which) {		switch (which) {
case SO_RCV:		case SO_RCV:
sb = &so->so_rcv;		sb = &so->so_rcv;
break;		break;
case SO_SND:		case SO_SND:
sb = &so->so_snd;		sb = &so->so_snd;
break;		break;
default:		default:
panic("soupcall_set: bad which");		panic("soupcall_set: bad which");
}		}
SOCKBUF_LOCK_ASSERT(sb);		SOCKBUF_LOCK_ASSERT(sb);
sb->sb_upcall = func;		sb->sb_upcall = func;
sb->sb_upcallarg = arg;		sb->sb_upcallarg = arg;
sb->sb_flags \|= SB_UPCALL;		sb->sb_flags \|= SB_UPCALL;
}		}

void		void
soupcall_clear(struct socket *so, int which)		soupcall_clear(struct socket *so, int which)
{		{
struct sockbuf *sb;		struct sockbuf *sb;

		KASSERT(!SOLISTENING(so), ("%s: so %p listening", __func__, so));

switch (which) {		switch (which) {
case SO_RCV:		case SO_RCV:
sb = &so->so_rcv;		sb = &so->so_rcv;
break;		break;
case SO_SND:		case SO_SND:
sb = &so->so_snd;		sb = &so->so_snd;
break;		break;
default:		default:
panic("soupcall_clear: bad which");		panic("soupcall_clear: bad which");
}		}
SOCKBUF_LOCK_ASSERT(sb);		SOCKBUF_LOCK_ASSERT(sb);
KASSERT(sb->sb_upcall != NULL, ("soupcall_clear: no upcall to clear"));		KASSERT(sb->sb_upcall != NULL,
		("%s: so %p no upcall to clear", __func__, so));
sb->sb_upcall = NULL;		sb->sb_upcall = NULL;
sb->sb_upcallarg = NULL;		sb->sb_upcallarg = NULL;
sb->sb_flags &= ~SB_UPCALL;		sb->sb_flags &= ~SB_UPCALL;
}		}

		void
		solisten_upcall_set(struct socket so, so_upcall_t func, void arg)
		{

		SOLISTEN_LOCK_ASSERT(so);
		so->sol_upcall = func;
		so->sol_upcallarg = arg;
		}

		static void
		so_rdknl_lock(void *arg)
		{
		struct socket *so = arg;

		if (SOLISTENING(so))
		SOCK_LOCK(so);
		else
		SOCKBUF_LOCK(&so->so_rcv);
		}

		static void
		so_rdknl_unlock(void *arg)
		{
		struct socket *so = arg;

		if (SOLISTENING(so))
		SOCK_UNLOCK(so);
		else
		SOCKBUF_UNLOCK(&so->so_rcv);
		}

		static void
		so_rdknl_assert_locked(void *arg)
		{
		struct socket *so = arg;

		if (SOLISTENING(so))
		SOCK_LOCK_ASSERT(so);
		else
		SOCKBUF_LOCK_ASSERT(&so->so_rcv);
		}

		static void
		so_rdknl_assert_unlocked(void *arg)
		{
		struct socket *so = arg;

		if (SOLISTENING(so))
		SOCK_UNLOCK_ASSERT(so);
		else
		SOCKBUF_UNLOCK_ASSERT(&so->so_rcv);
		}

		static void
		so_wrknl_lock(void *arg)
		{
		struct socket *so = arg;

		if (SOLISTENING(so))
		SOCK_LOCK(so);
		else
		SOCKBUF_LOCK(&so->so_snd);
		}

		static void
		so_wrknl_unlock(void *arg)
		{
		struct socket *so = arg;

		if (SOLISTENING(so))
		SOCK_UNLOCK(so);
		else
		SOCKBUF_UNLOCK(&so->so_snd);
		}

		static void
		so_wrknl_assert_locked(void *arg)
		{
		struct socket *so = arg;

		if (SOLISTENING(so))
		SOCK_LOCK_ASSERT(so);
		else
		SOCKBUF_LOCK_ASSERT(&so->so_snd);
		}

		static void
		so_wrknl_assert_unlocked(void *arg)
		{
		struct socket *so = arg;

		if (SOLISTENING(so))
		SOCK_UNLOCK_ASSERT(so);
		else
		SOCKBUF_UNLOCK_ASSERT(&so->so_snd);
		}

