libvirt/src/util/virnetdev.c

2899 lines
76 KiB
C
Raw Normal View History

/*
* Copyright (C) 2007-2015 Red Hat, Inc.
*
* This 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.
*
* This 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 this library. If not, see
* <http://www.gnu.org/licenses/>.
*
* Authors:
* Mark McLoughlin <markmc@redhat.com>
* Daniel P. Berrange <berrange@redhat.com>
*/
#include <config.h>
#include "virnetdev.h"
#include "virmacaddr.h"
#include "virfile.h"
#include "virerror.h"
#include "vircommand.h"
2012-12-12 18:06:53 +00:00
#include "viralloc.h"
2012-12-13 14:52:25 +00:00
#include "virpci.h"
2012-12-12 17:59:27 +00:00
#include "virlog.h"
#include "virstring.h"
#include "virutil.h"
#include <sys/ioctl.h>
#include <net/if.h>
#include <fcntl.h>
#ifdef __linux__
# include <linux/sockios.h>
# include <linux/if_vlan.h>
# define VIR_NETDEV_FAMILY AF_PACKET
#elif defined(HAVE_STRUCT_IFREQ) && defined(AF_LOCAL)
# define VIR_NETDEV_FAMILY AF_LOCAL
#else
# undef HAVE_STRUCT_IFREQ
#endif
#if defined(SIOCETHTOOL) && defined(HAVE_STRUCT_IFREQ)
# include <linux/types.h>
# include <linux/ethtool.h>
#endif
#if HAVE_DECL_LINK_ADDR
# include <sys/sockio.h>
# include <net/if_dl.h>
#endif
#ifndef IFNAMSIZ
# define IFNAMSIZ 16
#endif
#define VIR_FROM_THIS VIR_FROM_NONE
VIR_LOG_INIT("util.netdev");
#define PROC_NET_DEV_MCAST "/proc/net/dev_mcast"
#define MAX_MCAST_SIZE 50*14336
#define VIR_MCAST_NAME_LEN (IFNAMSIZ + 1)
#define VIR_MCAST_TOKEN_DELIMS " \n"
#define VIR_MCAST_ADDR_LEN (VIR_MAC_HEXLEN + 1)
#if defined(SIOCSIFFLAGS) && defined(HAVE_STRUCT_IFREQ)
# define VIR_IFF_UP IFF_UP
# define VIR_IFF_PROMISC IFF_PROMISC
# define VIR_IFF_MULTICAST IFF_MULTICAST
# define VIR_IFF_ALLMULTI IFF_ALLMULTI
#else
# define VIR_IFF_UP 0
# define VIR_IFF_PROMISC 0
# define VIR_IFF_MULTICAST 0
# define VIR_IFF_ALLMULTI 0
#endif
typedef enum {
VIR_MCAST_TYPE_INDEX_TOKEN,
VIR_MCAST_TYPE_NAME_TOKEN,
VIR_MCAST_TYPE_USERS_TOKEN,
VIR_MCAST_TYPE_GLOBAL_TOKEN,
VIR_MCAST_TYPE_ADDR_TOKEN,
VIR_MCAST_TYPE_LAST
} virMCastType;
typedef struct _virNetDevMcastEntry virNetDevMcastEntry;
typedef virNetDevMcastEntry *virNetDevMcastEntryPtr;
struct _virNetDevMcastEntry {
int index;
char name[VIR_MCAST_NAME_LEN];
int users;
bool global;
virMacAddr macaddr;
};
typedef struct _virNetDevMcastList virNetDevMcastList;
typedef virNetDevMcastList *virNetDevMcastListPtr;
struct _virNetDevMcastList {
size_t nentries;
virNetDevMcastEntryPtr *entries;
};
#if defined(HAVE_STRUCT_IFREQ)
static int virNetDevSetupControlFull(const char *ifname,
struct ifreq *ifr,
int domain,
int type)
{
int fd;
if (ifr && ifname) {
memset(ifr, 0, sizeof(*ifr));
if (virStrcpyStatic(ifr->ifr_name, ifname) == NULL) {
virReportSystemError(ERANGE,
_("Network interface name '%s' is too long"),
ifname);
return -1;
}
}
if ((fd = socket(domain, type, 0)) < 0) {
virReportSystemError(errno, "%s",
_("Cannot open network interface control socket"));
return -1;
}
if (virSetInherit(fd, false) < 0) {
virReportSystemError(errno, "%s",
_("Cannot set close-on-exec flag for socket"));
VIR_FORCE_CLOSE(fd);
return -1;
}
return fd;
}
int
virNetDevSetupControl(const char *ifname,
struct ifreq *ifr)
{
return virNetDevSetupControlFull(ifname, ifr, VIR_NETDEV_FAMILY, SOCK_DGRAM);
}
#else /* !HAVE_STRUCT_IFREQ */
int
virNetDevSetupControl(const char *ifname ATTRIBUTE_UNUSED,
void *ifr ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
_("Network device configuration is not supported "
"on this platform"));
return -1;
}
#endif /* HAVE_STRUCT_IFREQ */
#if defined(SIOCGIFFLAGS) && defined(HAVE_STRUCT_IFREQ)
/**
* virNetDevExists:
* @ifname
*
* Check if the network device @ifname exists
*
* Returns 1 if it exists, 0 if it does not, -1 on error
*/
int virNetDevExists(const char *ifname)
{
int fd = -1;
int ret = -1;
struct ifreq ifr;
if ((fd = virNetDevSetupControl(ifname, &ifr)) < 0)
return -1;
if (ioctl(fd, SIOCGIFFLAGS, &ifr)) {
if (errno == ENODEV || errno == ENXIO)
ret = 0;
else
virReportSystemError(errno,
_("Unable to check interface flags for %s"), ifname);
goto cleanup;
}
ret = 1;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#else
int virNetDevExists(const char *ifname)
{
virReportSystemError(ENOSYS,
_("Unable to check interface %s"), ifname);
return -1;
}
#endif
#if defined(SIOCGIFHWADDR) && defined(SIOCSIFHWADDR) && \
defined(HAVE_STRUCT_IFREQ)
/**
* virNetDevSetMAC:
* @ifname: interface name to set MTU for
* @macaddr: MAC address
*
* This function sets the @macaddr for a given interface @ifname. This
* gets rid of the kernel's automatically assigned random MAC.
*
* Returns 0 in case of success or -1 on failure
*/
int virNetDevSetMAC(const char *ifname,
maint: avoid 'const fooPtr' in virnet files 'const fooPtr' is the same as 'foo * const' (the pointer won't change, but it's contents can). But in general, if an interface is trying to be const-correct, it should be using 'const foo *' (the pointer is to data that can't be changed). Fix up remaining offenders in src/util. * src/util/virnetdev.h (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Use intended type. * src/util/virnetdevbandwidth.h (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.h (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.h (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.h (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.h (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.h (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate): Likewise. * src/util/virnetlink.h (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. * src/util/virnetdev.c (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Fix fallout. * src/util/virnetdevbandwidth.c (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.c (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.c (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.c (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.c (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.c (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate) (virNetDevVPortProfileOpSetLink, virNetDevVPortProfileOpCommon) (virNetDevVPortProfileOp8021Qbg, virNetDevVPortProfileOp8021Qbh): Likewise. * src/util/virnetlink.c (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. Signed-off-by: Eric Blake <eblake@redhat.com>
2013-10-05 19:41:44 +00:00
const virMacAddr *macaddr)
{
int fd = -1;
int ret = -1;
struct ifreq ifr;
if ((fd = virNetDevSetupControl(ifname, &ifr)) < 0)
return -1;
/* To fill ifr.ifr_hdaddr.sa_family field */
if (ioctl(fd, SIOCGIFHWADDR, &ifr) < 0) {
virReportSystemError(errno,
_("Cannot get interface MAC on '%s'"),
ifname);
goto cleanup;
}
virMacAddrGetRaw(macaddr, (unsigned char *)ifr.ifr_hwaddr.sa_data);
if (ioctl(fd, SIOCSIFHWADDR, &ifr) < 0) {
virReportSystemError(errno,
_("Cannot set interface MAC on '%s'"),
ifname);
goto cleanup;
}
ret = 0;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#elif defined(SIOCSIFLLADDR) && defined(HAVE_STRUCT_IFREQ) && \
HAVE_DECL_LINK_ADDR
int virNetDevSetMAC(const char *ifname,
maint: avoid 'const fooPtr' in virnet files 'const fooPtr' is the same as 'foo * const' (the pointer won't change, but it's contents can). But in general, if an interface is trying to be const-correct, it should be using 'const foo *' (the pointer is to data that can't be changed). Fix up remaining offenders in src/util. * src/util/virnetdev.h (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Use intended type. * src/util/virnetdevbandwidth.h (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.h (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.h (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.h (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.h (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.h (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate): Likewise. * src/util/virnetlink.h (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. * src/util/virnetdev.c (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Fix fallout. * src/util/virnetdevbandwidth.c (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.c (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.c (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.c (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.c (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.c (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate) (virNetDevVPortProfileOpSetLink, virNetDevVPortProfileOpCommon) (virNetDevVPortProfileOp8021Qbg, virNetDevVPortProfileOp8021Qbh): Likewise. * src/util/virnetlink.c (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. Signed-off-by: Eric Blake <eblake@redhat.com>
2013-10-05 19:41:44 +00:00
const virMacAddr *macaddr)
{
struct ifreq ifr;
struct sockaddr_dl sdl;
char mac[VIR_MAC_STRING_BUFLEN + 1] = ":";
int s;
int ret = -1;
if ((s = virNetDevSetupControl(ifname, &ifr)) < 0)
return -1;
virMacAddrFormat(macaddr, mac + 1);
sdl.sdl_len = sizeof(sdl);
link_addr(mac, &sdl);
memcpy(ifr.ifr_addr.sa_data, sdl.sdl_data, VIR_MAC_BUFLEN);
ifr.ifr_addr.sa_len = VIR_MAC_BUFLEN;
if (ioctl(s, SIOCSIFLLADDR, &ifr) < 0) {
virReportSystemError(errno,
_("Cannot set interface MAC on '%s'"),
ifname);
goto cleanup;
}
ret = 0;
cleanup:
VIR_FORCE_CLOSE(s);
return ret;
}
#else
int virNetDevSetMAC(const char *ifname,
maint: avoid 'const fooPtr' in virnet files 'const fooPtr' is the same as 'foo * const' (the pointer won't change, but it's contents can). But in general, if an interface is trying to be const-correct, it should be using 'const foo *' (the pointer is to data that can't be changed). Fix up remaining offenders in src/util. * src/util/virnetdev.h (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Use intended type. * src/util/virnetdevbandwidth.h (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.h (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.h (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.h (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.h (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.h (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate): Likewise. * src/util/virnetlink.h (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. * src/util/virnetdev.c (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Fix fallout. * src/util/virnetdevbandwidth.c (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.c (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.c (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.c (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.c (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.c (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate) (virNetDevVPortProfileOpSetLink, virNetDevVPortProfileOpCommon) (virNetDevVPortProfileOp8021Qbg, virNetDevVPortProfileOp8021Qbh): Likewise. * src/util/virnetlink.c (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. Signed-off-by: Eric Blake <eblake@redhat.com>
2013-10-05 19:41:44 +00:00
const virMacAddr *macaddr ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS,
_("Cannot set interface MAC on '%s'"),
ifname);
return -1;
}
#endif
#if defined(SIOCGIFHWADDR) && defined(HAVE_STRUCT_IFREQ)
/**
* virNetDevGetMAC:
* @ifname: interface name to set MTU for
* @macaddr: MAC address
*
* This function gets the @macaddr for a given interface @ifname.
*
* Returns 0 in case of success or -1 on failure
*/
int virNetDevGetMAC(const char *ifname,
virMacAddrPtr macaddr)
{
int fd = -1;
int ret = -1;
struct ifreq ifr;
if ((fd = virNetDevSetupControl(ifname, &ifr)) < 0)
return -1;
if (ioctl(fd, SIOCGIFHWADDR, &ifr) < 0) {
virReportSystemError(errno,
_("Cannot get interface MAC on '%s'"),
ifname);
goto cleanup;
}
virMacAddrSetRaw(macaddr, (unsigned char *)ifr.ifr_hwaddr.sa_data);
ret = 0;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#else
int virNetDevGetMAC(const char *ifname,
virMacAddrPtr macaddr ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS,
_("Cannot get interface MAC on '%s'"),
ifname);
return -1;
}
#endif
/**
* virNetDevReplaceMacAddress:
* @macaddress: new MAC address for interface
* @linkdev: name of interface
* @stateDir: directory to store old MAC address
*
* Returns 0 on success, -1 on failure
*
*/
int
virNetDevReplaceMacAddress(const char *linkdev,
maint: avoid 'const fooPtr' in virnet files 'const fooPtr' is the same as 'foo * const' (the pointer won't change, but it's contents can). But in general, if an interface is trying to be const-correct, it should be using 'const foo *' (the pointer is to data that can't be changed). Fix up remaining offenders in src/util. * src/util/virnetdev.h (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Use intended type. * src/util/virnetdevbandwidth.h (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.h (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.h (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.h (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.h (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.h (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate): Likewise. * src/util/virnetlink.h (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. * src/util/virnetdev.c (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Fix fallout. * src/util/virnetdevbandwidth.c (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.c (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.c (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.c (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.c (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.c (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate) (virNetDevVPortProfileOpSetLink, virNetDevVPortProfileOpCommon) (virNetDevVPortProfileOp8021Qbg, virNetDevVPortProfileOp8021Qbh): Likewise. * src/util/virnetlink.c (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. Signed-off-by: Eric Blake <eblake@redhat.com>
2013-10-05 19:41:44 +00:00
const virMacAddr *macaddress,
const char *stateDir)
{
virMacAddr oldmac;
char *path = NULL;
char macstr[VIR_MAC_STRING_BUFLEN];
int ret = -1;
if (virNetDevGetMAC(linkdev, &oldmac) < 0)
return -1;
if (virAsprintf(&path, "%s/%s",
stateDir,
linkdev) < 0)
return -1;
virMacAddrFormat(&oldmac, macstr);
if (virFileWriteStr(path, macstr, O_CREAT|O_TRUNC|O_WRONLY) < 0) {
virReportSystemError(errno, _("Unable to preserve mac for %s"),
linkdev);
goto cleanup;
}
if (virNetDevSetMAC(linkdev, macaddress) < 0)
goto cleanup;
ret = 0;
cleanup:
VIR_FREE(path);
return ret;
}
/**
* virNetDevRestoreMacAddress:
* @linkdev: name of interface
* @stateDir: directory containing old MAC address
*
* Returns 0 on success, -errno on failure.
*
*/
int
virNetDevRestoreMacAddress(const char *linkdev,
const char *stateDir)
{
int rc = -1;
char *oldmacname = NULL;
char *macstr = NULL;
char *path = NULL;
virMacAddr oldmac;
if (virAsprintf(&path, "%s/%s",
stateDir,
linkdev) < 0)
return -1;
if (virFileReadAll(path, VIR_MAC_STRING_BUFLEN, &macstr) < 0)
goto cleanup;
if (virMacAddrParse(macstr, &oldmac) != 0) {
virReportError(VIR_ERR_INTERNAL_ERROR,
_("Cannot parse MAC address from '%s'"),
oldmacname);
goto cleanup;
}
/*reset mac and remove file-ignore results*/
rc = virNetDevSetMAC(linkdev, &oldmac);
ignore_value(unlink(path));
cleanup:
VIR_FREE(macstr);
VIR_FREE(path);
return rc;
}
#if defined(SIOCGIFMTU) && defined(HAVE_STRUCT_IFREQ)
/**
* virNetDevGetMTU:
* @ifname: interface name get MTU for
*
* This function gets the @mtu value set for a given interface @ifname.
