/*
* Copyright (C) 2014 The Android Open Source Project
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "RouteController.h"
#include <arpa/inet.h>
#include <errno.h>
#include <fcntl.h>
#include <linux/fib_rules.h>
#include <net/if.h>
#include <sys/stat.h>
#include <private/android_filesystem_config.h>
#include <map>
#include "DummyNetwork.h"
#include "Fwmark.h"
#include "NetdConstants.h"
#include "NetlinkCommands.h"
#include "UidRanges.h"
#include "android-base/file.h"
#include <android-base/stringprintf.h>
#define LOG_TAG "Netd"
#include "log/log.h"
#include "logwrap/logwrap.h"
#include "netutils/ifc.h"
#include "resolv_netid.h"
using android::base::StringPrintf;
using android::base::WriteStringToFile;
using android::net::UidRange;
namespace android {
namespace net {
auto RouteController::iptablesRestoreCommandFunction = execIptablesRestoreCommand;
// BEGIN CONSTANTS --------------------------------------------------------------------------------
const uint32_t RULE_PRIORITY_VPN_OVERRIDE_SYSTEM = 10000;
const uint32_t RULE_PRIORITY_VPN_OVERRIDE_OIF = 10500;
const uint32_t RULE_PRIORITY_VPN_OUTPUT_TO_LOCAL = 11000;
const uint32_t RULE_PRIORITY_SECURE_VPN = 12000;
const uint32_t RULE_PRIORITY_PROHIBIT_NON_VPN = 12500;
const uint32_t RULE_PRIORITY_EXPLICIT_NETWORK = 13000;
const uint32_t RULE_PRIORITY_OUTPUT_INTERFACE = 14000;
const uint32_t RULE_PRIORITY_LEGACY_SYSTEM = 15000;
const uint32_t RULE_PRIORITY_LEGACY_NETWORK = 16000;
const uint32_t RULE_PRIORITY_LOCAL_NETWORK = 17000;
const uint32_t RULE_PRIORITY_TETHERING = 18000;
const uint32_t RULE_PRIORITY_IMPLICIT_NETWORK = 19000;
const uint32_t RULE_PRIORITY_BYPASSABLE_VPN = 20000;
const uint32_t RULE_PRIORITY_VPN_FALLTHROUGH = 21000;
const uint32_t RULE_PRIORITY_DEFAULT_NETWORK = 22000;
const uint32_t RULE_PRIORITY_UNREACHABLE = 32000;
const uint32_t ROUTE_TABLE_LOCAL_NETWORK = 97;
const uint32_t ROUTE_TABLE_LEGACY_NETWORK = 98;
const uint32_t ROUTE_TABLE_LEGACY_SYSTEM = 99;
const char* const ROUTE_TABLE_NAME_LOCAL_NETWORK = "local_network";
const char* const ROUTE_TABLE_NAME_LEGACY_NETWORK = "legacy_network";
const char* const ROUTE_TABLE_NAME_LEGACY_SYSTEM = "legacy_system";
const char* const ROUTE_TABLE_NAME_LOCAL = "local";
const char* const ROUTE_TABLE_NAME_MAIN = "main";
// None of our regular routes specify priority, which causes them to have the default priority.
// For default throw routes, we use a fixed priority of 100000.
uint32_t PRIO_THROW = 100000;
const char* const RouteController::LOCAL_MANGLE_INPUT = "routectrl_mangle_INPUT";
const uint8_t AF_FAMILIES[] = {AF_INET, AF_INET6};
const uid_t UID_ROOT = 0;
const uint32_t FWMARK_NONE = 0;
const uint32_t MASK_NONE = 0;
const char* const IIF_LOOPBACK = "lo";
const char* const IIF_NONE = NULL;
const char* const OIF_NONE = NULL;
const bool ACTION_ADD = true;
const bool ACTION_DEL = false;
const bool MODIFY_NON_UID_BASED_RULES = true;
const char* const RT_TABLES_PATH = "/data/misc/net/rt_tables";
const mode_t RT_TABLES_MODE = S_IRUSR | S_IWUSR | S_IRGRP | S_IROTH; // mode 0644, rw-r--r--
// Avoids "non-constant-expression cannot be narrowed from type 'unsigned int' to 'unsigned short'"
// warnings when using RTA_LENGTH(x) inside static initializers (even when x is already uint16_t).
constexpr uint16_t U16_RTA_LENGTH(uint16_t x) {
return RTA_LENGTH(x);
}
// These are practically const, but can't be declared so, because they are used to initialize
// non-const pointers ("void* iov_base") in iovec arrays.
