/* getifaddrs -- get names and addresses of all network interfaces
Copyright (C) 2003-2023 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
The GNU C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with the GNU C Library; if not, see
. */
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include "netlinkaccess.h"
/* There is a problem with this type. The address length for
Infiniband sockets is much longer than the 8 bytes allocated in the
sockaddr_ll definition. Hence we use here a special
definition. */
struct sockaddr_ll_max
{
unsigned short int sll_family;
unsigned short int sll_protocol;
int sll_ifindex;
unsigned short int sll_hatype;
unsigned char sll_pkttype;
unsigned char sll_halen;
unsigned char sll_addr[24];
};
/* struct to hold the data for one ifaddrs entry, so we can allocate
everything at once. */
struct ifaddrs_storage
{
struct ifaddrs ifa;
union
{
/* Save space for the biggest of the four used sockaddr types and
avoid a lot of casts. */
struct sockaddr sa;
struct sockaddr_ll_max sl;
struct sockaddr_in s4;
struct sockaddr_in6 s6;
} addr, netmask, broadaddr;
char name[IF_NAMESIZE + 1];
};
void
__netlink_free_handle (struct netlink_handle *h)
{
struct netlink_res *ptr;
int saved_errno = errno;
ptr = h->nlm_list;
while (ptr != NULL)
{
struct netlink_res *tmpptr;
tmpptr = ptr->next;
free (ptr);
ptr = tmpptr;
}
__set_errno (saved_errno);
}
static int
__netlink_sendreq (struct netlink_handle *h, int type)
{
struct req
{
struct nlmsghdr nlh;
struct rtgenmsg g;
char pad[0];
} req;
struct sockaddr_nl nladdr;
if (h->seq == 0)
h->seq = time_now ();
req.nlh.nlmsg_len = sizeof (req);
req.nlh.nlmsg_type = type;
req.nlh.nlmsg_flags = NLM_F_ROOT | NLM_F_MATCH | NLM_F_REQUEST;
req.nlh.nlmsg_pid = 0;
req.nlh.nlmsg_seq = h->seq;
req.g.rtgen_family = AF_UNSPEC;
if (sizeof (req) != offsetof (struct req, pad))
memset (req.pad, '\0', sizeof (req) - offsetof (struct req, pad));
memset (&nladdr, '\0', sizeof (nladdr));
nladdr.nl_family = AF_NETLINK;
return TEMP_FAILURE_RETRY (__sendto (h->fd, (void *) &req, sizeof (req), 0,
(struct sockaddr *) &nladdr,
sizeof (nladdr)));
}
int
__netlink_request (struct netlink_handle *h, int type)
{
struct netlink_res *nlm_next;
struct sockaddr_nl nladdr;
struct nlmsghdr *nlmh;
ssize_t read_len;
bool done = false;
#ifdef PAGE_SIZE
/* Help the compiler optimize out the malloc call if PAGE_SIZE
is constant and smaller or equal to PTHREAD_STACK_MIN/4. */
const size_t buf_size = PAGE_SIZE;
#else
const size_t buf_size = __getpagesize ();
#endif
bool use_malloc = false;
char *buf;
if (__libc_use_alloca (buf_size))
buf = alloca (buf_size);
else
{
buf = malloc (buf_size);
if (buf != NULL)
use_malloc = true;
else
goto out_fail;
}
struct iovec iov = { buf, buf_size };
if (__netlink_sendreq (h, type) < 0)
goto out_fail;
while (! done)
{
struct msghdr msg =
{
.msg_name = (void *) &nladdr,
.msg_namelen = sizeof (nladdr),
.msg_iov = &iov,
.msg_iovlen = 1,
.msg_control = NULL,
.msg_controllen = 0,
.msg_flags = 0
};
read_len = TEMP_FAILURE_RETRY (__recvmsg (h->fd, &msg, 0));
__netlink_assert_response (h->fd, read_len);
if (read_len < 0)
goto out_fail;
if (nladdr.nl_pid != 0)
continue;
if (__glibc_unlikely (msg.msg_flags & MSG_TRUNC))
goto out_fail;
size_t count = 0;
size_t remaining_len = read_len;
for (nlmh = (struct nlmsghdr *) buf;
NLMSG_OK (nlmh, remaining_len);
nlmh = (struct nlmsghdr *) NLMSG_NEXT (nlmh, remaining_len))
{
if ((pid_t) nlmh->nlmsg_pid != h->pid
|| nlmh->nlmsg_seq != h->seq)