/*		/*
* Create an external-format (``xsocket'') structure using the information in		* Create an external-format (``xsocket'') structure using the information in
* the kernel-format socket structure pointed to by so. This is done to		* the kernel-format socket structure pointed to by so. This is done to
* reduce the spew of irrelevant information over this interface, to isolate		* reduce the spew of irrelevant information over this interface, to isolate
* user code from changes in the kernel structure, and potentially to provide		* user code from changes in the kernel structure, and potentially to provide
* information-hiding if we decide that some of this information should be		* information-hiding if we decide that some of this information should be
* hidden from users.		* hidden from users.
*/		*/
void		void
sotoxsocket(struct socket so, struct xsocket xso)		sotoxsocket(struct socket so, struct xsocket xso)
{		{

xso->xso_len = sizeof *xso;		xso->xso_len = sizeof *xso;
xso->xso_so = so;		xso->xso_so = so;
xso->so_type = so->so_type;		xso->so_type = so->so_type;
xso->so_options = so->so_options;		xso->so_options = so->so_options;
xso->so_linger = so->so_linger;		xso->so_linger = so->so_linger;
xso->so_state = so->so_state;		xso->so_state = so->so_state;
xso->so_pcb = so->so_pcb;		xso->so_pcb = so->so_pcb;
xso->xso_protocol = so->so_proto->pr_protocol;		xso->xso_protocol = so->so_proto->pr_protocol;
xso->xso_family = so->so_proto->pr_domain->dom_family;		xso->xso_family = so->so_proto->pr_domain->dom_family;
xso->so_qlen = so->so_qlen;
xso->so_incqlen = so->so_incqlen;
xso->so_qlimit = so->so_qlimit;
xso->so_timeo = so->so_timeo;		xso->so_timeo = so->so_timeo;
xso->so_error = so->so_error;		xso->so_error = so->so_error;
		xso->so_uid = so->so_cred->cr_uid;
xso->so_pgid = so->so_sigio ? so->so_sigio->sio_pgid : 0;		xso->so_pgid = so->so_sigio ? so->so_sigio->sio_pgid : 0;
		if (SOLISTENING(so)) {
		xso->so_qlen = so->sol_qlen;
		xso->so_incqlen = so->sol_incqlen;
		xso->so_qlimit = so->sol_qlimit;
		xso->so_oobmark = 0;
		bzero(&xso->so_snd, sizeof(xso->so_snd));
		bzero(&xso->so_rcv, sizeof(xso->so_rcv));
		} else {
		xso->so_state \|= so->so_qstate;
		xso->so_qlen = xso->so_incqlen = xso->so_qlimit = 0;
xso->so_oobmark = so->so_oobmark;		xso->so_oobmark = so->so_oobmark;
sbtoxsockbuf(&so->so_snd, &xso->so_snd);		sbtoxsockbuf(&so->so_snd, &xso->so_snd);
sbtoxsockbuf(&so->so_rcv, &xso->so_rcv);		sbtoxsockbuf(&so->so_rcv, &xso->so_rcv);
xso->so_uid = so->so_cred->cr_uid;
}		}


/*
* Socket accessor functions to provide external consumers with
* a safe interface to socket state
*
*/

void
so_listeners_apply_all(struct socket so, void (func)(struct socket , void ),
void *arg)
{

TAILQ_FOREACH(so, &so->so_comp, so_list)
func(so, arg);
}		}

struct sockbuf *		struct sockbuf *
so_sockbuf_rcv(struct socket *so)		so_sockbuf_rcv(struct socket *so)
{		{

return (&so->so_rcv);		return (&so->so_rcv);
}		}
▲ Show 20 Lines • Show All 119 Lines • Show Last 20 Lines