diff options
author | zec <zec@FreeBSD.org> | 2009-05-05 10:56:12 +0000 |
---|---|---|
committer | zec <zec@FreeBSD.org> | 2009-05-05 10:56:12 +0000 |
commit | d78a1b1a824c4f5eb8cb3583bb5265f73dcc24dd (patch) | |
tree | 79a0bccccf2c92504cdf23ad15f7c1813bb3f926 /sys/net/if.c | |
parent | 8e4ffe653f6c9ff6da3eed58566ef35e77d530d0 (diff) | |
download | FreeBSD-src-d78a1b1a824c4f5eb8cb3583bb5265f73dcc24dd.zip FreeBSD-src-d78a1b1a824c4f5eb8cb3583bb5265f73dcc24dd.tar.gz |
Change the curvnet variable from a global const struct vnet *,
previously always pointing to the default vnet context, to a
dynamically changing thread-local one. The currvnet context
should be set on entry to networking code via CURVNET_SET() macros,
and reverted to previous state via CURVNET_RESTORE(). Recursions
on curvnet are permitted, though strongly discuouraged.
This change should have no functional impact on nooptions VIMAGE
kernel builds, where CURVNET_* macros expand to whitespace.
The curthread->td_vnet (aka curvnet) variable's purpose is to be an
indicator of the vnet context in which the current network-related
operation takes place, in case we cannot deduce the current vnet
context from any other source, such as by looking at mbuf's
m->m_pkthdr.rcvif->if_vnet, sockets's so->so_vnet etc. Moreover, so
far curvnet has turned out to be an invaluable consistency checking
aid: it helps to catch cases when sockets, ifnets or any other
vnet-aware structures may have leaked from one vnet to another.
The exact placement of the CURVNET_SET() / CURVNET_RESTORE() macros
was a result of an empirical iterative process, whith an aim to
reduce recursions on CURVNET_SET() to a minimum, while still reducing
the scope of CURVNET_SET() to networking only operations - the
alternative would be calling CURVNET_SET() on each system call entry.
In general, curvnet has to be set in three typicall cases: when
processing socket-related requests from userspace or from within the
kernel; when processing inbound traffic flowing from device drivers
to upper layers of the networking stack, and when executing
timer-driven networking functions.
This change also introduces a DDB subcommand to show the list of all
vnet instances.
Approved by: julian (mentor)
Diffstat (limited to 'sys/net/if.c')
-rw-r--r-- | sys/net/if.c | 42 |
1 files changed, 28 insertions, 14 deletions
diff --git a/sys/net/if.c b/sys/net/if.c index 92bf8a6..a67f31b 100644 --- a/sys/net/if.c +++ b/sys/net/if.c @@ -53,6 +53,7 @@ #include <sys/kernel.h> #include <sys/lock.h> #include <sys/refcount.h> +#include <sys/module.h> #include <sys/rwlock.h> #include <sys/sockio.h> #include <sys/syslog.h> @@ -126,7 +127,6 @@ static void if_attachdomain(void *); static void if_attachdomain1(struct ifnet *); static int ifconf(u_long, caddr_t); static void if_freemulti(struct ifmultiaddr *); -static void if_grow(void); static void if_init(void *); static void if_check(void *); static void if_route(struct ifnet *, int flag, int fam); @@ -202,7 +202,7 @@ MALLOC_DEFINE(M_IFNET, "ifnet", "interface internals"); MALLOC_DEFINE(M_IFADDR, "ifaddr", "interface address"); MALLOC_DEFINE(M_IFMADDR, "ether_multi", "link-level multicast address"); -static struct ifnet * +struct ifnet * ifnet_byindex_locked(u_short idx) { INIT_VNET_NET(curvnet); @@ -239,7 +239,7 @@ ifnet_byindex_ref(u_short idx) return (ifp); } -static void +void ifnet_setbyindex(u_short idx, struct ifnet *ifp) { INIT_VNET_NET(curvnet); @@ -445,7 +445,7 @@ vnet_net_iattach(const void *unused __unused) return (0); } -static void +void if_grow(void) { INIT_VNET_NET(curvnet); @@ -696,11 +696,13 @@ if_attach(struct ifnet *ifp) mac_ifnet_create(ifp); #endif - ifdev_setbyindex(ifp->if_index, make_dev(&net_cdevsw, - ifp->if_index, UID_ROOT, GID_WHEEL, 0600, "%s/%s", - net_cdevsw.d_name, ifp->if_xname)); - make_dev_alias(ifdev_byindex(ifp->if_index), "%s%d", - net_cdevsw.d_name, ifp->if_index); + if (IS_DEFAULT_VNET(curvnet)) { + ifdev_setbyindex(ifp->if_index, make_dev(&net_cdevsw, + ifp->if_index, UID_ROOT, GID_WHEEL, 0600, "%s/%s", + net_cdevsw.d_name, ifp->if_xname)); + make_dev_alias(ifdev_byindex(ifp->if_index), "%s%d", + net_cdevsw.d_name, ifp->if_index); + } ifq_attach(&ifp->if_snd, ifp); @@ -742,13 +744,17 @@ if_attach(struct ifnet *ifp) IFNET_WLOCK(); TAILQ_INSERT_TAIL(&V_ifnet, ifp, if_link); +#ifdef VIMAGE + curvnet->ifccnt++; +#endif IFNET_WUNLOCK(); if (domain_init_status >= 2) if_attachdomain1(ifp); EVENTHANDLER_INVOKE(ifnet_arrival_event, ifp); - devctl_notify("IFNET", ifp->if_xname, "ATTACH", NULL); + if (IS_DEFAULT_VNET(curvnet)) + devctl_notify("IFNET", ifp->if_xname, "ATTACH", NULL); /* Announce the interface. */ rt_ifannouncemsg(ifp, IFAN_ARRIVAL); @@ -895,6 +901,10 @@ if_detach(struct ifnet *ifp) found = 1; break; } +#ifdef VIMAGE + if (found) + curvnet->ifccnt--; +#endif IFNET_WUNLOCK(); if (!found) return; @@ -943,7 +953,8 @@ if_detach(struct ifnet *ifp) * Clean up all addresses. */ ifp->if_addr = NULL; - destroy_dev(ifdev_byindex(ifp->if_index)); + if (IS_DEFAULT_VNET(curvnet)) + destroy_dev(ifdev_byindex(ifp->if_index)); ifdev_setbyindex(ifp->if_index, NULL); /* We can now free link ifaddr. */ @@ -972,7 +983,8 @@ if_detach(struct ifnet *ifp) /* Announce that the interface is gone. */ rt_ifannouncemsg(ifp, IFAN_DEPARTURE); EVENTHANDLER_INVOKE(ifnet_departure_event, ifp); - devctl_notify("IFNET", ifp->if_xname, "DETACH", NULL); + if (IS_DEFAULT_VNET(curvnet)) + devctl_notify("IFNET", ifp->if_xname, "DETACH", NULL); if_delgroups(ifp); IF_AFDATA_LOCK(ifp); @@ -1701,8 +1713,10 @@ do_link_state_change(void *arg, int pending) (*lagg_linkstate_p)(ifp, link_state); } - devctl_notify("IFNET", ifp->if_xname, - (link_state == LINK_STATE_UP) ? "LINK_UP" : "LINK_DOWN", NULL); + if (IS_DEFAULT_VNET(curvnet)) + devctl_notify("IFNET", ifp->if_xname, + (link_state == LINK_STATE_UP) ? "LINK_UP" : "LINK_DOWN", + NULL); if (pending > 1) if_printf(ifp, "%d link states coalesced\n", pending); if (log_link_state_change) |