continue;
++count;
if (nlmh->nlmsg_type == NLMSG_DONE)
{
/* We found the end, leave the loop. */
done = true;
break;
}
if (nlmh->nlmsg_type == NLMSG_ERROR)
{
struct nlmsgerr *nlerr = (struct nlmsgerr *) NLMSG_DATA (nlmh);
if (nlmh->nlmsg_len < NLMSG_LENGTH (sizeof (struct nlmsgerr)))
errno = EIO;
else
errno = -nlerr->error;
goto out_fail;
}
}
/* If there was nothing with the expected nlmsg_pid and nlmsg_seq,
there is no point to record it. */
if (count == 0)
continue;
nlm_next = (struct netlink_res *) malloc (sizeof (struct netlink_res)
+ read_len);
if (nlm_next == NULL)
goto out_fail;
nlm_next->next = NULL;
nlm_next->nlh = memcpy (nlm_next + 1, buf, read_len);
nlm_next->size = read_len;
nlm_next->seq = h->seq;
if (h->nlm_list == NULL)
h->nlm_list = nlm_next;
else
h->end_ptr->next = nlm_next;
h->end_ptr = nlm_next;
}
if (use_malloc)
free (buf);
return 0;
out_fail:
if (use_malloc)
free (buf);
return -1;
}
void
__netlink_close (struct netlink_handle *h)
{
/* Don't modify errno. */
INTERNAL_SYSCALL_CALL (close, h->fd);
}
/* Open a NETLINK socket. */
int
__netlink_open (struct netlink_handle *h)
{
struct sockaddr_nl nladdr;
h->fd = __socket (PF_NETLINK, SOCK_RAW | SOCK_CLOEXEC, NETLINK_ROUTE);
if (h->fd < 0)
goto out;
memset (&nladdr, '\0', sizeof (nladdr));
nladdr.nl_family = AF_NETLINK;
if (__bind (h->fd, (struct sockaddr *) &nladdr, sizeof (nladdr)) < 0)
{
close_and_out:
__netlink_close (h);
out:
return -1;
}
/* Determine the ID the kernel assigned for this netlink connection.
It is not necessarily the PID if there is more than one socket
open. */
socklen_t addr_len = sizeof (nladdr);
if (__getsockname (h->fd, (struct sockaddr *) &nladdr, &addr_len) < 0)
goto close_and_out;
h->pid = nladdr.nl_pid;
return 0;
}
/* We know the number of RTM_NEWLINK entries, so we reserve the first
# of entries for this type. All RTM_NEWADDR entries have an index
pointer to the RTM_NEWLINK entry. To find the entry, create
a table to map kernel index entries to our index numbers.
Since we get at first all RTM_NEWLINK entries, it can never happen
that a RTM_NEWADDR index is not known to this map. */
static int
map_newlink (int index, struct ifaddrs_storage *ifas, int *map, int max)
{
int i;
for (i = 0; i < max; i++)
{
if (map[i] == -1)
{
map[i] = index;
if (i > 0)
ifas[i - 1].ifa.ifa_next = &ifas[i].ifa;
return i;
}
else if (map[i] == index)
return i;
}
/* This means interfaces changed between the reading of the
RTM_GETLINK and RTM_GETADDR information. We have to repeat
everything. */
return -1;
}
/* Create a linked list of `struct ifaddrs' structures, one for each
network interface on the host machine. If successful, store the
list in *IFAP and return 0. On errors, return -1 and set `errno'. */
static int
getifaddrs_internal (struct ifaddrs **ifap)
{
struct netlink_handle nh = { 0, 0, 0, NULL, NULL };
struct netlink_res *nlp;
struct ifaddrs_storage *ifas;
unsigned int i, newlink, newaddr, newaddr_idx;
int *map_newlink_data;
size_t ifa_data_size = 0; /* Size to allocate for all ifa_data. */
char *ifa_data_ptr; /* Pointer to the unused part of memory for
ifa_data. */
int result = 0;
*ifap = NULL;
if (__netlink_open (&nh) < 0)
return -1;
/* Tell the kernel that we wish to get a list of all
active interfaces, collect all data for every interface. */
if (__netlink_request (&nh, RTM_GETLINK) < 0)
{
result = -1;
goto exit_free;
}
/* Now ask the kernel for all addresses which are assigned
to an interface and collect all data for every interface.