*
* Returns the MTU value in case of success, or -1 on failure.
*/
int virNetDevGetMTU(const char *ifname)
{
int fd = -1;
int ret = -1;
struct ifreq ifr;
if ((fd = virNetDevSetupControl(ifname, &ifr)) < 0)
return -1;
if (ioctl(fd, SIOCGIFMTU, &ifr) < 0) {
virReportSystemError(errno,
_("Cannot get interface MTU on '%s'"),
ifname);
goto cleanup;
}
ret = ifr.ifr_mtu;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#else
int virNetDevGetMTU(const char *ifname)
{
virReportSystemError(ENOSYS,
_("Cannot get interface MTU on '%s'"),
ifname);
return -1;
}
#endif
#if defined(SIOCSIFMTU) && defined(HAVE_STRUCT_IFREQ)
/**
* virNetDevSetMTU:
* @ifname: interface name to set MTU for
* @mtu: MTU value
*
* This function sets the @mtu for a given interface @ifname. Typically
* used on a tap device to set up for Jumbo Frames.
*
* Returns 0 in case of success, or -1 on failure
*/
int virNetDevSetMTU(const char *ifname, int mtu)
{
int fd = -1;
int ret = -1;
struct ifreq ifr;
if ((fd = virNetDevSetupControl(ifname, &ifr)) < 0)
return -1;
ifr.ifr_mtu = mtu;
if (ioctl(fd, SIOCSIFMTU, &ifr) < 0) {
virReportSystemError(errno,
_("Cannot set interface MTU on '%s'"),
ifname);
goto cleanup;
}
ret = 0;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#else
int virNetDevSetMTU(const char *ifname, int mtu ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS,
_("Cannot set interface MTU on '%s'"),
ifname);
return -1;
}
#endif
/**
* virNetDevSetMTUFromDevice:
* @ifname: name of the interface whose MTU we want to set
* @otherifname: name of the interface whose MTU we want to copy
*
* Sets the interface mtu to the same MTU as another interface
*
* Returns 0 in case of success, or -1 on failure
*/
int virNetDevSetMTUFromDevice(const char *ifname,
const char *otherifname)
{
int mtu = virNetDevGetMTU(otherifname);
if (mtu < 0)
return -1;
return virNetDevSetMTU(ifname, mtu);
}
/**
* virNetDevSetNamespace:
* @ifname: name of device
* @pidInNs: PID of process in target net namespace
*
* Moves the given device into the target net namespace specified by the given
* pid using this command:
* ip link set @iface netns @pidInNs
*
* Returns 0 on success or -1 in case of error
*/
build: use correct type for pid and similar types No thanks to 64-bit windows, with 64-bit pid_t, we have to avoid constructs like 'int pid'. Our API in libvirt-qemu cannot be changed without breaking ABI; but then again, libvirt-qemu can only be used on systems that support UNIX sockets, which rules out Windows (even if qemu could be compiled there) - so for all points on the call chain that interact with this API decision, we require a different variable name to make it clear that we audited the use for safety. Adding a syntax-check rule only solves half the battle; anywhere that uses printf on a pid_t still needs to be converted, but that will be a separate patch. * cfg.mk (sc_correct_id_types): New syntax check. * src/libvirt-qemu.c (virDomainQemuAttach): Document why we didn't use pid_t for pid, and validate for overflow. * include/libvirt/libvirt-qemu.h (virDomainQemuAttach): Tweak name for syntax check. * src/vmware/vmware_conf.c (vmwareExtractPid): Likewise. * src/driver.h (virDrvDomainQemuAttach): Likewise. * tools/virsh.c (cmdQemuAttach): Likewise. * src/remote/qemu_protocol.x (qemu_domain_attach_args): Likewise. * src/qemu_protocol-structs (qemu_domain_attach_args): Likewise. * src/util/cgroup.c (virCgroupPidCode, virCgroupKillInternal): Likewise. * src/qemu/qemu_command.c(qemuParseProcFileStrings): Likewise. (qemuParseCommandLinePid): Use pid_t for pid. * daemon/libvirtd.c (daemonForkIntoBackground): Likewise. * src/conf/domain_conf.h (_virDomainObj): Likewise. * src/probes.d (rpc_socket_new): Likewise. * src/qemu/qemu_command.h (qemuParseCommandLinePid): Likewise. * src/qemu/qemu_driver.c (qemudGetProcessInfo, qemuDomainAttach): Likewise. * src/qemu/qemu_process.c (qemuProcessAttach): Likewise. * src/qemu/qemu_process.h (qemuProcessAttach): Likewise. * src/uml/uml_driver.c (umlGetProcessInfo): Likewise. * src/util/virnetdev.h (virNetDevSetNamespace): Likewise. * src/util/virnetdev.c (virNetDevSetNamespace): Likewise. * tests/testutils.c (virtTestCaptureProgramOutput): Likewise. * src/conf/storage_conf.h (_virStoragePerms): Use mode_t, uid_t, and gid_t rather than int. * src/security/security_dac.c (virSecurityDACSetOwnership): Likewise. * src/conf/storage_conf.c (virStorageDefParsePerms): Avoid compiler warning.
2012-02-10 23:08:11 +00:00
int virNetDevSetNamespace(const char *ifname, pid_t pidInNs)
{
int rc;
char *pid = NULL;
const char *argv[] = {
"ip", "link", "set", ifname, "netns", NULL, NULL
};
if (virAsprintf(&pid, "%lld", (long long) pidInNs) == -1)
return -1;
argv[5] = pid;
rc = virRun(argv, NULL);
VIR_FREE(pid);
return rc;
}
#if defined(SIOCSIFNAME) && defined(HAVE_STRUCT_IFREQ)
/**
* virNetDevSetName:
* @ifname: name of device
* @newifname: new name of @ifname
*
* Changes the name of the given device.
*
* Returns 0 on success, -1 on error
*/
int virNetDevSetName(const char* ifname, const char *newifname)
{
int fd = -1;
int ret = -1;
struct ifreq ifr;
if ((fd = virNetDevSetupControl(ifname, &ifr)) < 0)
return -1;
# ifdef HAVE_STRUCT_IFREQ_IFR_NEWNAME
if (virStrcpyStatic(ifr.ifr_newname, newifname) == NULL) {
virReportSystemError(ERANGE,
_("Network interface name '%s' is too long"),
newifname);
goto cleanup;
}
# else
ifr.ifr_data = (caddr_t)newifname;
# endif
if (ioctl(fd, SIOCSIFNAME, &ifr)) {
virReportSystemError(errno,
_("Unable to rename '%s' to '%s'"),
ifname, newifname);
goto cleanup;
}
ret = 0;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#else
int virNetDevSetName(const char* ifname, const char *newifname)
{
virReportSystemError(ENOSYS,
_("Cannot rename interface '%s' to '%s' on this platform"),
ifname, newifname);
return -1;
}
#endif
#if defined(SIOCSIFFLAGS) && defined(HAVE_STRUCT_IFREQ)
static int
virNetDevSetIFFlag(const char *ifname, int flag, bool val)
{
int fd = -1;
int ret = -1;
struct ifreq ifr;
int ifflags;
if ((fd = virNetDevSetupControl(ifname, &ifr)) < 0)
return -1;
if (ioctl(fd, SIOCGIFFLAGS, &ifr) < 0) {
virReportSystemError(errno,
_("Cannot get interface flags on '%s'"),
ifname);
goto cleanup;
}
if (val)
ifflags = ifr.ifr_flags | flag;
else
ifflags = ifr.ifr_flags & ~flag;
if (ifr.ifr_flags != ifflags) {
ifr.ifr_flags = ifflags;
if (ioctl(fd, SIOCSIFFLAGS, &ifr) < 0) {
virReportSystemError(errno,
_("Cannot set interface flags on '%s'"),
ifname);
goto cleanup;
}
}
ret = 0;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#else
static int
virNetDevSetIFFlag(const char *ifname,
int flag ATTRIBUTE_UNUSED,
bool val ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS,
_("Cannot set interface flags on '%s'"),
ifname);
return -1;
}
#endif
/**
* virNetDevSetOnline:
* @ifname: the interface name
* @online: true for up, false for down
*
* Function to control if an interface is activated (up, true) or not (down, false)
*
* Returns 0 in case of success or -1 on error.
*/
int
virNetDevSetOnline(const char *ifname,
bool online)
{
return virNetDevSetIFFlag(ifname, VIR_IFF_UP, online);
}
/**
* virNetDevSetPromiscuous:
* @ifname: the interface name
* @promiscuous: true for receive all packets, false for do not receive
* all packets
*
* Function to control if an interface is to receive all
* packets (receive all, true) or not (do not receive all, false)
*
* Returns 0 in case of success or -1 on error.
*/
int
virNetDevSetPromiscuous(const char *ifname,
bool promiscuous)
{
return virNetDevSetIFFlag(ifname, VIR_IFF_PROMISC, promiscuous);
}
/**
* virNetDevSetRcvMulti:
* @ifname: the interface name
* @:receive true for receive multicast packets, false for do not receive
* multicast packets
*
* Function to control if an interface is to receive multicast
* packets in which it is interested (receive, true)
* or not (do not receive, false)
*
* Returns 0 in case of success or -1 on error.
*/
int
virNetDevSetRcvMulti(const char *ifname,
bool receive)
{
return virNetDevSetIFFlag(ifname, VIR_IFF_MULTICAST, receive);
}
/**
* virNetDevSetRcvAllMulti:
* @ifname: the interface name
* @:receive true for receive all packets, false for do not receive all packets
*
* Function to control if an interface is to receive all multicast
* packets (receive, true) or not (do not receive, false)
*
* Returns 0 in case of success or -1 on error.
*/
int
virNetDevSetRcvAllMulti(const char *ifname,
bool receive)
{
return virNetDevSetIFFlag(ifname, VIR_IFF_ALLMULTI, receive);
}
#if defined(SIOCGIFFLAGS) && defined(HAVE_STRUCT_IFREQ)
static int
virNetDevGetIFFlag(const char *ifname, int flag, bool *val)
{
int fd = -1;
int ret = -1;
struct ifreq ifr;
if ((fd = virNetDevSetupControl(ifname, &ifr)) < 0)
return -1;
if (ioctl(fd, SIOCGIFFLAGS, &ifr) < 0) {
virReportSystemError(errno,
_("Cannot get interface flags on '%s'"),
ifname);
goto cleanup;
}
*val = (ifr.ifr_flags & flag) ? true : false;
ret = 0;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#else
static int
virNetDevGetIFFlag(const char *ifname,
int flag ATTRIBUTE_UNUSED,
bool *val ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS,
_("Cannot get interface flags on '%s'"),
ifname);
return -1;
}
#endif
/**
* virNetDevGetOnline:
* @ifname: the interface name
* @online: where to store the status
*
* Function to query if an interface is activated (true) or not (false)
*
* Returns 0 in case of success or an errno code in case of failure.
*/
int
virNetDevGetOnline(const char *ifname,
bool *online)
{
return virNetDevGetIFFlag(ifname, VIR_IFF_UP, online);
}
/**
* virNetDevIsPromiscuous:
* @ifname: the interface name
* @promiscuous: where to store the status
*
* Function to query if an interface is receiving all packets (true) or
* not (false)
*
* Returns 0 in case of success or an errno code in case of failure.
*/
int
virNetDevGetPromiscuous(const char *ifname,
bool *promiscuous)
{
return virNetDevGetIFFlag(ifname, VIR_IFF_PROMISC, promiscuous);
}
/**
* virNetDevIsRcvMulti:
* @ifname: the interface name
* @receive where to store the status
*
* Function to query whether an interface is receiving multicast packets (true)
* in which it is interested, or not (false)
*
* Returns 0 in case of success or -1 on error.
*/
int
virNetDevGetRcvMulti(const char *ifname,
bool *receive)
{
return virNetDevGetIFFlag(ifname, VIR_IFF_MULTICAST, receive);
}
/**
* virNetDevIsRcvAllMulti:
* @ifname: the interface name
* @:receive where to store the status
*
* Function to query whether an interface is receiving all multicast
* packets (receiving, true) or not (is not receiving, false)
*
* Returns 0 in case of success or -1 on error.
*/
int
virNetDevGetRcvAllMulti(const char *ifname,
bool *receive)
{
return virNetDevGetIFFlag(ifname, VIR_IFF_ALLMULTI, receive);
}
/**
* virNetDevGetIndex:
* @ifname : Name of the interface whose index is to be found
* @ifindex: Pointer to int where the index will be written into
*
* Get the index of an interface given its name.
*
* Returns 0 on success, -1 on failure
*/
#if defined(SIOCGIFINDEX) && defined(HAVE_STRUCT_IFREQ)
int virNetDevGetIndex(const char *ifname, int *ifindex)
{
int ret = -1;
struct ifreq ifreq;
int fd = socket(VIR_NETDEV_FAMILY, SOCK_DGRAM, 0);
if (fd < 0) {
virReportSystemError(errno, "%s",
_("Unable to open control socket"));
return -1;
}
memset(&ifreq, 0, sizeof(ifreq));
if (virStrncpy(ifreq.ifr_name, ifname, strlen(ifname),
sizeof(ifreq.ifr_name)) == NULL) {
virReportSystemError(ERANGE,
_("invalid interface name %s"),
ifname);
goto cleanup;
}
if (ioctl(fd, SIOCGIFINDEX, &ifreq) < 0) {
virReportSystemError(errno,
_("Unable to get index for interface %s"), ifname);
goto cleanup;
}
# ifdef HAVE_STRUCT_IFREQ_IFR_INDEX
*ifindex = ifreq.ifr_index;
# else
*ifindex = ifreq.ifr_ifindex;
# endif
ret = 0;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#else /* ! SIOCGIFINDEX */
int virNetDevGetIndex(const char *ifname ATTRIBUTE_UNUSED,
int *ifindex ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
_("Unable to get interface index on this platform"));
return -1;
}
#endif /* ! SIOCGIFINDEX */
#if defined(SIOCGIFVLAN) && defined(HAVE_STRUCT_IFREQ) && HAVE_DECL_GET_VLAN_VID_CMD
int virNetDevGetVLanID(const char *ifname, int *vlanid)
{
struct vlan_ioctl_args vlanargs = {
.cmd = GET_VLAN_VID_CMD,
};
int ret = -1;
int fd = socket(PF_PACKET, SOCK_DGRAM, 0);
if (fd < 0) {
virReportSystemError(errno, "%s",
_("Unable to open control socket"));
return -1;
}
if (virStrcpyStatic(vlanargs.device1, ifname) == NULL) {
virReportSystemError(ERANGE,
_("invalid interface name %s"),
ifname);
goto cleanup;
}
if (ioctl(fd, SIOCGIFVLAN, &vlanargs) != 0) {
virReportSystemError(errno,
_("Unable to get VLAN for interface %s"), ifname);
goto cleanup;
}
*vlanid = vlanargs.u.VID;
ret = 0;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#else /* ! SIOCGIFVLAN */
int virNetDevGetVLanID(const char *ifname ATTRIBUTE_UNUSED,
int *vlanid ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
_("Unable to get VLAN on this platform"));
return -1;
}
#endif /* ! SIOCGIFVLAN */
#if defined(__linux__) && defined(HAVE_LIBNL)
static int
virNetDevGetIPAddressBinary(virSocketAddr *addr, void **data, size_t *len)
{
if (!addr)
return -1;
switch (VIR_SOCKET_ADDR_FAMILY(addr)) {
case AF_INET:
*data = &addr->data.inet4.sin_addr;
*len = sizeof(struct in_addr);
break;
case AF_INET6:
*data = &addr->data.inet6.sin6_addr;
*len = sizeof(struct in6_addr);
break;
default:
return -1;
}
return 0;
}
static struct nl_msg *
virNetDevCreateNetlinkAddressMessage(int messageType,
const char *ifname,
virSocketAddr *addr,
unsigned int prefix,
virSocketAddr *broadcast)
{
struct nl_msg *nlmsg = NULL;
struct ifaddrmsg ifa;
unsigned int ifindex;
void *addrData = NULL;
void *broadcastData = NULL;
size_t addrDataLen;
if (virNetDevGetIPAddressBinary(addr, &addrData, &addrDataLen) < 0)
return NULL;
if (broadcast && virNetDevGetIPAddressBinary(broadcast, &broadcastData,
&addrDataLen) < 0)
return NULL;
/* Get the interface index */
if ((ifindex = if_nametoindex(ifname)) == 0)
return NULL;
if (!(nlmsg = nlmsg_alloc_simple(messageType,
NLM_F_REQUEST | NLM_F_CREATE | NLM_F_EXCL))) {
virReportOOMError();
return NULL;
}
memset(&ifa, 0, sizeof(ifa));
ifa.ifa_prefixlen = prefix;
ifa.ifa_family = VIR_SOCKET_ADDR_FAMILY(addr);
ifa.ifa_index = ifindex;
ifa.ifa_scope = 0;
if (nlmsg_append(nlmsg, &ifa, sizeof(ifa), NLMSG_ALIGNTO) < 0)
goto buffer_too_small;
if (nla_put(nlmsg, IFA_LOCAL, addrDataLen, addrData) < 0)
goto buffer_too_small;
if (nla_put(nlmsg, IFA_ADDRESS, addrDataLen, addrData) < 0)
goto buffer_too_small;
if (broadcastData &&
nla_put(nlmsg, IFA_BROADCAST, addrDataLen, broadcastData) < 0)
goto buffer_too_small;
return nlmsg;
buffer_too_small:
virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
_("allocated netlink buffer is too small"));
nlmsg_free(nlmsg);
return NULL;
}
/**
* virNetDevSetIPAddress:
* @ifname: the interface name
* @addr: the IP address (IPv4 or IPv6)
* @prefix: number of 1 bits in the netmask
*
* Add an IP address to an interface. This function *does not* remove
* any previously added IP addresses - that must be done separately with
* brDelInetAddress.