rtattr FRATTR_PRIORITY = { U16_RTA_LENGTH(sizeof(uint32_t)), FRA_PRIORITY };
rtattr FRATTR_TABLE = { U16_RTA_LENGTH(sizeof(uint32_t)), FRA_TABLE };
rtattr FRATTR_FWMARK = { U16_RTA_LENGTH(sizeof(uint32_t)), FRA_FWMARK };
rtattr FRATTR_FWMASK = { U16_RTA_LENGTH(sizeof(uint32_t)), FRA_FWMASK };
rtattr FRATTR_UID_RANGE = { U16_RTA_LENGTH(sizeof(fib_rule_uid_range)), FRA_UID_RANGE };
rtattr RTATTR_TABLE = { U16_RTA_LENGTH(sizeof(uint32_t)), RTA_TABLE };
rtattr RTATTR_OIF = { U16_RTA_LENGTH(sizeof(uint32_t)), RTA_OIF };
rtattr RTATTR_PRIO = { U16_RTA_LENGTH(sizeof(uint32_t)), RTA_PRIORITY };
uint8_t PADDING_BUFFER[RTA_ALIGNTO] = {0, 0, 0, 0};
// END CONSTANTS ----------------------------------------------------------------------------------
const char *actionName(uint16_t action) {
static const char *ops[4] = {"adding", "deleting", "getting", "???"};
return ops[action % 4];
}
const char *familyName(uint8_t family) {
switch (family) {
case AF_INET: return "IPv4";
case AF_INET6: return "IPv6";
default: return "???";
}
}
// Caller must hold sInterfaceToTableLock.
uint32_t RouteController::getRouteTableForInterfaceLocked(const char* interface) {
uint32_t index = if_nametoindex(interface);
if (index) {
index += RouteController::ROUTE_TABLE_OFFSET_FROM_INDEX;
sInterfaceToTable[interface] = index;
return index;
}
// If the interface goes away if_nametoindex() will return 0 but we still need to know
// the index so we can remove the rules and routes.
auto iter = sInterfaceToTable.find(interface);
if (iter == sInterfaceToTable.end()) {
ALOGE("cannot find interface %s", interface);
return RT_TABLE_UNSPEC;
}
return iter->second;
}
uint32_t RouteController::getIfIndex(const char* interface) {
android::RWLock::AutoRLock lock(sInterfaceToTableLock);
auto iter = sInterfaceToTable.find(interface);
if (iter == sInterfaceToTable.end()) {
ALOGE("getIfIndex: cannot find interface %s", interface);
return 0;
}
return iter->second - ROUTE_TABLE_OFFSET_FROM_INDEX;
}
uint32_t RouteController::getRouteTableForInterface(const char* interface) {
android::RWLock::AutoRLock lock(sInterfaceToTableLock);
return getRouteTableForInterfaceLocked(interface);
}
void addTableName(uint32_t table, const std::string& name, std::string* contents) {
char tableString[UINT32_STRLEN];
snprintf(tableString, sizeof(tableString), "%u", table);
*contents += tableString;
*contents += " ";
*contents += name;
*contents += "\n";
}
// Doesn't return success/failure as the file is optional; it's okay if we fail to update it.
void RouteController::updateTableNamesFile() {
std::string contents;
addTableName(RT_TABLE_LOCAL, ROUTE_TABLE_NAME_LOCAL, &contents);
addTableName(RT_TABLE_MAIN, ROUTE_TABLE_NAME_MAIN, &contents);
addTableName(ROUTE_TABLE_LOCAL_NETWORK, ROUTE_TABLE_NAME_LOCAL_NETWORK, &contents);
addTableName(ROUTE_TABLE_LEGACY_NETWORK, ROUTE_TABLE_NAME_LEGACY_NETWORK, &contents);
addTableName(ROUTE_TABLE_LEGACY_SYSTEM, ROUTE_TABLE_NAME_LEGACY_SYSTEM, &contents);
android::RWLock::AutoRLock lock(sInterfaceToTableLock);
for (const auto& entry : sInterfaceToTable) {
addTableName(entry.second, entry.first, &contents);
}
if (!WriteStringToFile(contents, RT_TABLES_PATH, RT_TABLES_MODE, AID_SYSTEM, AID_WIFI)) {
ALOGE("failed to write to %s (%s)", RT_TABLES_PATH, strerror(errno));
return;
}
}
// Returns 0 on success or negative errno on failure.
int padInterfaceName(const char* input, char* name, size_t* length, uint16_t* padding) {
if (!input) {
*length = 0;
*padding = 0;
return 0;
}
*length = strlcpy(name, input, IFNAMSIZ) + 1;
if (*length > IFNAMSIZ) {
ALOGE("interface name too long (%zu > %u)", *length, IFNAMSIZ);
return -ENAMETOOLONG;
}
*padding = RTA_SPACE(*length) - RTA_LENGTH(*length);
return 0;
}
// Adds or removes a routing rule for IPv4 and IPv6.