Since we store the addresses after the interfaces in the
list, we will later always find the interface before the
corresponding addresses. */
++nh.seq;
if (__netlink_request (&nh, RTM_GETADDR) < 0)
{
result = -1;
goto exit_free;
}
/* Count all RTM_NEWLINK and RTM_NEWADDR entries to allocate
enough memory. */
newlink = newaddr = 0;
for (nlp = nh.nlm_list; nlp; nlp = nlp->next)
{
struct nlmsghdr *nlh;
size_t size = nlp->size;
if (nlp->nlh == NULL)
continue;
/* Walk through all entries we got from the kernel and look, which
message type they contain. */
for (nlh = nlp->nlh; NLMSG_OK (nlh, size); nlh = NLMSG_NEXT (nlh, size))
{
/* Check if the message is what we want. */
if ((pid_t) nlh->nlmsg_pid != nh.pid || nlh->nlmsg_seq != nlp->seq)
continue;
/* If the dump got interrupted, we can't rely on the results
so try again. */
if (nlh->nlmsg_flags & NLM_F_DUMP_INTR)
{
result = -EAGAIN;
goto exit_free;
}
if (nlh->nlmsg_type == NLMSG_DONE)
break; /* ok */
if (nlh->nlmsg_type == RTM_NEWLINK)
{
/* A RTM_NEWLINK message can have IFLA_STATS data. We need to
know the size before creating the list to allocate enough
memory. */
struct ifinfomsg *ifim = (struct ifinfomsg *) NLMSG_DATA (nlh);
struct rtattr *rta = IFLA_RTA (ifim);
size_t rtasize = IFLA_PAYLOAD (nlh);
while (RTA_OK (rta, rtasize))
{
size_t rta_payload = RTA_PAYLOAD (rta);
if (rta->rta_type == IFLA_STATS)
{
ifa_data_size += rta_payload;
break;
}
else
rta = RTA_NEXT (rta, rtasize);
}
++newlink;
}
else if (nlh->nlmsg_type == RTM_NEWADDR)
++newaddr;
}
}
/* Return if no interface is up. */
if ((newlink + newaddr) == 0)
goto exit_free;
/* Allocate memory for all entries we have and initialize next
pointer. */
ifas = (struct ifaddrs_storage *) calloc (1,
(newlink + newaddr)
* sizeof (struct ifaddrs_storage)
+ ifa_data_size);
if (ifas == NULL)
{
result = -1;
goto exit_free;
}
/* Table for mapping kernel index to entry in our list. */
map_newlink_data = alloca (newlink * sizeof (int));
memset (map_newlink_data, '\xff', newlink * sizeof (int));
ifa_data_ptr = (char *) &ifas[newlink + newaddr];
newaddr_idx = 0; /* Counter for newaddr index. */
/* Walk through the list of data we got from the kernel. */
for (nlp = nh.nlm_list; nlp; nlp = nlp->next)
{
struct nlmsghdr *nlh;
size_t size = nlp->size;
if (nlp->nlh == NULL)
continue;
/* Walk through one message and look at the type: If it is our
message, we need RTM_NEWLINK/RTM_NEWADDR and stop if we reach
the end or we find the end marker (in this case we ignore the
following data. */
for (nlh = nlp->nlh; NLMSG_OK (nlh, size); nlh = NLMSG_NEXT (nlh, size))
{
int ifa_index = 0;
/* Check if the message is the one we want */
if ((pid_t) nlh->nlmsg_pid != nh.pid || nlh->nlmsg_seq != nlp->seq)
continue;
if (nlh->nlmsg_type == NLMSG_DONE)
break; /* ok */
if (nlh->nlmsg_type == RTM_NEWLINK)
{
/* We found a new interface. Now extract everything from the
interface data we got and need. */
struct ifinfomsg *ifim = (struct ifinfomsg *) NLMSG_DATA (nlh);
struct rtattr *rta = IFLA_RTA (ifim);