*
* Returns 0 in case of success or -1 in case of error.
*/
int virNetDevSetIPAddress(const char *ifname,
virSocketAddr *addr,
unsigned int prefix)
{
virSocketAddr *broadcast = NULL;
int ret = -1;
struct nl_msg *nlmsg = NULL;
struct nlmsghdr *resp = NULL;
unsigned int recvbuflen;
/* The caller needs to provide a correct address */
if (VIR_SOCKET_ADDR_FAMILY(addr) == AF_INET) {
/* compute a broadcast address if this is IPv4 */
if (VIR_ALLOC(broadcast) < 0)
return -1;
if (virSocketAddrBroadcastByPrefix(addr, prefix, broadcast) < 0)
goto cleanup;
}
if (!(nlmsg = virNetDevCreateNetlinkAddressMessage(RTM_NEWADDR, ifname,
addr, prefix,
broadcast)))
goto cleanup;
if (virNetlinkCommand(nlmsg, &resp, &recvbuflen, 0, 0,
NETLINK_ROUTE, 0) < 0)
goto cleanup;
if (virNetlinkGetErrorCode(resp, recvbuflen) < 0) {
virReportError(VIR_ERR_SYSTEM_ERROR,
_("Error adding IP address to %s"), ifname);
goto cleanup;
}
ret = 0;
cleanup:
nlmsg_free(nlmsg);
VIR_FREE(resp);
VIR_FREE(broadcast);
return ret;
}
/**
* virNetDevAddRoute:
* @ifname: the interface name
* @addr: the IP network address (IPv4 or IPv6)
* @prefix: number of 1 bits in the netmask
* @gateway: via address for route (same as @addr)
*
* Add a route for a network IP address to an interface. This function
* *does not* remove any previously added IP static routes.
*
* Returns 0 in case of success or -1 in case of error.
*/
int
virNetDevAddRoute(const char *ifname,
virSocketAddrPtr addr,
unsigned int prefix,
virSocketAddrPtr gateway,
unsigned int metric)
{
int ret = -1;
struct nl_msg *nlmsg = NULL;
struct nlmsghdr *resp = NULL;
unsigned int recvbuflen;
unsigned int ifindex;
struct rtmsg rtmsg;
void *gatewayData = NULL;
void *addrData = NULL;
size_t addrDataLen;
int errCode;
virSocketAddr defaultAddr;
virSocketAddrPtr actualAddr;
char *toStr = NULL;
char *viaStr = NULL;
actualAddr = addr;
/* If we have no valid network address, then use the default one */
if (!addr || !VIR_SOCKET_ADDR_VALID(addr)) {
VIR_DEBUG("computing default address");
int family = VIR_SOCKET_ADDR_FAMILY(gateway);
if (family == AF_INET) {
if (virSocketAddrParseIPv4(&defaultAddr, VIR_SOCKET_ADDR_IPV4_ALL) < 0)
goto cleanup;
} else {
if (virSocketAddrParseIPv6(&defaultAddr, VIR_SOCKET_ADDR_IPV6_ALL) < 0)
goto cleanup;
}
actualAddr = &defaultAddr;
}
toStr = virSocketAddrFormat(actualAddr);
viaStr = virSocketAddrFormat(gateway);
VIR_DEBUG("Adding route %s/%d via %s", toStr, prefix, viaStr);
if (virNetDevGetIPAddressBinary(actualAddr, &addrData, &addrDataLen) < 0 ||
virNetDevGetIPAddressBinary(gateway, &gatewayData, &addrDataLen) < 0)
goto cleanup;
/* Get the interface index */
if ((ifindex = if_nametoindex(ifname)) == 0)
goto cleanup;
if (!(nlmsg = nlmsg_alloc_simple(RTM_NEWROUTE,
NLM_F_REQUEST | NLM_F_CREATE |
NLM_F_EXCL))) {
virReportOOMError();
goto cleanup;
}
memset(&rtmsg, 0, sizeof(rtmsg));
rtmsg.rtm_family = VIR_SOCKET_ADDR_FAMILY(gateway);
rtmsg.rtm_table = RT_TABLE_MAIN;
rtmsg.rtm_scope = RT_SCOPE_UNIVERSE;
rtmsg.rtm_protocol = RTPROT_BOOT;
rtmsg.rtm_type = RTN_UNICAST;
rtmsg.rtm_dst_len = prefix;
if (nlmsg_append(nlmsg, &rtmsg, sizeof(rtmsg), NLMSG_ALIGNTO) < 0)
goto buffer_too_small;
if (prefix > 0 && nla_put(nlmsg, RTA_DST, addrDataLen, addrData) < 0)
goto buffer_too_small;
if (nla_put(nlmsg, RTA_GATEWAY, addrDataLen, gatewayData) < 0)
goto buffer_too_small;
if (nla_put_u32(nlmsg, RTA_OIF, ifindex) < 0)
goto buffer_too_small;
if (metric > 0 && nla_put_u32(nlmsg, RTA_PRIORITY, metric) < 0)
goto buffer_too_small;
if (virNetlinkCommand(nlmsg, &resp, &recvbuflen, 0, 0,
NETLINK_ROUTE, 0) < 0)
goto cleanup;
if ((errCode = virNetlinkGetErrorCode(resp, recvbuflen)) < 0) {
virReportSystemError(errCode, _("Error adding route to %s"), ifname);
goto cleanup;
}
ret = 0;
cleanup:
VIR_FREE(toStr);
VIR_FREE(viaStr);
nlmsg_free(nlmsg);
return ret;
buffer_too_small:
virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
_("allocated netlink buffer is too small"));
goto cleanup;
}
/**
* virNetDevClearIPAddress:
* @ifname: the interface name
* @addr: the IP address (IPv4 or IPv6)
* @prefix: number of 1 bits in the netmask
*
* Delete an IP address from an interface.
*
* Returns 0 in case of success or -1 in case of error.
*/
int virNetDevClearIPAddress(const char *ifname,
virSocketAddr *addr,
unsigned int prefix)
{
int ret = -1;
struct nl_msg *nlmsg = NULL;
struct nlmsghdr *resp = NULL;
unsigned int recvbuflen;
if (!(nlmsg = virNetDevCreateNetlinkAddressMessage(RTM_DELADDR, ifname,
addr, prefix,
NULL)))
goto cleanup;
if (virNetlinkCommand(nlmsg, &resp, &recvbuflen, 0, 0,
NETLINK_ROUTE, 0) < 0)
goto cleanup;
if (virNetlinkGetErrorCode(resp, recvbuflen) < 0) {
virReportError(VIR_ERR_SYSTEM_ERROR,
_("Error removing IP address from %s"), ifname);
goto cleanup;
}
ret = 0;
cleanup:
nlmsg_free(nlmsg);
VIR_FREE(resp);
return ret;
}
#else /* defined(__linux__) && defined(HAVE_LIBNL) */
int virNetDevSetIPAddress(const char *ifname,
virSocketAddr *addr,
unsigned int prefix)
{
virCommandPtr cmd = NULL;
char *addrstr = NULL, *bcaststr = NULL;
virSocketAddr broadcast;
int ret = -1;
if (!(addrstr = virSocketAddrFormat(addr)))
goto cleanup;
/* format up a broadcast address if this is IPv4 */
if ((VIR_SOCKET_ADDR_IS_FAMILY(addr, AF_INET)) &&
((virSocketAddrBroadcastByPrefix(addr, prefix, &broadcast) < 0) ||
!(bcaststr = virSocketAddrFormat(&broadcast)))) {
goto cleanup;
}
# ifdef IFCONFIG_PATH
cmd = virCommandNew(IFCONFIG_PATH);
virCommandAddArg(cmd, ifname);
if (VIR_SOCKET_ADDR_IS_FAMILY(addr, AF_INET6))
virCommandAddArg(cmd, "inet6");
else
virCommandAddArg(cmd, "inet");
virCommandAddArgFormat(cmd, "%s/%u", addrstr, prefix);
if (bcaststr)
virCommandAddArgList(cmd, "broadcast", bcaststr, NULL);
virCommandAddArg(cmd, "alias");
# else
cmd = virCommandNew(IP_PATH);
virCommandAddArgList(cmd, "addr", "add", NULL);
virCommandAddArgFormat(cmd, "%s/%u", addrstr, prefix);
if (bcaststr)
virCommandAddArgList(cmd, "broadcast", bcaststr, NULL);
virCommandAddArgList(cmd, "dev", ifname, NULL);
# endif
if (virCommandRun(cmd, NULL) < 0)
goto cleanup;
ret = 0;
cleanup:
VIR_FREE(addrstr);
VIR_FREE(bcaststr);
virCommandFree(cmd);
return ret;
}
Support for static routes on a virtual bridge network: static route support for <network> This patch adds the <route> subelement of <network> to define a static route. the address and prefix (or netmask) attribute identify the destination network, and the gateway attribute specifies the next hop address (which must be directly reachable from the containing <network>) which is to receive the packets destined for "address/(prefix|netmask)". These attributes are translated into an "ip route add" command that is executed when the network is started. The command used is of the following form: ip route add <address>/<prefix> via <gateway> \ dev <virbr-bridge> proto static metric <metric> Tests are done to validate that the input data are correct. For example, for a static route ip definition, the address must be a network address and not a host address. Additional checks are added to ensure that the specified gateway is directly reachable via this network (i.e. that the gateway IP address is in the same subnet as one of the IP's defined for the network). prefix='0' is supported for both family='ipv4' address='0.0.0.0' netmask='0.0.0.0' or prefix='0', and for family='ipv6' address='::', prefix=0', although care should be taken to not override a desired system default route. Anytime an attempt is made to define a static route which *exactly* duplicates an existing static route (for example, address=::, prefix=0, metric=1), the following error message will be sent to syslog: RTNETLINK answers: File exists This can be overridden by decreasing the metric value for the route that should be preferred, or increasing the metric for the route that shouldn't be preferred (and is thus in place only in anticipation that the preferred route may be removed in the future). Caution should be used when manipulating route metrics, especially for a default route. Note: The use of the command-line interface should be replaced by direct use of libnl so that error conditions can be handled better. But, that is being left as an exercise for another day. Signed-off-by: Gene Czarcinski <gene@czarc.net> Signed-off-by: Laine Stump <laine@laine.org>
2013-05-07 17:42:55 +00:00
int
virNetDevAddRoute(const char *ifname,
virSocketAddrPtr addr,
unsigned int prefix,
virSocketAddrPtr gateway,
unsigned int metric)
{
virCommandPtr cmd = NULL;
char *addrstr = NULL, *gatewaystr = NULL;
int ret = -1;
if (!(addrstr = virSocketAddrFormat(addr)))
goto cleanup;
if (!(gatewaystr = virSocketAddrFormat(gateway)))
goto cleanup;
cmd = virCommandNew(IP_PATH);
virCommandAddArgList(cmd, "route", "add", NULL);
virCommandAddArgFormat(cmd, "%s/%u", addrstr, prefix);
virCommandAddArgList(cmd, "via", gatewaystr, "dev", ifname,
"proto", "static", "metric", NULL);
virCommandAddArgFormat(cmd, "%u", metric);
if (virCommandRun(cmd, NULL) < 0)
goto cleanup;
ret = 0;
cleanup:
Support for static routes on a virtual bridge network: static route support for <network> This patch adds the <route> subelement of <network> to define a static route. the address and prefix (or netmask) attribute identify the destination network, and the gateway attribute specifies the next hop address (which must be directly reachable from the containing <network>) which is to receive the packets destined for "address/(prefix|netmask)". These attributes are translated into an "ip route add" command that is executed when the network is started. The command used is of the following form: ip route add <address>/<prefix> via <gateway> \ dev <virbr-bridge> proto static metric <metric> Tests are done to validate that the input data are correct. For example, for a static route ip definition, the address must be a network address and not a host address. Additional checks are added to ensure that the specified gateway is directly reachable via this network (i.e. that the gateway IP address is in the same subnet as one of the IP's defined for the network). prefix='0' is supported for both family='ipv4' address='0.0.0.0' netmask='0.0.0.0' or prefix='0', and for family='ipv6' address='::', prefix=0', although care should be taken to not override a desired system default route. Anytime an attempt is made to define a static route which *exactly* duplicates an existing static route (for example, address=::, prefix=0, metric=1), the following error message will be sent to syslog: RTNETLINK answers: File exists This can be overridden by decreasing the metric value for the route that should be preferred, or increasing the metric for the route that shouldn't be preferred (and is thus in place only in anticipation that the preferred route may be removed in the future). Caution should be used when manipulating route metrics, especially for a default route. Note: The use of the command-line interface should be replaced by direct use of libnl so that error conditions can be handled better. But, that is being left as an exercise for another day. Signed-off-by: Gene Czarcinski <gene@czarc.net> Signed-off-by: Laine Stump <laine@laine.org>
2013-05-07 17:42:55 +00:00
VIR_FREE(addrstr);
VIR_FREE(gatewaystr);
virCommandFree(cmd);
return ret;
}
int virNetDevClearIPAddress(const char *ifname,
virSocketAddr *addr,
unsigned int prefix)
{
virCommandPtr cmd = NULL;
char *addrstr;
int ret = -1;
if (!(addrstr = virSocketAddrFormat(addr)))
goto cleanup;
# ifdef IFCONFIG_PATH
cmd = virCommandNew(IFCONFIG_PATH);
virCommandAddArg(cmd, ifname);
if (VIR_SOCKET_ADDR_IS_FAMILY(addr, AF_INET6))
virCommandAddArg(cmd, "inet6");
else
virCommandAddArg(cmd, "inet");
virCommandAddArgFormat(cmd, "%s/%u", addrstr, prefix);
virCommandAddArg(cmd, "-alias");
# else
cmd = virCommandNew(IP_PATH);
virCommandAddArgList(cmd, "addr", "del", NULL);
virCommandAddArgFormat(cmd, "%s/%u", addrstr, prefix);
virCommandAddArgList(cmd, "dev", ifname, NULL);
# endif
if (virCommandRun(cmd, NULL) < 0)
goto cleanup;
ret = 0;
cleanup:
VIR_FREE(addrstr);
virCommandFree(cmd);
return ret;
}
#endif /* defined(__linux__) && defined(HAVE_LIBNL) */
/**
* virNetDevGetIPv4AddressIoctl:
* @ifname: name of the interface whose IP address we want
* @addr: filled with the IPv4 address
*
* This function gets the IPv4 address for the interface @ifname
* and stores it in @addr
*
* Returns 0 on success, -errno on failure.