size_t rtasize = IFLA_PAYLOAD (nlh);
/* Interfaces are stored in the first "newlink" entries
of our list, starting in the order as we got from the
kernel. */
ifa_index = map_newlink (ifim->ifi_index - 1, ifas,
map_newlink_data, newlink);
if (__glibc_unlikely (ifa_index == -1))
{
try_again:
result = -EAGAIN;
free (ifas);
goto exit_free;
}
ifas[ifa_index].ifa.ifa_flags = ifim->ifi_flags;
while (RTA_OK (rta, rtasize))
{
char *rta_data = RTA_DATA (rta);
size_t rta_payload = RTA_PAYLOAD (rta);
switch (rta->rta_type)
{
case IFLA_ADDRESS:
if (rta_payload <= sizeof (ifas[ifa_index].addr))
{
ifas[ifa_index].addr.sl.sll_family = AF_PACKET;
memcpy (ifas[ifa_index].addr.sl.sll_addr,
(char *) rta_data, rta_payload);
ifas[ifa_index].addr.sl.sll_halen = rta_payload;
ifas[ifa_index].addr.sl.sll_ifindex
= ifim->ifi_index;
ifas[ifa_index].addr.sl.sll_hatype = ifim->ifi_type;
ifas[ifa_index].ifa.ifa_addr
= &ifas[ifa_index].addr.sa;
}
break;
case IFLA_BROADCAST:
if (rta_payload <= sizeof (ifas[ifa_index].broadaddr))
{
ifas[ifa_index].broadaddr.sl.sll_family = AF_PACKET;
memcpy (ifas[ifa_index].broadaddr.sl.sll_addr,
(char *) rta_data, rta_payload);
ifas[ifa_index].broadaddr.sl.sll_halen = rta_payload;
ifas[ifa_index].broadaddr.sl.sll_ifindex
= ifim->ifi_index;
ifas[ifa_index].broadaddr.sl.sll_hatype
= ifim->ifi_type;
ifas[ifa_index].ifa.ifa_broadaddr
= &ifas[ifa_index].broadaddr.sa;
}
break;
case IFLA_IFNAME: /* Name of Interface */
if ((rta_payload + 1) <= sizeof (ifas[ifa_index].name))
{
ifas[ifa_index].ifa.ifa_name = ifas[ifa_index].name;
*(char *) __mempcpy (ifas[ifa_index].name, rta_data,
rta_payload) = '\0';
}
break;
case IFLA_STATS: /* Statistics of Interface */
ifas[ifa_index].ifa.ifa_data = ifa_data_ptr;
ifa_data_ptr += rta_payload;
memcpy (ifas[ifa_index].ifa.ifa_data, rta_data,
rta_payload);
break;
case IFLA_UNSPEC:
break;
case IFLA_MTU:
break;
case IFLA_LINK:
break;
case IFLA_QDISC:
break;
default:
break;
}
rta = RTA_NEXT (rta, rtasize);
}
}
else if (nlh->nlmsg_type == RTM_NEWADDR)
{
struct ifaddrmsg *ifam = (struct ifaddrmsg *) NLMSG_DATA (nlh);
struct rtattr *rta = IFA_RTA (ifam);
size_t rtasize = IFA_PAYLOAD (nlh);
/* New Addresses are stored in the order we got them from
the kernel after the interfaces. Theoretically it is possible
that we have holes in the interface part of the list,
but we always have already the interface for this address. */
ifa_index = newlink + newaddr_idx;
int idx = map_newlink (ifam->ifa_index - 1, ifas,
map_newlink_data, newlink);
if (__glibc_unlikely (idx == -1))
goto try_again;
ifas[ifa_index].ifa.ifa_flags = ifas[idx].ifa.ifa_flags;
if (ifa_index > 0)
ifas[ifa_index - 1].ifa.ifa_next = &ifas[ifa_index].ifa;
++newaddr_idx;
while (RTA_OK (rta, rtasize))
{
char *rta_data = RTA_DATA (rta);
size_t rta_payload = RTA_PAYLOAD (rta);
switch (rta->rta_type)
{
case IFA_ADDRESS:
{
struct sockaddr *sa;
if (ifas[ifa_index].ifa.ifa_addr != NULL)
{
/* In a point-to-poing network IFA_ADDRESS
contains the destination address, local
address is supplied in IFA_LOCAL attribute.
destination address and broadcast address
are stored in an union, so it doesn't matter
which name we use. */
ifas[ifa_index].ifa.ifa_broadaddr
= &ifas[ifa_index].broadaddr.sa;
sa = &ifas[ifa_index].broadaddr.sa;
}
else
{
ifas[ifa_index].ifa.ifa_addr
= &ifas[ifa_index].addr.sa;
sa = &ifas[ifa_index].addr.sa;
}
sa->sa_family = ifam->ifa_family;
switch (ifam->ifa_family)
{
case AF_INET:
/* Size must match that of an address for IPv4. */
if (rta_payload == 4)
memcpy (&((struct sockaddr_in *) sa)->sin_addr,
rta_data, rta_payload);
break;
case AF_INET6:
/* Size must match that of an address for IPv6. */
if (rta_payload == 16)
{
memcpy (&((struct sockaddr_in6 *) sa)->sin6_addr,
rta_data, rta_payload);
if (IN6_IS_ADDR_LINKLOCAL (rta_data)
|| IN6_IS_ADDR_MC_LINKLOCAL (rta_data))
((struct sockaddr_in6 *) sa)->sin6_scope_id
= ifam->ifa_index;
}
break;
default:
if (rta_payload <= sizeof (ifas[ifa_index].addr))
memcpy (sa->sa_data, rta_data, rta_payload);
break;
}
}
break;
case IFA_LOCAL:
if (ifas[ifa_index].ifa.ifa_addr != NULL)
{
/* If ifa_addr is set and we get IFA_LOCAL,
assume we have a point-to-point network.
Move address to correct field. */
ifas[ifa_index].broadaddr = ifas[ifa_index].addr;
ifas[ifa_index].ifa.ifa_broadaddr
= &ifas[ifa_index].broadaddr.sa;
memset (&ifas[ifa_index].addr, '\0',
sizeof (ifas[ifa_index].addr));
}
ifas[ifa_index].ifa.ifa_addr = &ifas[ifa_index].addr.sa;
ifas[ifa_index].ifa.ifa_addr->sa_family
= ifam->ifa_family;
switch (ifam->ifa_family)
{
case AF_INET:
/* Size must match that of an address for IPv4. */
if (rta_payload == 4)
memcpy (&ifas[ifa_index].addr.s4.sin_addr,
rta_data, rta_payload);
break;
case AF_INET6:
/* Size must match that of an address for IPv6. */
if (rta_payload == 16)
{
memcpy (&ifas[ifa_index].addr.s6.sin6_addr,
rta_data, rta_payload);
if (IN6_IS_ADDR_LINKLOCAL (rta_data)
|| IN6_IS_ADDR_MC_LINKLOCAL (rta_data))
ifas[ifa_index].addr.s6.sin6_scope_id =
ifam->ifa_index;
}
break;
default:
if (rta_payload <= sizeof (ifas[ifa_index].addr))
memcpy (ifas[ifa_index].addr.sa.sa_data,
rta_data, rta_payload);
break;
}
break;
case IFA_BROADCAST:
/* We get IFA_BROADCAST, so IFA_LOCAL was too much. */
if (ifas[ifa_index].ifa.ifa_broadaddr != NULL)
memset (&ifas[ifa_index].broadaddr, '\0',
sizeof (ifas[ifa_index].broadaddr));
ifas[ifa_index].ifa.ifa_broadaddr
= &ifas[ifa_index].broadaddr.sa;
ifas[ifa_index].ifa.ifa_broadaddr->sa_family
= ifam->ifa_family;
switch (ifam->ifa_family)
{
case AF_INET:
/* Size must match that of an address for IPv4. */
if (rta_payload == 4)
memcpy (&ifas[ifa_index].broadaddr.s4.sin_addr,
rta_data, rta_payload);
break;
case AF_INET6:
/* Size must match that of an address for IPv6. */
if (rta_payload == 16)
{
memcpy (&ifas[ifa_index].broadaddr.s6.sin6_addr,
rta_data, rta_payload);
if (IN6_IS_ADDR_LINKLOCAL (rta_data)
|| IN6_IS_ADDR_MC_LINKLOCAL (rta_data))
ifas[ifa_index].broadaddr.s6.sin6_scope_id