*/
#if defined(SIOCGIFADDR) && defined(HAVE_STRUCT_IFREQ)
static int
virNetDevGetIPv4AddressIoctl(const char *ifname,
virSocketAddrPtr addr)
{
int fd = -1;
int ret = -1;
struct ifreq ifr;
if ((fd = virNetDevSetupControl(ifname, &ifr)) < 0)
return -1;
if (ioctl(fd, SIOCGIFADDR, (char *)&ifr) < 0) {
virReportSystemError(errno,
_("Unable to get IPv4 address for interface %s via ioctl"),
ifname);
goto cleanup;
}
addr->data.stor.ss_family = AF_INET;
addr->len = sizeof(addr->data.inet4);
memcpy(&addr->data.inet4, &ifr.ifr_addr, addr->len);
ret = 0;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#else /* ! SIOCGIFADDR */
static int
virNetDevGetIPv4AddressIoctl(const char *ifname ATTRIBUTE_UNUSED,
virSocketAddrPtr addr ATTRIBUTE_UNUSED)
{
return -2;
}
#endif /* ! SIOCGIFADDR */
/**
* virNetDevGetIPAddress:
* @ifname: name of the interface whose IP address we want
* @addr: filled with the IPv4 address
*
* This function gets the IPv4 address for the interface @ifname
* and stores it in @addr
*
* Returns 0 on success, -errno on failure.
*/
int
virNetDevGetIPAddress(const char *ifname,
virSocketAddrPtr addr)
{
int ret;
memset(addr, 0, sizeof(*addr));
addr->data.stor.ss_family = AF_UNSPEC;
if ((ret = virNetDevGetIPv4AddressIoctl(ifname, addr)) != -2)
return ret;
virReportSystemError(ENOSYS, "%s",
_("Unable to get IP address on this platform"));
return -1;
}
/**
* virNetDevValidateConfig:
* @ifname: Name of the interface
* @macaddr: expected MAC address of the interface; not checked if NULL
* @ifindex: expected index of the interface; not checked if '-1'
*
* Determine whether a given interface is still available. If so,
* it must have the given MAC address and if an interface index is
* passed, it must also match the interface index.
*
* Returns 1 if the config matches, 0 if the config does not match, or interface does not exist, -1 on error
*/
#if defined(SIOCGIFHWADDR) && defined(HAVE_STRUCT_IFREQ)
int virNetDevValidateConfig(const char *ifname,
maint: avoid 'const fooPtr' in virnet files 'const fooPtr' is the same as 'foo * const' (the pointer won't change, but it's contents can). But in general, if an interface is trying to be const-correct, it should be using 'const foo *' (the pointer is to data that can't be changed). Fix up remaining offenders in src/util. * src/util/virnetdev.h (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Use intended type. * src/util/virnetdevbandwidth.h (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.h (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.h (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.h (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.h (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.h (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate): Likewise. * src/util/virnetlink.h (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. * src/util/virnetdev.c (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Fix fallout. * src/util/virnetdevbandwidth.c (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.c (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.c (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.c (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.c (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.c (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate) (virNetDevVPortProfileOpSetLink, virNetDevVPortProfileOpCommon) (virNetDevVPortProfileOp8021Qbg, virNetDevVPortProfileOp8021Qbh): Likewise. * src/util/virnetlink.c (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. Signed-off-by: Eric Blake <eblake@redhat.com>
2013-10-05 19:41:44 +00:00
const virMacAddr *macaddr, int ifindex)
{
int fd = -1;
int ret = -1;
struct ifreq ifr;
int idx;
int rc;
if ((rc = virNetDevExists(ifname)) < 0)
return -1;
if (rc == 0) {
ret = 0;
goto cleanup;
}
if (macaddr != NULL) {
if ((fd = virNetDevSetupControl(ifname, &ifr)) < 0)
return -1;
if (ioctl(fd, SIOCGIFHWADDR, &ifr) < 0) {
if (errno == ENODEV) {
ret = 0;
goto cleanup;
}
virReportSystemError(errno,
_("could not get MAC address of interface %s"),
ifname);
goto cleanup;
}
if (virMacAddrCmpRaw(macaddr,
(unsigned char *)ifr.ifr_hwaddr.sa_data) != 0) {
ret = 0;
goto cleanup;
}
}
if (ifindex != -1) {
if (virNetDevGetIndex(ifname, &idx) < 0)
goto cleanup;
maint: use consistent if-else braces in remaining spots I'm about to add a syntax check that enforces our documented HACKING style of always using matching {} on if-else statements. This patch focuses on all remaining problems, where there weren't enough issues to warrant splitting it further. * src/remote/remote_driver.c (doRemoteOpen): Correct use of {}. * src/security/virt-aa-helper.c (vah_add_path, valid_path, main): Likewise. * src/rpc/virnetsocket.c (virNetSocketNewConnectLibSSH2): Likewise. * src/esx/esx_vi_types.c (esxVI_Type_FromString): Likewise. * src/uml/uml_driver.c (umlDomainDetachDevice): Likewise. * src/util/viralloc.c (virShrinkN): Likewise. * src/util/virbuffer.c (virBufferURIEncodeString): Likewise. * src/util/virdbus.c (virDBusCall): Likewise. * src/util/virnetdev.c (virNetDevValidateConfig): Likewise. * src/util/virnetdevvportprofile.c (virNetDevVPortProfileGetNthParent): Likewise. * src/util/virpci.c (virPCIDeviceIterDevices) (virPCIDeviceWaitForCleanup) (virPCIDeviceIsBehindSwitchLackingACS): Likewise. * src/util/virsocketaddr.c (virSocketAddrGetNumNetmaskBits): Likewise. * src/util/viruri.c (virURIParseParams): Likewise. * daemon/stream.c (daemonStreamHandleAbort): Likewise. * tests/testutils.c (virtTestResult): Likewise. * tests/cputest.c (cpuTestBaseline): Likewise. * tools/virsh-domain.c (cmdDomPMSuspend): Likewise. * tools/virsh-host.c (cmdNodeSuspend): Likewise. * src/esx/esx_vi_generator.py (Type.generate_typefromstring): Tweak generated code. Signed-off-by: Eric Blake <eblake@redhat.com>
2014-09-03 19:39:21 +00:00
if (idx != ifindex) {
ret = 0;
goto cleanup;
}
}
ret = 1;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#else
int virNetDevValidateConfig(const char *ifname ATTRIBUTE_UNUSED,
maint: avoid 'const fooPtr' in virnet files 'const fooPtr' is the same as 'foo * const' (the pointer won't change, but it's contents can). But in general, if an interface is trying to be const-correct, it should be using 'const foo *' (the pointer is to data that can't be changed). Fix up remaining offenders in src/util. * src/util/virnetdev.h (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Use intended type. * src/util/virnetdevbandwidth.h (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.h (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.h (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.h (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.h (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.h (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate): Likewise. * src/util/virnetlink.h (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. * src/util/virnetdev.c (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Fix fallout. * src/util/virnetdevbandwidth.c (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.c (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.c (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.c (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.c (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.c (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate) (virNetDevVPortProfileOpSetLink, virNetDevVPortProfileOpCommon) (virNetDevVPortProfileOp8021Qbg, virNetDevVPortProfileOp8021Qbh): Likewise. * src/util/virnetlink.c (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. Signed-off-by: Eric Blake <eblake@redhat.com>
2013-10-05 19:41:44 +00:00
const virMacAddr *macaddr ATTRIBUTE_UNUSED,
int ifindex ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
_("Unable to check interface config on this platform"));
return -1;
}
#endif
#ifdef __linux__
# define NET_SYSFS "/sys/class/net/"
static int
virNetDevSysfsFile(char **pf_sysfs_device_link, const char *ifname,
const char *file)
{
if (virAsprintf(pf_sysfs_device_link, NET_SYSFS "%s/%s", ifname, file) < 0)
return -1;
return 0;
}
static int
virNetDevSysfsDeviceFile(char **pf_sysfs_device_link, const char *ifname,
const char *file)
{
if (virAsprintf(pf_sysfs_device_link, NET_SYSFS "%s/device/%s", ifname,
file) < 0)
return -1;
return 0;
}
/**
* virNetDevGetVirtualFunctions:
*
* @pfname : name of the physical function interface name
* @vfname: array that will hold the interface names of the virtual_functions
* @n_vfname: pointer to the number of virtual functions
*
* Returns 0 on success and -1 on failure
*/
int
virNetDevGetVirtualFunctions(const char *pfname,
char ***vfname,
virPCIDeviceAddressPtr **virt_fns,
size_t *n_vfname)
{
int ret = -1;
size_t i;
char *pf_sysfs_device_link = NULL;
char *pci_sysfs_device_link = NULL;
char *pciConfigAddr = NULL;
*virt_fns = NULL;
*n_vfname = 0;
if (virNetDevSysfsFile(&pf_sysfs_device_link, pfname, "device") < 0)
return ret;
if (virPCIGetVirtualFunctions(pf_sysfs_device_link, virt_fns,
n_vfname) < 0)
goto cleanup;
if (VIR_ALLOC_N(*vfname, *n_vfname) < 0)
goto cleanup;
for (i = 0; i < *n_vfname; i++) {
if (virPCIGetAddrString((*virt_fns)[i]->domain,
(*virt_fns)[i]->bus,
(*virt_fns)[i]->slot,
(*virt_fns)[i]->function,
&pciConfigAddr) < 0) {
virReportSystemError(ENOSYS, "%s",
_("Failed to get PCI Config Address String"));
goto cleanup;
}
if (virPCIGetSysfsFile(pciConfigAddr, &pci_sysfs_device_link) < 0) {
virReportSystemError(ENOSYS, "%s",
_("Failed to get PCI SYSFS file"));
goto cleanup;
}
if (virPCIGetNetName(pci_sysfs_device_link, &((*vfname)[i])) < 0)
VIR_INFO("VF does not have an interface name");
}
ret = 0;
cleanup:
if (ret < 0) {
VIR_FREE(*vfname);
VIR_FREE(*virt_fns);
}
VIR_FREE(pf_sysfs_device_link);
VIR_FREE(pci_sysfs_device_link);
VIR_FREE(pciConfigAddr);
return ret;
}
/**
* virNetDevIsVirtualFunction:
* @ifname : name of the interface
*
* Checks if an interface is a SRIOV virtual function.
*
* Returns 1 if interface is SRIOV virtual function, 0 if not and -1 if error
*
*/
int
virNetDevIsVirtualFunction(const char *ifname)
{
char *if_sysfs_device_link = NULL;
int ret = -1;
if (virNetDevSysfsFile(&if_sysfs_device_link, ifname, "device") < 0)
return ret;
ret = virPCIIsVirtualFunction(if_sysfs_device_link);
VIR_FREE(if_sysfs_device_link);
return ret;
}
/**
* virNetDevGetVirtualFunctionIndex
*
* @pfname : name of the physical function interface name
* @vfname : name of the virtual function interface name
* @vf_index : Pointer to int. Contains vf index of interface upon successful
* return
*
* Returns 0 on success, -1 on failure
*
*/
int
virNetDevGetVirtualFunctionIndex(const char *pfname, const char *vfname,
int *vf_index)
{
char *pf_sysfs_device_link = NULL, *vf_sysfs_device_link = NULL;
int ret = -1;
if (virNetDevSysfsFile(&pf_sysfs_device_link, pfname, "device") < 0)
return ret;
if (virNetDevSysfsFile(&vf_sysfs_device_link, vfname, "device") < 0) {
VIR_FREE(pf_sysfs_device_link);
return ret;
}
ret = virPCIGetVirtualFunctionIndex(pf_sysfs_device_link,
vf_sysfs_device_link,
vf_index);
VIR_FREE(pf_sysfs_device_link);
VIR_FREE(vf_sysfs_device_link);
return ret;
}
/**
* virNetDevGetPhysicalFunction
*
* @ifname : name of the physical function interface name
* @pfname : Contains sriov physical function for interface ifname
* upon successful return
*
* Returns 0 on success, -1 on failure
*
*/
int
virNetDevGetPhysicalFunction(const char *ifname, char **pfname)
{
char *physfn_sysfs_path = NULL;
int ret = -1;
if (virNetDevSysfsDeviceFile(&physfn_sysfs_path, ifname, "physfn") < 0)
return ret;
ret = virPCIGetNetName(physfn_sysfs_path, pfname);
VIR_FREE(physfn_sysfs_path);
return ret;
}
/**
* virNetDevGetVirtualFunctionInfo:
* @vfname: name of the virtual function interface
* @pfname: name of the physical function
* @vf: vf index
*
* Returns 0 on success, -errno on failure.