= ifam->ifa_index;
}
break;
default:
if (rta_payload <= sizeof (ifas[ifa_index].addr))
memcpy (&ifas[ifa_index].broadaddr.sa.sa_data,
rta_data, rta_payload);
break;
}
break;
case IFA_LABEL:
if (rta_payload + 1 <= sizeof (ifas[ifa_index].name))
{
ifas[ifa_index].ifa.ifa_name = ifas[ifa_index].name;
*(char *) __mempcpy (ifas[ifa_index].name, rta_data,
rta_payload) = '\0';
}
else
abort ();
break;
case IFA_UNSPEC:
break;
case IFA_CACHEINFO:
break;
default:
break;
}
rta = RTA_NEXT (rta, rtasize);
}
/* If we didn't get the interface name with the
address, use the name from the interface entry. */
if (ifas[ifa_index].ifa.ifa_name == NULL)
{
int idx = map_newlink (ifam->ifa_index - 1, ifas,
map_newlink_data, newlink);
if (__glibc_unlikely (idx == -1))
goto try_again;
ifas[ifa_index].ifa.ifa_name = ifas[idx].ifa.ifa_name;
}
/* Calculate the netmask. */
if (ifas[ifa_index].ifa.ifa_addr
&& ifas[ifa_index].ifa.ifa_addr->sa_family != AF_UNSPEC
&& ifas[ifa_index].ifa.ifa_addr->sa_family != AF_PACKET)
{
uint32_t max_prefixlen = 0;
char *cp = NULL;
ifas[ifa_index].ifa.ifa_netmask
= &ifas[ifa_index].netmask.sa;
switch (ifas[ifa_index].ifa.ifa_addr->sa_family)
{
case AF_INET:
cp = (char *) &ifas[ifa_index].netmask.s4.sin_addr;
max_prefixlen = 32;
break;
case AF_INET6:
cp = (char *) &ifas[ifa_index].netmask.s6.sin6_addr;
max_prefixlen = 128;
break;
}
ifas[ifa_index].ifa.ifa_netmask->sa_family
= ifas[ifa_index].ifa.ifa_addr->sa_family;
if (cp != NULL)
{
unsigned int preflen;
if (ifam->ifa_prefixlen > max_prefixlen)
preflen = max_prefixlen;
else
preflen = ifam->ifa_prefixlen;
for (i = 0; i < preflen / 8; i++)
*cp++ = 0xff;
if (preflen % 8)
*cp = 0xff << (8 - preflen % 8);
}
}
}
}
}
assert (ifa_data_ptr <= (char *) &ifas[newlink + newaddr] + ifa_data_size);
if (newaddr_idx > 0)
{
for (i = 0; i < newlink; ++i)
if (map_newlink_data[i] == -1)
{
/* We have fewer links then we anticipated. Adjust the
forward pointer to the first address entry. */
ifas[i - 1].ifa.ifa_next = &ifas[newlink].ifa;
}
if (i == 0 && newlink > 0)
/* No valid link, but we allocated memory. We have to
populate the first entry. */
memmove (ifas, &ifas[newlink], sizeof (struct ifaddrs_storage));
}
*ifap = &ifas[0].ifa;
exit_free:
__netlink_free_handle (&nh);
__netlink_close (&nh);
return result;
}
/* Create a linked list of `struct ifaddrs' structures, one for each
network interface on the host machine. If successful, store the
list in *IFAP and return 0. On errors, return -1 and set `errno'. */
int
__getifaddrs (struct ifaddrs **ifap)
{
int res;
do
res = getifaddrs_internal (ifap);
while (res == -EAGAIN);
return res;
}
weak_alias (__getifaddrs, getifaddrs)
libc_hidden_def (__getifaddrs)
libc_hidden_weak (getifaddrs)
void
__freeifaddrs (struct ifaddrs *ifa)
{
free (ifa);
}
weak_alias (__freeifaddrs, freeifaddrs)
libc_hidden_def (__freeifaddrs)
libc_hidden_weak (freeifaddrs)