*
*/
int
virNetDevGetVirtualFunctionInfo(const char *vfname, char **pfname,
int *vf)
{
char *pf_sysfs_path = NULL, *vf_sysfs_path = NULL;
int ret = -1;
*pfname = NULL;
if (virNetDevGetPhysicalFunction(vfname, pfname) < 0)
return ret;
if (virNetDevSysfsFile(&pf_sysfs_path, *pfname, "device") < 0)
goto cleanup;
if (virNetDevSysfsFile(&vf_sysfs_path, vfname, "device") < 0)
goto cleanup;
ret = virPCIGetVirtualFunctionIndex(pf_sysfs_path, vf_sysfs_path, vf);
cleanup:
if (ret < 0)
VIR_FREE(*pfname);
VIR_FREE(vf_sysfs_path);
VIR_FREE(pf_sysfs_path);
return ret;
}
#else /* !__linux__ */
int
virNetDevGetVirtualFunctions(const char *pfname ATTRIBUTE_UNUSED,
char ***vfname ATTRIBUTE_UNUSED,
virPCIDeviceAddressPtr **virt_fns ATTRIBUTE_UNUSED,
size_t *n_vfname ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
2012-04-01 09:23:56 +00:00
_("Unable to get virtual functions on this platform"));
return -1;
}
int
virNetDevIsVirtualFunction(const char *ifname ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
_("Unable to check virtual function status on this platform"));
return -1;
}
int
virNetDevGetVirtualFunctionIndex(const char *pfname ATTRIBUTE_UNUSED,
const char *vfname ATTRIBUTE_UNUSED,
int *vf_index ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
_("Unable to get virtual function index on this platform"));
return -1;
}
int
virNetDevGetPhysicalFunction(const char *ifname ATTRIBUTE_UNUSED,
char **pfname ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
_("Unable to get physical function status on this platform"));
return -1;
}
int
virNetDevGetVirtualFunctionInfo(const char *vfname ATTRIBUTE_UNUSED,
char **pfname ATTRIBUTE_UNUSED,
int *vf ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
_("Unable to get virtual function info on this platform"));
return -1;
}
#endif /* !__linux__ */
#if defined(__linux__) && defined(HAVE_LIBNL) && defined(IFLA_VF_MAX)
static struct nla_policy ifla_vf_policy[IFLA_VF_MAX+1] = {
[IFLA_VF_MAC] = { .type = NLA_UNSPEC,
.maxlen = sizeof(struct ifla_vf_mac) },
[IFLA_VF_VLAN] = { .type = NLA_UNSPEC,
.maxlen = sizeof(struct ifla_vf_vlan) },
};
/**
* virNetDevLinkDump:
*
util: eliminate "use after free" in callers of virNetDevLinkDump virNetDevLinkDump() gets a message from netlink into "resp", then calls nlmsg_parse() to fill the table "tb" with pointers into resp. It then returns tb to its caller, but not before freeing the buffer at resp. That means that all the callers of virNetDevLinkDump() are examining memory that has already been freed. This can be verified by filling the buffer at resp with garbage prior to freeing it (or, I suppose, just running libvirtd under valgrind) then performing some operation that calls virNetDevLinkDump(). The code has been like this ever since virNetDevLinkDump() was written - the original author didn't notice it, and neither did later additional users of the function. It has only been pure luck (or maybe a lack of heavy load, and/or maybe an allocation algorithm in malloc() that delays re-use of just-freed memory) that has kept this from causing errors, for example when configuring a PCI passthrough or macvtap passthrough network interface. The solution taken in this patch is the simplest - just return resp to the caller along with tb, then have the caller free it after they are finished using the data (pointers) in tb. I alternately could have made a cleaner interface by creating a new struct that put tb and resp together along with a vir*Free() function for it, but this function is only used in a couple places, and I'm not sure there will be additional new uses of virNetDevLinkDump(), so the value of adding a new type, extra APIs, etc. is dubious.
2014-10-15 22:49:01 +00:00
* @ifname: The name of the interface; only use if ifindex <= 0
* @ifindex: The interface index; may be <= 0 if ifname is given
* @data: Gets a pointer to the raw data from netlink.
MUST BE FREED BY CALLER!
* @nlattr: Pointer to a pointer of netlink attributes that will contain
* the results
util: set src_pid for virNetlinkCommand when appropriate Until now, the nl_pid of the source address of every message sent by virNetlinkCommand has been set to the value of getpid(). Most of the time this doesn't matter, and in the one case where it does (communication with lldpad), it previously was the proper thing to do, because the netlink event service (which listens on a netlink socket for unsolicited messages from lldpad) coincidentally always happened to bind with a local nl_pid == getpid(). With the fix for: https://bugzilla.redhat.com/show_bug.cgi?id=816465 that particular nl_pid is now effectively a reserved value, so the netlink event service will always bind to something else (coincidentally "getpid() + (1 << 22)", but it really could be anything). The result is that communication between lldpad and libvirtd is broken (lldpad gets a "disconnected" error when it tries to send a directed message). The solution to this problem caused by a solution, is to query the netlink event service's nlhandle for its "local_port", and send that as the source nl_pid (but only when sending to lldpad, of course - in other cases we maintain the old behavior of sending getpid()). There are two cases where a message is being directed at lldpad - one in virNetDevLinkDump, and one in virNetDevVPortProfileOpSetLink. The case of virNetDevVPortProfileOpSetLink is simplest to explain - only if !nltarget_kernel, i.e. the message isn't targetted for the kernel, is the dst_pid set (by calling virNetDevVPortProfileGetLldpadPid()), so only in that case do we call virNetlinkEventServiceLocalPid() to set src_pid. For virNetDevLinkDump, it's a bit more complicated. The call to virNetDevVPortProfileGetLldpadPid() was effectively up one level (in virNetDevVPortProfileOpCommon), although obscured by an unnecessary passing of a function pointer. This patch removes the function pointer, and calls virNetDevVPortProfileGetLldpadPid() directly in virNetDevVPortProfileOpCommon - if it's doing this, it knows that it should also call virNetlinkEventServiceLocalPid() to set src_pid too; then it just passes src_pid and dst_pid down to virNetDevLinkDump. Since (src_pid == 0 && dst_pid == 0) implies that the kernel is the destination, there is no longer any need to send nltarget_kernel as an arg to virNetDevLinkDump, so it's been removed. The disparity between src_pid being int and dst_pid being uint32_t may be a bit disconcerting to some, but I didn't want to complicate virNetlinkEventServiceLocalPid() by having status returned separately from the value.
2012-05-04 17:48:20 +00:00
* @src_pid: pid used for nl_pid of the local end of the netlink message
* (0 == "use getpid()")
* @dst_pid: pid of destination nl_pid if the kernel
* is not the target of the netlink message but it is to be
* sent to another process (0 if sending to the kernel)
*
util: eliminate "use after free" in callers of virNetDevLinkDump virNetDevLinkDump() gets a message from netlink into "resp", then calls nlmsg_parse() to fill the table "tb" with pointers into resp. It then returns tb to its caller, but not before freeing the buffer at resp. That means that all the callers of virNetDevLinkDump() are examining memory that has already been freed. This can be verified by filling the buffer at resp with garbage prior to freeing it (or, I suppose, just running libvirtd under valgrind) then performing some operation that calls virNetDevLinkDump(). The code has been like this ever since virNetDevLinkDump() was written - the original author didn't notice it, and neither did later additional users of the function. It has only been pure luck (or maybe a lack of heavy load, and/or maybe an allocation algorithm in malloc() that delays re-use of just-freed memory) that has kept this from causing errors, for example when configuring a PCI passthrough or macvtap passthrough network interface. The solution taken in this patch is the simplest - just return resp to the caller along with tb, then have the caller free it after they are finished using the data (pointers) in tb. I alternately could have made a cleaner interface by creating a new struct that put tb and resp together along with a vir*Free() function for it, but this function is only used in a couple places, and I'm not sure there will be additional new uses of virNetDevLinkDump(), so the value of adding a new type, extra APIs, etc. is dubious.
2014-10-15 22:49:01 +00:00
* Get information from netlink about an interface given its name or index.
*
* Returns 0 on success, -1 on fatal error.
*/
int
virNetDevLinkDump(const char *ifname, int ifindex,
util: eliminate "use after free" in callers of virNetDevLinkDump virNetDevLinkDump() gets a message from netlink into "resp", then calls nlmsg_parse() to fill the table "tb" with pointers into resp. It then returns tb to its caller, but not before freeing the buffer at resp. That means that all the callers of virNetDevLinkDump() are examining memory that has already been freed. This can be verified by filling the buffer at resp with garbage prior to freeing it (or, I suppose, just running libvirtd under valgrind) then performing some operation that calls virNetDevLinkDump(). The code has been like this ever since virNetDevLinkDump() was written - the original author didn't notice it, and neither did later additional users of the function. It has only been pure luck (or maybe a lack of heavy load, and/or maybe an allocation algorithm in malloc() that delays re-use of just-freed memory) that has kept this from causing errors, for example when configuring a PCI passthrough or macvtap passthrough network interface. The solution taken in this patch is the simplest - just return resp to the caller along with tb, then have the caller free it after they are finished using the data (pointers) in tb. I alternately could have made a cleaner interface by creating a new struct that put tb and resp together along with a vir*Free() function for it, but this function is only used in a couple places, and I'm not sure there will be additional new uses of virNetDevLinkDump(), so the value of adding a new type, extra APIs, etc. is dubious.
2014-10-15 22:49:01 +00:00
void **nlData, struct nlattr **tb,
util: set src_pid for virNetlinkCommand when appropriate Until now, the nl_pid of the source address of every message sent by virNetlinkCommand has been set to the value of getpid(). Most of the time this doesn't matter, and in the one case where it does (communication with lldpad), it previously was the proper thing to do, because the netlink event service (which listens on a netlink socket for unsolicited messages from lldpad) coincidentally always happened to bind with a local nl_pid == getpid(). With the fix for: https://bugzilla.redhat.com/show_bug.cgi?id=816465 that particular nl_pid is now effectively a reserved value, so the netlink event service will always bind to something else (coincidentally "getpid() + (1 << 22)", but it really could be anything). The result is that communication between lldpad and libvirtd is broken (lldpad gets a "disconnected" error when it tries to send a directed message). The solution to this problem caused by a solution, is to query the netlink event service's nlhandle for its "local_port", and send that as the source nl_pid (but only when sending to lldpad, of course - in other cases we maintain the old behavior of sending getpid()). There are two cases where a message is being directed at lldpad - one in virNetDevLinkDump, and one in virNetDevVPortProfileOpSetLink. The case of virNetDevVPortProfileOpSetLink is simplest to explain - only if !nltarget_kernel, i.e. the message isn't targetted for the kernel, is the dst_pid set (by calling virNetDevVPortProfileGetLldpadPid()), so only in that case do we call virNetlinkEventServiceLocalPid() to set src_pid. For virNetDevLinkDump, it's a bit more complicated. The call to virNetDevVPortProfileGetLldpadPid() was effectively up one level (in virNetDevVPortProfileOpCommon), although obscured by an unnecessary passing of a function pointer. This patch removes the function pointer, and calls virNetDevVPortProfileGetLldpadPid() directly in virNetDevVPortProfileOpCommon - if it's doing this, it knows that it should also call virNetlinkEventServiceLocalPid() to set src_pid too; then it just passes src_pid and dst_pid down to virNetDevLinkDump. Since (src_pid == 0 && dst_pid == 0) implies that the kernel is the destination, there is no longer any need to send nltarget_kernel as an arg to virNetDevLinkDump, so it's been removed. The disparity between src_pid being int and dst_pid being uint32_t may be a bit disconcerting to some, but I didn't want to complicate virNetlinkEventServiceLocalPid() by having status returned separately from the value.
2012-05-04 17:48:20 +00:00
uint32_t src_pid, uint32_t dst_pid)
{
int rc = -1;
struct nlmsghdr *resp = NULL;
struct nlmsgerr *err;
struct ifinfomsg ifinfo = {
.ifi_family = AF_UNSPEC,
.ifi_index = ifindex
};
unsigned int recvbuflen;
struct nl_msg *nl_msg;
if (ifname && ifindex <= 0 && virNetDevGetIndex(ifname, &ifindex) < 0)
return -1;
ifinfo.ifi_index = ifindex;
nl_msg = nlmsg_alloc_simple(RTM_GETLINK, NLM_F_REQUEST);
if (!nl_msg) {
virReportOOMError();
return -1;
}
if (nlmsg_append(nl_msg, &ifinfo, sizeof(ifinfo), NLMSG_ALIGNTO) < 0)
goto buffer_too_small;
if (ifname) {
if (nla_put(nl_msg, IFLA_IFNAME, strlen(ifname)+1, ifname) < 0)
goto buffer_too_small;
}
util: fix botched check for new netlink request filters This is an adjustment to the fix for https://bugzilla.redhat.com/show_bug.cgi?id=889319 to account for two bonehead mistakes I made. commit ac2797cf2af2fd0e64c58a48409a8175d24d6f86 attempted to fix a problem with netlink in newer kernels requiring an extra attribute with a filter flag set in order to receive an IFLA_VFINFO_LIST from netlink. Unfortunately, the #ifdef that protected against compiling it in on systems without the new flag went a bit too far, assuring that the new code would *never* be compiled, and even if it had, the code was incorrect. The first problem was that, while some IFLA_* enum values are also their existence at compile time, IFLA_EXT_MASK *isn't* #defined, so checking to see if it's #defined is not a valid method of determining whether or not to add the attribute. Fortunately, the flag that is being set (RTEXT_FILTER_VF) *is* #defined, and it is never present if IFLA_EXT_MASK isn't, so it's sufficient to just check for that flag. And to top it off, due to the code not actually compiling when I thought it did, I didn't realize that I'd been given the wrong arglist to nla_put() - you can't just send a const value to nla_put, you have to send it a pointer to memory containing what you want to add to the message, along with the length of that memory. This time I've actually sent the patch over to the other machine that's experiencing the problem, applied it to the branch being used (0.10.2) and verified that it works properly, i.e. it does fix the problem it's supposed to fix. :-/
2012-12-21 20:09:33 +00:00
# ifdef RTEXT_FILTER_VF
util: fix functions that retrieve SRIOV VF info This patch resolves: https://bugzilla.redhat.com/show_bug.cgi?id=889319 When assigning an SRIOV virtual function to a guest using "intelligent PCI passthrough" (<interface type='hostdev'>, which sets the MAC address and vlan tag of the VF before passing its info to qemu), libvirt first learns the current MAC address and vlan tag by sending an NLM_F_REQUEST message for the VF's PF (physical function) to the kernel via a NETLINK_ROUTE socket (see virNetDevLinkDump()); the response message's IFLA_VFINFO_LIST section is examined to extract the info for the particular VF being assigned. This worked fine with kernels up until kernel commit 115c9b81928360d769a76c632bae62d15206a94a (first appearing in upstream kernel 3.3) which changed the ABI to not return IFLA_VFINFO_LIST in the response until a newly introduced IFLA_EXT_MASK field was included in the request, with the (newly introduced, of course) RTEXT_FILTER_VF flag set. The justification for this ABI change was that new fields had been added to the VFINFO, causing NLM_F_REQUEST messages to fail on systems with large numbers of VFs if the requesting application didn't have a large enough buffer for all the info. The idea is that most applications doing an NLM_F_REQUEST don't care about VFINFO anyway, so eliminating it from the response would lower the requirements on buffer size. Apparently, the people who pushed this patch made the mistaken assumption that iproute2 (the "ip" command) was the only package that used IFLA_VFINFO_LIST, so it wouldn't break anything else (and they made sure that iproute2 was fixed. The logic of this "fix" is debatable at best (one could claim that the proper fix would be for the applications in question to be fixed so that they properly sized the buffer, which is what libvirt does (purely by virtue of using libnl), but it is what it is and we have to deal with it. In order for <interface type='hostdev'> to work properly on systems with a kernel 3.3 or later, libvirt needs to add the afore-mentioned IFLA_EXT_MASK field with RTEXT_FILTER_VF set. Of course we also need to continue working on systems with older kernels, so that one bit of code is compiled conditionally. The one time this could cause problems is if the libvirt binary was built on a system without IFLA_EXT_MASK which was subsequently updated to a kernel that *did* have it. That could be solved by manually providing the values of IFLA_EXT_MASK and RTEXT_FILTER_VF and adding it to the message anyway, but I'm uncertain what that might actually do on a system that didn't support the message, so for the time being we'll just fail in that case (which will very likely never happen anyway).
2012-12-20 18:22:17 +00:00
/* if this filter exists in the kernel's netlink implementation,
* we need to set it, otherwise the response message will not
* contain the IFLA_VFINFO_LIST that we're looking for.
*/
util: fix botched check for new netlink request filters This is an adjustment to the fix for https://bugzilla.redhat.com/show_bug.cgi?id=889319 to account for two bonehead mistakes I made. commit ac2797cf2af2fd0e64c58a48409a8175d24d6f86 attempted to fix a problem with netlink in newer kernels requiring an extra attribute with a filter flag set in order to receive an IFLA_VFINFO_LIST from netlink. Unfortunately, the #ifdef that protected against compiling it in on systems without the new flag went a bit too far, assuring that the new code would *never* be compiled, and even if it had, the code was incorrect. The first problem was that, while some IFLA_* enum values are also their existence at compile time, IFLA_EXT_MASK *isn't* #defined, so checking to see if it's #defined is not a valid method of determining whether or not to add the attribute. Fortunately, the flag that is being set (RTEXT_FILTER_VF) *is* #defined, and it is never present if IFLA_EXT_MASK isn't, so it's sufficient to just check for that flag. And to top it off, due to the code not actually compiling when I thought it did, I didn't realize that I'd been given the wrong arglist to nla_put() - you can't just send a const value to nla_put, you have to send it a pointer to memory containing what you want to add to the message, along with the length of that memory. This time I've actually sent the patch over to the other machine that's experiencing the problem, applied it to the branch being used (0.10.2) and verified that it works properly, i.e. it does fix the problem it's supposed to fix. :-/
2012-12-21 20:09:33 +00:00
{
uint32_t ifla_ext_mask = RTEXT_FILTER_VF;
if (nla_put(nl_msg, IFLA_EXT_MASK,
sizeof(ifla_ext_mask), &ifla_ext_mask) < 0) {
goto buffer_too_small;
}
}
util: fix functions that retrieve SRIOV VF info This patch resolves: https://bugzilla.redhat.com/show_bug.cgi?id=889319 When assigning an SRIOV virtual function to a guest using "intelligent PCI passthrough" (<interface type='hostdev'>, which sets the MAC address and vlan tag of the VF before passing its info to qemu), libvirt first learns the current MAC address and vlan tag by sending an NLM_F_REQUEST message for the VF's PF (physical function) to the kernel via a NETLINK_ROUTE socket (see virNetDevLinkDump()); the response message's IFLA_VFINFO_LIST section is examined to extract the info for the particular VF being assigned. This worked fine with kernels up until kernel commit 115c9b81928360d769a76c632bae62d15206a94a (first appearing in upstream kernel 3.3) which changed the ABI to not return IFLA_VFINFO_LIST in the response until a newly introduced IFLA_EXT_MASK field was included in the request, with the (newly introduced, of course) RTEXT_FILTER_VF flag set. The justification for this ABI change was that new fields had been added to the VFINFO, causing NLM_F_REQUEST messages to fail on systems with large numbers of VFs if the requesting application didn't have a large enough buffer for all the info. The idea is that most applications doing an NLM_F_REQUEST don't care about VFINFO anyway, so eliminating it from the response would lower the requirements on buffer size. Apparently, the people who pushed this patch made the mistaken assumption that iproute2 (the "ip" command) was the only package that used IFLA_VFINFO_LIST, so it wouldn't break anything else (and they made sure that iproute2 was fixed. The logic of this "fix" is debatable at best (one could claim that the proper fix would be for the applications in question to be fixed so that they properly sized the buffer, which is what libvirt does (purely by virtue of using libnl), but it is what it is and we have to deal with it. In order for <interface type='hostdev'> to work properly on systems with a kernel 3.3 or later, libvirt needs to add the afore-mentioned IFLA_EXT_MASK field with RTEXT_FILTER_VF set. Of course we also need to continue working on systems with older kernels, so that one bit of code is compiled conditionally. The one time this could cause problems is if the libvirt binary was built on a system without IFLA_EXT_MASK which was subsequently updated to a kernel that *did* have it. That could be solved by manually providing the values of IFLA_EXT_MASK and RTEXT_FILTER_VF and adding it to the message anyway, but I'm uncertain what that might actually do on a system that didn't support the message, so for the time being we'll just fail in that case (which will very likely never happen anyway).
2012-12-20 18:22:17 +00:00
# endif
if (virNetlinkCommand(nl_msg, &resp, &recvbuflen,
src_pid, dst_pid, NETLINK_ROUTE, 0) < 0)
goto cleanup;
if (recvbuflen < NLMSG_LENGTH(0) || resp == NULL)
goto malformed_resp;
switch (resp->nlmsg_type) {
case NLMSG_ERROR:
err = (struct nlmsgerr *)NLMSG_DATA(resp);
if (resp->nlmsg_len < NLMSG_LENGTH(sizeof(*err)))
goto malformed_resp;
if (err->error) {
virReportSystemError(-err->error,
_("error dumping %s (%d) interface"),
ifname, ifindex);
goto cleanup;
}
break;
case GENL_ID_CTRL:
case NLMSG_DONE:
rc = nlmsg_parse(resp, sizeof(struct ifinfomsg),
tb, IFLA_MAX, NULL);
if (rc < 0)
goto malformed_resp;
break;
default:
goto malformed_resp;
}
rc = 0;
cleanup:
nlmsg_free(nl_msg);
util: eliminate "use after free" in callers of virNetDevLinkDump virNetDevLinkDump() gets a message from netlink into "resp", then calls nlmsg_parse() to fill the table "tb" with pointers into resp. It then returns tb to its caller, but not before freeing the buffer at resp. That means that all the callers of virNetDevLinkDump() are examining memory that has already been freed. This can be verified by filling the buffer at resp with garbage prior to freeing it (or, I suppose, just running libvirtd under valgrind) then performing some operation that calls virNetDevLinkDump(). The code has been like this ever since virNetDevLinkDump() was written - the original author didn't notice it, and neither did later additional users of the function. It has only been pure luck (or maybe a lack of heavy load, and/or maybe an allocation algorithm in malloc() that delays re-use of just-freed memory) that has kept this from causing errors, for example when configuring a PCI passthrough or macvtap passthrough network interface. The solution taken in this patch is the simplest - just return resp to the caller along with tb, then have the caller free it after they are finished using the data (pointers) in tb. I alternately could have made a cleaner interface by creating a new struct that put tb and resp together along with a vir*Free() function for it, but this function is only used in a couple places, and I'm not sure there will be additional new uses of virNetDevLinkDump(), so the value of adding a new type, extra APIs, etc. is dubious.
2014-10-15 22:49:01 +00:00
if (rc < 0)
VIR_FREE(resp);
*nlData = resp;
return rc;
malformed_resp:
virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
_("malformed netlink response message"));
goto cleanup;
buffer_too_small:
virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
_("allocated netlink buffer is too small"));
goto cleanup;
}
static int
virNetDevSetVfConfig(const char *ifname, int ifindex, int vf,
maint: avoid 'const fooPtr' in virnet files 'const fooPtr' is the same as 'foo * const' (the pointer won't change, but it's contents can). But in general, if an interface is trying to be const-correct, it should be using 'const foo *' (the pointer is to data that can't be changed). Fix up remaining offenders in src/util. * src/util/virnetdev.h (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Use intended type. * src/util/virnetdevbandwidth.h (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.h (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.h (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.h (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.h (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.h (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate): Likewise. * src/util/virnetlink.h (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. * src/util/virnetdev.c (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Fix fallout. * src/util/virnetdevbandwidth.c (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.c (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.c (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.c (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.c (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.c (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate) (virNetDevVPortProfileOpSetLink, virNetDevVPortProfileOpCommon) (virNetDevVPortProfileOp8021Qbg, virNetDevVPortProfileOp8021Qbh): Likewise. * src/util/virnetlink.c (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. Signed-off-by: Eric Blake <eblake@redhat.com>
2013-10-05 19:41:44 +00:00
bool nltarget_kernel, const virMacAddr *macaddr,
int vlanid, uint32_t (*getPidFunc)(void))
{
int rc = -1;
struct nlmsghdr *resp = NULL;
struct nlmsgerr *err;
unsigned int recvbuflen = 0;
uint32_t pid = 0;
struct nl_msg *nl_msg;
struct nlattr *vfinfolist, *vfinfo;
struct ifinfomsg ifinfo = {
.ifi_family = AF_UNSPEC,
.ifi_index = ifindex
};
if (!macaddr && vlanid < 0)
return -1;
nl_msg = nlmsg_alloc_simple(RTM_SETLINK, NLM_F_REQUEST);
if (!nl_msg) {
virReportOOMError();
return rc;
}
if (nlmsg_append(nl_msg, &ifinfo, sizeof(ifinfo), NLMSG_ALIGNTO) < 0)
goto buffer_too_small;
if (ifname &&
nla_put(nl_msg, IFLA_IFNAME, strlen(ifname)+1, ifname) < 0)
goto buffer_too_small;
if (!(vfinfolist = nla_nest_start(nl_msg, IFLA_VFINFO_LIST)))
goto buffer_too_small;
if (!(vfinfo = nla_nest_start(nl_msg, IFLA_VF_INFO)))
goto buffer_too_small;
if (macaddr) {
struct ifla_vf_mac ifla_vf_mac = {
.vf = vf,
.mac = { 0, },
};
virMacAddrGetRaw(macaddr, ifla_vf_mac.mac);
if (nla_put(nl_msg, IFLA_VF_MAC, sizeof(ifla_vf_mac),
&ifla_vf_mac) < 0)
goto buffer_too_small;
}
if (vlanid >= 0) {
struct ifla_vf_vlan ifla_vf_vlan = {
.vf = vf,
.vlan = vlanid,
.qos = 0,
};
if (nla_put(nl_msg, IFLA_VF_VLAN, sizeof(ifla_vf_vlan),
&ifla_vf_vlan) < 0)
goto buffer_too_small;
}
nla_nest_end(nl_msg, vfinfo);
nla_nest_end(nl_msg, vfinfolist);
if (!nltarget_kernel) {
pid = getPidFunc();
if (pid == 0) {
rc = -1;
goto cleanup;
}
}
if (virNetlinkCommand(nl_msg, &resp, &recvbuflen, 0, pid,
NETLINK_ROUTE, 0) < 0)
goto cleanup;
if (recvbuflen < NLMSG_LENGTH(0) || resp == NULL)
goto malformed_resp;
switch (resp->nlmsg_type) {
case NLMSG_ERROR:
err = (struct nlmsgerr *)NLMSG_DATA(resp);
if (resp->nlmsg_len < NLMSG_LENGTH(sizeof(*err)))
goto malformed_resp;
if (err->error) {
virReportSystemError(-err->error,
_("error during set %s of ifindex %d"),
(macaddr ? (vlanid >= 0 ? "mac/vlan" : "mac") : "vlanid"),
ifindex);
goto cleanup;
}
break;
case NLMSG_DONE:
break;
default:
goto malformed_resp;
}
rc = 0;
cleanup:
nlmsg_free(nl_msg);
VIR_FREE(resp);
return rc;
malformed_resp:
virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
_("malformed netlink response message"));
goto cleanup;
buffer_too_small:
virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
_("allocated netlink buffer is too small"));
goto cleanup;
}
static int
virNetDevParseVfConfig(struct nlattr **tb, int32_t vf, virMacAddrPtr mac,
int *vlanid)
{
int rc = -1;
struct ifla_vf_mac *vf_mac;
struct ifla_vf_vlan *vf_vlan;
struct nlattr *tb_vf_info = {NULL, };
struct nlattr *tb_vf[IFLA_VF_MAX+1];
int rem;
if (!tb[IFLA_VFINFO_LIST]) {
virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
_("missing IFLA_VF_INFO in netlink response"));
goto cleanup;
}
nla_for_each_nested(tb_vf_info, tb[IFLA_VFINFO_LIST], rem) {
if (nla_type(tb_vf_info) != IFLA_VF_INFO)
continue;
if (nla_parse_nested(tb_vf, IFLA_VF_MAX, tb_vf_info,
ifla_vf_policy)) {
virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
_("error parsing IFLA_VF_INFO"));
goto cleanup;
}
if (tb[IFLA_VF_MAC]) {
vf_mac = RTA_DATA(tb_vf[IFLA_VF_MAC]);
if (vf_mac && vf_mac->vf == vf) {
virMacAddrSetRaw(mac, vf_mac->mac);
2013-05-24 10:29:28 +00:00
rc = 0;
}
}
if (tb[IFLA_VF_VLAN]) {
vf_vlan = RTA_DATA(tb_vf[IFLA_VF_VLAN]);
if (vf_vlan && vf_vlan->vf == vf) {
*vlanid = vf_vlan->vlan;
2013-05-24 10:29:28 +00:00
rc = 0;
}
}
2013-05-24 10:29:28 +00:00
if (rc == 0)
break;
}
2013-05-24 10:29:28 +00:00
if (rc < 0)
virReportError(VIR_ERR_INTERNAL_ERROR,
_("couldn't find IFLA_VF_INFO for VF %d "
"in netlink response"), vf);
cleanup:
return rc;
}
static int
virNetDevGetVfConfig(const char *ifname, int vf, virMacAddrPtr mac,
int *vlanid)
{
int rc = -1;
util: eliminate "use after free" in callers of virNetDevLinkDump virNetDevLinkDump() gets a message from netlink into "resp", then calls nlmsg_parse() to fill the table "tb" with pointers into resp. It then returns tb to its caller, but not before freeing the buffer at resp. That means that all the callers of virNetDevLinkDump() are examining memory that has already been freed. This can be verified by filling the buffer at resp with garbage prior to freeing it (or, I suppose, just running libvirtd under valgrind) then performing some operation that calls virNetDevLinkDump(). The code has been like this ever since virNetDevLinkDump() was written - the original author didn't notice it, and neither did later additional users of the function. It has only been pure luck (or maybe a lack of heavy load, and/or maybe an allocation algorithm in malloc() that delays re-use of just-freed memory) that has kept this from causing errors, for example when configuring a PCI passthrough or macvtap passthrough network interface. The solution taken in this patch is the simplest - just return resp to the caller along with tb, then have the caller free it after they are finished using the data (pointers) in tb. I alternately could have made a cleaner interface by creating a new struct that put tb and resp together along with a vir*Free() function for it, but this function is only used in a couple places, and I'm not sure there will be additional new uses of virNetDevLinkDump(), so the value of adding a new type, extra APIs, etc. is dubious.
2014-10-15 22:49:01 +00:00
void *nlData = NULL;
struct nlattr *tb[IFLA_MAX + 1] = {NULL, };
int ifindex = -1;
util: eliminate "use after free" in callers of virNetDevLinkDump virNetDevLinkDump() gets a message from netlink into "resp", then calls nlmsg_parse() to fill the table "tb" with pointers into resp. It then returns tb to its caller, but not before freeing the buffer at resp. That means that all the callers of virNetDevLinkDump() are examining memory that has already been freed. This can be verified by filling the buffer at resp with garbage prior to freeing it (or, I suppose, just running libvirtd under valgrind) then performing some operation that calls virNetDevLinkDump(). The code has been like this ever since virNetDevLinkDump() was written - the original author didn't notice it, and neither did later additional users of the function. It has only been pure luck (or maybe a lack of heavy load, and/or maybe an allocation algorithm in malloc() that delays re-use of just-freed memory) that has kept this from causing errors, for example when configuring a PCI passthrough or macvtap passthrough network interface. The solution taken in this patch is the simplest - just return resp to the caller along with tb, then have the caller free it after they are finished using the data (pointers) in tb. I alternately could have made a cleaner interface by creating a new struct that put tb and resp together along with a vir*Free() function for it, but this function is only used in a couple places, and I'm not sure there will be additional new uses of virNetDevLinkDump(), so the value of adding a new type, extra APIs, etc. is dubious.
2014-10-15 22:49:01 +00:00
rc = virNetDevLinkDump(ifname, ifindex, &nlData, tb, 0, 0);
if (rc < 0)
util: eliminate "use after free" in callers of virNetDevLinkDump virNetDevLinkDump() gets a message from netlink into "resp", then calls nlmsg_parse() to fill the table "tb" with pointers into resp. It then returns tb to its caller, but not before freeing the buffer at resp. That means that all the callers of virNetDevLinkDump() are examining memory that has already been freed. This can be verified by filling the buffer at resp with garbage prior to freeing it (or, I suppose, just running libvirtd under valgrind) then performing some operation that calls virNetDevLinkDump(). The code has been like this ever since virNetDevLinkDump() was written - the original author didn't notice it, and neither did later additional users of the function. It has only been pure luck (or maybe a lack of heavy load, and/or maybe an allocation algorithm in malloc() that delays re-use of just-freed memory) that has kept this from causing errors, for example when configuring a PCI passthrough or macvtap passthrough network interface. The solution taken in this patch is the simplest - just return resp to the caller along with tb, then have the caller free it after they are finished using the data (pointers) in tb. I alternately could have made a cleaner interface by creating a new struct that put tb and resp together along with a vir*Free() function for it, but this function is only used in a couple places, and I'm not sure there will be additional new uses of virNetDevLinkDump(), so the value of adding a new type, extra APIs, etc. is dubious.
2014-10-15 22:49:01 +00:00
goto cleanup;
rc = virNetDevParseVfConfig(tb, vf, mac, vlanid);
util: eliminate "use after free" in callers of virNetDevLinkDump virNetDevLinkDump() gets a message from netlink into "resp", then calls nlmsg_parse() to fill the table "tb" with pointers into resp. It then returns tb to its caller, but not before freeing the buffer at resp. That means that all the callers of virNetDevLinkDump() are examining memory that has already been freed. This can be verified by filling the buffer at resp with garbage prior to freeing it (or, I suppose, just running libvirtd under valgrind) then performing some operation that calls virNetDevLinkDump(). The code has been like this ever since virNetDevLinkDump() was written - the original author didn't notice it, and neither did later additional users of the function. It has only been pure luck (or maybe a lack of heavy load, and/or maybe an allocation algorithm in malloc() that delays re-use of just-freed memory) that has kept this from causing errors, for example when configuring a PCI passthrough or macvtap passthrough network interface. The solution taken in this patch is the simplest - just return resp to the caller along with tb, then have the caller free it after they are finished using the data (pointers) in tb. I alternately could have made a cleaner interface by creating a new struct that put tb and resp together along with a vir*Free() function for it, but this function is only used in a couple places, and I'm not sure there will be additional new uses of virNetDevLinkDump(), so the value of adding a new type, extra APIs, etc. is dubious.
2014-10-15 22:49:01 +00:00
cleanup:
VIR_FREE(nlData);
return rc;
}
static int
virNetDevReplaceVfConfig(const char *pflinkdev, int vf,
maint: avoid 'const fooPtr' in virnet files 'const fooPtr' is the same as 'foo * const' (the pointer won't change, but it's contents can). But in general, if an interface is trying to be const-correct, it should be using 'const foo *' (the pointer is to data that can't be changed). Fix up remaining offenders in src/util. * src/util/virnetdev.h (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Use intended type. * src/util/virnetdevbandwidth.h (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.h (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.h (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.h (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.h (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.h (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate): Likewise. * src/util/virnetlink.h (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. * src/util/virnetdev.c (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Fix fallout. * src/util/virnetdevbandwidth.c (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.c (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.c (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.c (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.c (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.c (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate) (virNetDevVPortProfileOpSetLink, virNetDevVPortProfileOpCommon) (virNetDevVPortProfileOp8021Qbg, virNetDevVPortProfileOp8021Qbh): Likewise. * src/util/virnetlink.c (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. Signed-off-by: Eric Blake <eblake@redhat.com>
2013-10-05 19:41:44 +00:00
const virMacAddr *macaddress,
int vlanid,
const char *stateDir)
{
int ret = -1;
virMacAddr oldmac;
int oldvlanid = -1;
char *path = NULL;
char macstr[VIR_MAC_STRING_BUFLEN];
char *fileData = NULL;
int ifindex = -1;
if (virNetDevGetVfConfig(pflinkdev, vf, &oldmac, &oldvlanid) < 0)
goto cleanup;
if (virAsprintf(&path, "%s/%s_vf%d",
stateDir, pflinkdev, vf) < 0)
goto cleanup;
if (virAsprintf(&fileData, "%s\n%d\n",
virMacAddrFormat(&oldmac, macstr), oldvlanid) < 0)
goto cleanup;
if (virFileWriteStr(path, fileData, O_CREAT|O_TRUNC|O_WRONLY) < 0) {
virReportSystemError(errno, _("Unable to preserve mac/vlan tag "
"for pf = %s, vf = %d"), pflinkdev, vf);
goto cleanup;
}
ret = virNetDevSetVfConfig(pflinkdev, ifindex, vf, true,
macaddress, vlanid, NULL);
cleanup:
VIR_FREE(path);
VIR_FREE(fileData);
return ret;
}
static int
virNetDevRestoreVfConfig(const char *pflinkdev, int vf,
const char *stateDir)
{
int rc = -1;
char *path = NULL;
char *fileData = NULL;
char *vlan = NULL;
virMacAddr oldmac;
int vlanid = -1;
int ifindex = -1;
if (virAsprintf(&path, "%s/%s_vf%d",
stateDir, pflinkdev, vf) < 0)
return rc;
if (virFileReadAll(path, 128, &fileData) < 0)
goto cleanup;
if ((vlan = strchr(fileData, '\n'))) {
char *endptr;
*vlan++ = 0; /* NULL terminate the mac address */
if (*vlan) {
if ((virStrToLong_i(vlan, &endptr, 10, &vlanid) < 0) ||
(endptr && *endptr != '\n' && *endptr != 0)) {
virReportError(VIR_ERR_INTERNAL_ERROR,
_("Cannot parse vlan tag from '%s'"),
vlan);
goto cleanup;
}
}
}
if (virMacAddrParse(fileData, &oldmac) != 0) {
virReportError(VIR_ERR_INTERNAL_ERROR,
_("Cannot parse MAC address from '%s'"),
fileData);
goto cleanup;
}
/*reset mac and remove file-ignore results*/
rc = virNetDevSetVfConfig(pflinkdev, ifindex, vf, true,
&oldmac, vlanid, NULL);
ignore_value(unlink(path));
cleanup:
VIR_FREE(path);
VIR_FREE(fileData);
return rc;
}
/**
* virNetDevReplaceNetConfig:
* @linkdev: name of the interface
* @vf: vf index if linkdev is a pf
* @macaddress: new MAC address for interface
* @vlanid: new vlanid
* @stateDir: directory to store old net config
*
* Returns 0 on success, -1 on failure
*
*/
int
virNetDevReplaceNetConfig(const char *linkdev, int vf,
maint: avoid 'const fooPtr' in virnet files 'const fooPtr' is the same as 'foo * const' (the pointer won't change, but it's contents can). But in general, if an interface is trying to be const-correct, it should be using 'const foo *' (the pointer is to data that can't be changed). Fix up remaining offenders in src/util. * src/util/virnetdev.h (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Use intended type. * src/util/virnetdevbandwidth.h (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.h (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.h (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.h (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.h (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.h (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate): Likewise. * src/util/virnetlink.h (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. * src/util/virnetdev.c (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Fix fallout. * src/util/virnetdevbandwidth.c (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.c (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.c (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.c (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.c (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.c (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate) (virNetDevVPortProfileOpSetLink, virNetDevVPortProfileOpCommon) (virNetDevVPortProfileOp8021Qbg, virNetDevVPortProfileOp8021Qbh): Likewise. * src/util/virnetlink.c (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. Signed-off-by: Eric Blake <eblake@redhat.com>
2013-10-05 19:41:44 +00:00
const virMacAddr *macaddress, int vlanid,
const char *stateDir)
{
if (vf == -1)
return virNetDevReplaceMacAddress(linkdev, macaddress, stateDir);
else
return virNetDevReplaceVfConfig(linkdev, vf, macaddress, vlanid,
stateDir);
}
/**
* virNetDevRestoreNetConfig:
* @linkdev: name of the interface
* @vf: vf index if linkdev is a pf
* @stateDir: directory containing old net config
*
* Returns 0 on success, -errno on failure.
*
*/
int
virNetDevRestoreNetConfig(const char *linkdev, int vf, const char *stateDir)
{
if (vf == -1)
return virNetDevRestoreMacAddress(linkdev, stateDir);
else
return virNetDevRestoreVfConfig(linkdev, vf, stateDir);
}
#else /* defined(__linux__) && defined(HAVE_LIBNL) */
int
virNetDevLinkDump(const char *ifname ATTRIBUTE_UNUSED,
int ifindex ATTRIBUTE_UNUSED,
util: eliminate "use after free" in callers of virNetDevLinkDump virNetDevLinkDump() gets a message from netlink into "resp", then calls nlmsg_parse() to fill the table "tb" with pointers into resp. It then returns tb to its caller, but not before freeing the buffer at resp. That means that all the callers of virNetDevLinkDump() are examining memory that has already been freed. This can be verified by filling the buffer at resp with garbage prior to freeing it (or, I suppose, just running libvirtd under valgrind) then performing some operation that calls virNetDevLinkDump(). The code has been like this ever since virNetDevLinkDump() was written - the original author didn't notice it, and neither did later additional users of the function. It has only been pure luck (or maybe a lack of heavy load, and/or maybe an allocation algorithm in malloc() that delays re-use of just-freed memory) that has kept this from causing errors, for example when configuring a PCI passthrough or macvtap passthrough network interface. The solution taken in this patch is the simplest - just return resp to the caller along with tb, then have the caller free it after they are finished using the data (pointers) in tb. I alternately could have made a cleaner interface by creating a new struct that put tb and resp together along with a vir*Free() function for it, but this function is only used in a couple places, and I'm not sure there will be additional new uses of virNetDevLinkDump(), so the value of adding a new type, extra APIs, etc. is dubious.
2014-10-15 22:49:01 +00:00
void **nlData ATTRIBUTE_UNUSED,
struct nlattr **tb ATTRIBUTE_UNUSED,
util: set src_pid for virNetlinkCommand when appropriate Until now, the nl_pid of the source address of every message sent by virNetlinkCommand has been set to the value of getpid(). Most of the time this doesn't matter, and in the one case where it does (communication with lldpad), it previously was the proper thing to do, because the netlink event service (which listens on a netlink socket for unsolicited messages from lldpad) coincidentally always happened to bind with a local nl_pid == getpid(). With the fix for: https://bugzilla.redhat.com/show_bug.cgi?id=816465 that particular nl_pid is now effectively a reserved value, so the netlink event service will always bind to something else (coincidentally "getpid() + (1 << 22)", but it really could be anything). The result is that communication between lldpad and libvirtd is broken (lldpad gets a "disconnected" error when it tries to send a directed message). The solution to this problem caused by a solution, is to query the netlink event service's nlhandle for its "local_port", and send that as the source nl_pid (but only when sending to lldpad, of course - in other cases we maintain the old behavior of sending getpid()). There are two cases where a message is being directed at lldpad - one in virNetDevLinkDump, and one in virNetDevVPortProfileOpSetLink. The case of virNetDevVPortProfileOpSetLink is simplest to explain - only if !nltarget_kernel, i.e. the message isn't targetted for the kernel, is the dst_pid set (by calling virNetDevVPortProfileGetLldpadPid()), so only in that case do we call virNetlinkEventServiceLocalPid() to set src_pid. For virNetDevLinkDump, it's a bit more complicated. The call to virNetDevVPortProfileGetLldpadPid() was effectively up one level (in virNetDevVPortProfileOpCommon), although obscured by an unnecessary passing of a function pointer. This patch removes the function pointer, and calls virNetDevVPortProfileGetLldpadPid() directly in virNetDevVPortProfileOpCommon - if it's doing this, it knows that it should also call virNetlinkEventServiceLocalPid() to set src_pid too; then it just passes src_pid and dst_pid down to virNetDevLinkDump. Since (src_pid == 0 && dst_pid == 0) implies that the kernel is the destination, there is no longer any need to send nltarget_kernel as an arg to virNetDevLinkDump, so it's been removed. The disparity between src_pid being int and dst_pid being uint32_t may be a bit disconcerting to some, but I didn't want to complicate virNetlinkEventServiceLocalPid() by having status returned separately from the value.
2012-05-04 17:48:20 +00:00
uint32_t src_pid ATTRIBUTE_UNUSED,
uint32_t dst_pid ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
_("Unable to dump link info on this platform"));
return -1;
}
int
virNetDevReplaceNetConfig(const char *linkdev ATTRIBUTE_UNUSED,
int vf ATTRIBUTE_UNUSED,
maint: avoid 'const fooPtr' in virnet files 'const fooPtr' is the same as 'foo * const' (the pointer won't change, but it's contents can). But in general, if an interface is trying to be const-correct, it should be using 'const foo *' (the pointer is to data that can't be changed). Fix up remaining offenders in src/util. * src/util/virnetdev.h (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Use intended type. * src/util/virnetdevbandwidth.h (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.h (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.h (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.h (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.h (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.h (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate): Likewise. * src/util/virnetlink.h (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. * src/util/virnetdev.c (virNetDevSetMAC) (virNetDevReplaceMacAddress, virNetDevValidateConfig) (virNetDevReplaceNetConfig): Fix fallout. * src/util/virnetdevbandwidth.c (virNetDevBandwidthCopy) (virNetDevBandwidthPlug): Likewise. * src/util/virnetdevmacvlan.c (virNetDevMacVLanCreate) (virNetDevMacVLanCreateWithVPortProfile) (virNetDevMacVLanDeleteWithVPortProfile) (virNetDevMacVLanRestartWithVPortProfile) (virNetDevMacVLanVPortProfileRegisterCallback): Likewise. * src/util/virnetdevopenvswitch.c (virNetDevOpenvswitchAddPort): Likewise. * src/util/virnetdevtap.c (virNetDevTapCreateInBridgePort): Likewise. * src/util/virnetdevvlan.c (virNetDevVlanEqual) (virNetDevVlanCopy): Likewise. * src/util/virnetdevvportprofile.c (virNetDevVPortProfileAssociate) (virNetDevVPortProfileDisassociate) (virNetDevVPortProfileOpSetLink, virNetDevVPortProfileOpCommon) (virNetDevVPortProfileOp8021Qbg, virNetDevVPortProfileOp8021Qbh): Likewise. * src/util/virnetlink.c (virNetlinkEventRemoveCallback) (virNetlinkEventAddClient, virNetlinkEventRemoveClient): Likewise. Signed-off-by: Eric Blake <eblake@redhat.com>
2013-10-05 19:41:44 +00:00
const virMacAddr *macaddress ATTRIBUTE_UNUSED,
int vlanid ATTRIBUTE_UNUSED,
const char *stateDir ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
_("Unable to replace net config on this platform"));
return -1;
}
int
virNetDevRestoreNetConfig(const char *linkdev ATTRIBUTE_UNUSED,
int vf ATTRIBUTE_UNUSED,
const char *stateDir ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
_("Unable to restore net config on this platform"));
return -1;
}
#endif /* defined(__linux__) && defined(HAVE_LIBNL) */
#ifdef __linux__
int
virNetDevGetLinkInfo(const char *ifname,
virInterfaceLinkPtr lnk)
{
int ret = -1;
char *path = NULL;
char *buf = NULL;
char *tmp;
int tmp_state;
unsigned int tmp_speed;
if (virNetDevSysfsFile(&path, ifname, "operstate") < 0)
goto cleanup;
if (virFileReadAll(path, 1024, &buf) < 0) {
virReportSystemError(errno,
_("unable to read: %s"),
path);
goto cleanup;
}
if (!(tmp = strchr(buf, '\n'))) {
virReportError(VIR_ERR_INTERNAL_ERROR,
_("Unable to parse: %s"),
buf);
goto cleanup;
}
*tmp = '\0';
/* We shouldn't allow 0 here, because
* virInterfaceState enum starts from 1. */
if ((tmp_state = virInterfaceStateTypeFromString(buf)) <= 0) {
virReportError(VIR_ERR_INTERNAL_ERROR,
_("Unable to parse: %s"),
buf);
goto cleanup;
}
lnk->state = tmp_state;
/* Shortcut to avoid some kernel issues. If link is not up several drivers
* report several misleading values. While igb reports 65535, realtek goes
* with 10. To avoid muddying XML with insane values, don't report link
* speed if that's the case. */
if (lnk->state != VIR_INTERFACE_STATE_UP) {
lnk->speed = 0;
ret = 0;
goto cleanup;
}
VIR_FREE(path);
VIR_FREE(buf);
if (virNetDevSysfsFile(&path, ifname, "speed") < 0)
goto cleanup;
if (virFileReadAllQuiet(path, 1024, &buf) < 0) {
/* Some devices doesn't report speed, in which case we get EINVAL */
if (errno == EINVAL) {
ret = 0;
goto cleanup;
}
virReportSystemError(errno,
_("unable to read: %s"),
path);
goto cleanup;
}
if (virStrToLong_ui(buf, &tmp, 10, &tmp_speed) < 0 ||
*tmp != '\n') {
virReportError(VIR_ERR_INTERNAL_ERROR,
_("Unable to parse: %s"),
buf);
goto cleanup;
}
lnk->speed = tmp_speed;
ret = 0;
cleanup:
VIR_FREE(buf);
VIR_FREE(path);
return ret;
}
#else
int
virNetDevGetLinkInfo(const char *ifname,
virInterfaceLinkPtr lnk)
{
/* Port me */
VIR_DEBUG("Getting link info on %s is not implemented on this platform",
ifname);
lnk->speed = lnk->state = 0;
return 0;
}
#endif /* defined(__linux__) */
#if defined(SIOCADDMULTI) && defined(HAVE_STRUCT_IFREQ) && \
defined(HAVE_STRUCT_IFREQ_IFR_HWADDR)
/**
* virNetDevAddMulti:
* @ifname: interface name to which to add multicast MAC address
* @macaddr: MAC address
*
* This function adds the @macaddr to the multicast list for a given interface
* @ifname.
*
* Returns 0 in case of success or -1 on failure
*/
int virNetDevAddMulti(const char *ifname,
virMacAddrPtr macaddr)
{
int fd = -1;
int ret = -1;
struct ifreq ifr;
if ((fd = virNetDevSetupControl(ifname, &ifr)) < 0)
return -1;
ifr.ifr_hwaddr.sa_family = AF_UNSPEC;
virMacAddrGetRaw(macaddr, (unsigned char *)ifr.ifr_hwaddr.sa_data);
if (ioctl(fd, SIOCADDMULTI, &ifr) < 0) {
char macstr[VIR_MAC_STRING_BUFLEN];
virReportSystemError(errno,
_("Cannot add multicast MAC %s on '%s' interface"),
virMacAddrFormat(macaddr, macstr), ifname);
goto cleanup;
}
ret = 0;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#else
int virNetDevAddMulti(const char *ifname ATTRIBUTE_UNUSED,
virMacAddrPtr macaddr ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
_("Unable to add address to interface "
"multicast list on this platform"));
return -1;
}
#endif
#if defined(SIOCDELMULTI) && defined(HAVE_STRUCT_IFREQ) && \
defined(HAVE_STRUCT_IFREQ_IFR_HWADDR)
/**
* virNetDevDelMulti:
* @ifname: interface name from which to delete the multicast MAC address
* @macaddr: MAC address
*
* This function deletes the @macaddr from the multicast list for a given
* interface @ifname.
*
* Returns 0 in case of success or -1 on failure
*/
int virNetDevDelMulti(const char *ifname,
virMacAddrPtr macaddr)
{
int fd = -1;
int ret = -1;
struct ifreq ifr;
if ((fd = virNetDevSetupControl(ifname, &ifr)) < 0)
return -1;
ifr.ifr_hwaddr.sa_family = AF_UNSPEC;
virMacAddrGetRaw(macaddr, (unsigned char *)ifr.ifr_hwaddr.sa_data);
if (ioctl(fd, SIOCDELMULTI, &ifr) < 0) {
char macstr[VIR_MAC_STRING_BUFLEN];
virReportSystemError(errno,
_("Cannot add multicast MAC %s on '%s' interface"),
virMacAddrFormat(macaddr, macstr), ifname);
goto cleanup;
}
ret = 0;
cleanup:
VIR_FORCE_CLOSE(fd);
return ret;
}
#else
int virNetDevDelMulti(const char *ifname ATTRIBUTE_UNUSED,
virMacAddrPtr macaddr ATTRIBUTE_UNUSED)
{
virReportSystemError(ENOSYS, "%s",
_("Unable to delete address from interface "
"multicast list on this platform"));
return -1;
}
#endif
static int virNetDevParseMcast(char *buf, virNetDevMcastEntryPtr mcast)
{
int ifindex;
int num;
char *next;
char *token;
char *saveptr;
char *endptr;
for (ifindex = 0, next = buf; ifindex < VIR_MCAST_TYPE_LAST; ifindex++,
next = NULL) {
token = strtok_r(next, VIR_MCAST_TOKEN_DELIMS, &saveptr);
if (token == NULL) {
virReportSystemError(EINVAL,
_("failed to parse multicast address from '%s'"),
buf);
return -1;
}
switch ((virMCastType)ifindex) {
case VIR_MCAST_TYPE_INDEX_TOKEN:
if (virStrToLong_i(token, &endptr, 10, &num) < 0) {
virReportSystemError(EINVAL,
_("Failed to parse interface index from '%s'"),
buf);
return -1;
}
mcast->index = num;
break;
case VIR_MCAST_TYPE_NAME_TOKEN:
if (virStrncpy(mcast->name, token, strlen(token),
VIR_MCAST_NAME_LEN) == NULL) {
virReportSystemError(EINVAL,
_("Failed to parse network device name from '%s'"),
buf);
return -1;
}
break;
case VIR_MCAST_TYPE_USERS_TOKEN:
if (virStrToLong_i(token, &endptr, 10, &num) < 0) {
virReportSystemError(EINVAL,
_("Failed to parse users from '%s'"),
buf);
return -1;
}
mcast->users = num;
break;
case VIR_MCAST_TYPE_GLOBAL_TOKEN:
if (virStrToLong_i(token, &endptr, 10, &num) < 0) {
virReportSystemError(EINVAL,
_("Failed to parse users from '%s'"),
buf);
return -1;
}
mcast->global = num;
break;
case VIR_MCAST_TYPE_ADDR_TOKEN:
if (virMacAddrParseHex((const char*)token,
&mcast->macaddr) < 0) {
virReportSystemError(EINVAL,
_("Failed to parse MAC address from '%s'"),
buf);
}
break;
/* coverity[dead_error_begin] */
case VIR_MCAST_TYPE_LAST:
break;
}
}
return 0;
}
static void virNetDevMcastListClear(virNetDevMcastListPtr mcast)
{
size_t i;
for (i = 0; i < mcast->nentries; i++)
VIR_FREE(mcast->entries[i]);
VIR_FREE(mcast->entries);
mcast->nentries = 0;
}
static int virNetDevGetMcastList(const char *ifname,
virNetDevMcastListPtr mcast)
{
char *cur = NULL;
char *buf = NULL;
char *next = NULL;
int ret = -1, len;
virNetDevMcastEntryPtr entry = NULL;
mcast->entries = NULL;
mcast->nentries = 0;
/* Read entire multicast table into memory */
if ((len = virFileReadAll(PROC_NET_DEV_MCAST, MAX_MCAST_SIZE, &buf)) <= 0)
goto cleanup;
cur = buf;
while (cur) {
if (!entry && VIR_ALLOC(entry) < 0)
goto cleanup;
next = strchr(cur, '\n');
if (next)
next++;
if (virNetDevParseMcast(cur, entry))
goto cleanup;
/* Only return global multicast MAC addresses for
* specified interface */
if (entry->global && STREQ(ifname, entry->name)) {
if (VIR_APPEND_ELEMENT(mcast->entries, mcast->nentries, entry))
goto cleanup;
} else {
memset(entry, 0, sizeof(virNetDevMcastEntry));
}
cur = next && ((next - buf) < len) ? next : NULL;
}
ret = 0;
cleanup:
VIR_FREE(buf);
VIR_FREE(entry);
return ret;
}
VIR_ENUM_IMPL(virNetDevRxFilterMode,
VIR_NETDEV_RX_FILTER_MODE_LAST,
"none",
"normal",
"all");
static int virNetDevGetMulticastTable(const char *ifname,
virNetDevRxFilterPtr filter)
{
size_t i;
int ret = -1;
virNetDevMcastList mcast;
filter->multicast.nTable = 0;
filter->multicast.table = NULL;
if (virNetDevGetMcastList(ifname, &mcast) < 0)
goto cleanup;
if (mcast.nentries > 0) {
if (VIR_ALLOC_N(filter->multicast.table, mcast.nentries) < 0)
goto cleanup;
for (i = 0; i < mcast.nentries; i++) {
virMacAddrSet(&filter->multicast.table[i],
&mcast.entries[i]->macaddr);
}
filter->multicast.nTable = mcast.nentries;
}
ret = 0;
cleanup:
virNetDevMcastListClear(&mcast);
return ret;
}
virNetDevRxFilterPtr
virNetDevRxFilterNew(void)
{
virNetDevRxFilterPtr filter;
if (VIR_ALLOC(filter) < 0)
return NULL;
return filter;
}
void
virNetDevRxFilterFree(virNetDevRxFilterPtr filter)
{
if (filter) {
VIR_FREE(filter->name);
VIR_FREE(filter->unicast.table);
VIR_FREE(filter->multicast.table);
VIR_FREE(filter->vlan.table);
VIR_FREE(filter);
}
}
/**
* virNetDevGetRxFilter:
* This function supplies the RX filter list for a given device interface
*
* @ifname: Name of the interface
* @filter: The RX filter list
*
* Returns 0 or -1 on failure.
*/
int virNetDevGetRxFilter(const char *ifname,
virNetDevRxFilterPtr *filter)
{
int ret = -1;
bool receive = false;
virNetDevRxFilterPtr fil = virNetDevRxFilterNew();
if (!fil)
goto cleanup;
if (virNetDevGetMAC(ifname, &fil->mac))
goto cleanup;
if (virNetDevGetMulticastTable(ifname, fil))
goto cleanup;
if (virNetDevGetPromiscuous(ifname, &fil->promiscuous))
goto cleanup;
if (virNetDevGetRcvAllMulti(ifname, &receive))
goto cleanup;
if (receive) {
fil->multicast.mode = VIR_NETDEV_RX_FILTER_MODE_ALL;
} else {
if (virNetDevGetRcvMulti(ifname, &receive))
goto cleanup;
if (receive)
fil->multicast.mode = VIR_NETDEV_RX_FILTER_MODE_NORMAL;
else
fil->multicast.mode = VIR_NETDEV_RX_FILTER_MODE_NONE;
}
ret = 0;
cleanup:
if (ret < 0) {
virNetDevRxFilterFree(fil);
fil = NULL;
}
*filter = fil;
return ret;
}
#if defined(SIOCETHTOOL) && defined(HAVE_STRUCT_IFREQ)
/**
* virNetDevFeatureAvailable
* This function checks for the availability of a network device feature
*
* @ifname: name of the interface
* @cmd: reference to an ethtool command structure
*
* Returns 0 on success, -1 on failure.
*/
static int
virNetDevFeatureAvailable(const char *ifname, struct ethtool_value *cmd)
{
int ret = -1;
int sock = -1;
virIfreq ifr;
sock = socket(AF_LOCAL, SOCK_DGRAM, 0);
if (sock < 0) {
virReportSystemError(errno, "%s", _("Cannot open control socket"));
goto cleanup;
}
memset(&ifr, 0, sizeof(ifr));
strcpy(ifr.ifr_name, ifname);
ifr.ifr_data = (void*) cmd;
if (ioctl(sock, SIOCETHTOOL, &ifr) != 0) {
switch (errno) {
case EPERM:
VIR_DEBUG("ethtool ioctl: permission denied");
break;
case EINVAL:
VIR_DEBUG("ethtool ioctl: invalid request");
break;
case EOPNOTSUPP:
VIR_DEBUG("ethtool ioctl: request not supported");
break;
default:
virReportSystemError(errno, "%s", _("ethtool ioctl error"));
goto cleanup;
}
}
ret = cmd->data > 0 ? 1: 0;
cleanup:
if (sock)
VIR_FORCE_CLOSE(sock);
return ret;
}
/**
* virNetDevGetFeatures:
* This function gets the nic offloads features available for ifname
*
* @ifname: name of the interface
* @features: network device feature structures
* @nfeatures: number of features available
*
* Returns 0 on success, -1 on failure.
*/
int
virNetDevGetFeatures(const char *ifname,
virBitmapPtr *out)
{
size_t i = -1;
struct ethtool_value cmd = { 0 };
struct elem{
const int cmd;
const virNetDevFeature feat;
};
/* legacy ethtool getters */
struct elem cmds[] = {
{ETHTOOL_GRXCSUM, VIR_NET_DEV_FEAT_GRXCSUM},
{ETHTOOL_GTXCSUM, VIR_NET_DEV_FEAT_GTXCSUM},
{ETHTOOL_GSG, VIR_NET_DEV_FEAT_GSG},
{ETHTOOL_GTSO, VIR_NET_DEV_FEAT_GTSO},
# if HAVE_DECL_ETHTOOL_GGSO
{ETHTOOL_GGSO, VIR_NET_DEV_FEAT_GGSO},
# endif
# if HAVE_DECL_ETHTOOL_GGRO
{ETHTOOL_GGRO, VIR_NET_DEV_FEAT_GGRO},
# endif
};
if (!(*out = virBitmapNew(VIR_NET_DEV_FEAT_LAST)))
return -1;
for (i = 0; i < ARRAY_CARDINALITY(cmds); i++) {
cmd.cmd = cmds[i].cmd;
if (virNetDevFeatureAvailable(ifname, &cmd))
ignore_value(virBitmapSetBit(*out, cmds[i].feat));
}
# if HAVE_DECL_ETHTOOL_GFLAGS
size_t j = -1;
/* ethtool masks */
struct elem flags[] = {
# if HAVE_DECL_ETH_FLAG_LRO
{ETH_FLAG_LRO, VIR_NET_DEV_FEAT_LRO},
# endif
# if HAVE_DECL_ETH_FLAG_TXVLAN
{ETH_FLAG_RXVLAN, VIR_NET_DEV_FEAT_RXVLAN},
{ETH_FLAG_TXVLAN, VIR_NET_DEV_FEAT_TXVLAN},
# endif
# if HAVE_DECL_ETH_FLAG_NTUBLE
{ETH_FLAG_NTUPLE, VIR_NET_DEV_FEAT_NTUPLE},
# endif
# if HAVE_DECL_ETH_FLAG_RXHASH
{ETH_FLAG_RXHASH, VIR_NET_DEV_FEAT_RXHASH},
# endif
};
cmd.cmd = ETHTOOL_GFLAGS;
if (virNetDevFeatureAvailable(ifname, &cmd)) {
for (j = 0; j < ARRAY_CARDINALITY(flags); j++) {
if (cmd.data & flags[j].cmd)
ignore_value(virBitmapSetBit(*out, flags[j].feat));
}
}
# endif
return 0;
}
#else
int
virNetDevGetFeatures(const char *ifname ATTRIBUTE_UNUSED,
virBitmapPtr *out ATTRIBUTE_UNUSED)
{
VIR_DEBUG("Getting network device features on %s is not implemented on this platform",
ifname);
return 0;
}
#endif