//
// + If |table| is non-zero, the rule points at the specified routing table. Otherwise, the table is
// unspecified. An unspecified table is not allowed when creating an FR_ACT_TO_TBL rule.
// + If |mask| is non-zero, the rule matches the specified fwmark and mask. Otherwise, |fwmark| is
// ignored.
// + If |iif| is non-NULL, the rule matches the specified incoming interface.
// + If |oif| is non-NULL, the rule matches the specified outgoing interface.
// + If |uidStart| and |uidEnd| are not INVALID_UID, the rule matches packets from UIDs in that
// range (inclusive). Otherwise, the rule matches packets from all UIDs.
//
// Returns 0 on success or negative errno on failure.
WARN_UNUSED_RESULT int modifyIpRule(uint16_t action, uint32_t priority, uint8_t ruleType,
uint32_t table, uint32_t fwmark, uint32_t mask, const char* iif,
const char* oif, uid_t uidStart, uid_t uidEnd) {
// Ensure that if you set a bit in the fwmark, it's not being ignored by the mask.
if (fwmark & ~mask) {
ALOGE("mask 0x%x does not select all the bits set in fwmark 0x%x", mask, fwmark);
return -ERANGE;
}
// Interface names must include exactly one terminating NULL and be properly padded, or older
// kernels will refuse to delete rules.
char iifName[IFNAMSIZ], oifName[IFNAMSIZ];
size_t iifLength, oifLength;
uint16_t iifPadding, oifPadding;
if (int ret = padInterfaceName(iif, iifName, &iifLength, &iifPadding)) {
return ret;
}
if (int ret = padInterfaceName(oif, oifName, &oifLength, &oifPadding)) {
return ret;
}
// Either both start and end UID must be specified, or neither.
if ((uidStart == INVALID_UID) != (uidEnd == INVALID_UID)) {
ALOGE("incompatible start and end UIDs (%u vs %u)", uidStart, uidEnd);
return -EUSERS;
}
bool isUidRule = (uidStart != INVALID_UID);
// Assemble a rule request and put it in an array of iovec structures.
fib_rule_hdr rule = {
.action = ruleType,
// Note that here we're implicitly setting rule.table to 0. When we want to specify a
// non-zero table, we do this via the FRATTR_TABLE attribute.
};
// Don't ever create a rule that looks up table 0, because table 0 is the local table.
// It's OK to specify a table ID of 0 when deleting a rule, because that doesn't actually select
// table 0, it's a wildcard that matches anything.
if (table == RT_TABLE_UNSPEC && rule.action == FR_ACT_TO_TBL && action != RTM_DELRULE) {
ALOGE("RT_TABLE_UNSPEC only allowed when deleting rules");
return -ENOTUNIQ;
}
rtattr fraIifName = { U16_RTA_LENGTH(iifLength), FRA_IIFNAME };
rtattr fraOifName = { U16_RTA_LENGTH(oifLength), FRA_OIFNAME };
struct fib_rule_uid_range uidRange = { uidStart, uidEnd };
iovec iov[] = {
{ NULL, 0 },
{ &rule, sizeof(rule) },
{ &FRATTR_PRIORITY, sizeof(FRATTR_PRIORITY) },
{ &priority, sizeof(priority) },
{ &FRATTR_TABLE, table != RT_TABLE_UNSPEC ? sizeof(FRATTR_TABLE) : 0 },
{ &table, table != RT_TABLE_UNSPEC ? sizeof(table) : 0 },
{ &FRATTR_FWMARK, mask ? sizeof(FRATTR_FWMARK) : 0 },
{ &fwmark, mask ? sizeof(fwmark) : 0 },
{ &FRATTR_FWMASK, mask ? sizeof(FRATTR_FWMASK) : 0 },
{ &mask, mask ? sizeof(mask) : 0 },
{ &FRATTR_UID_RANGE, isUidRule ? sizeof(FRATTR_UID_RANGE) : 0 },
{ &uidRange, isUidRule ? sizeof(uidRange) : 0 },
{ &fraIifName, iif != IIF_NONE ? sizeof(fraIifName) : 0 },
{ iifName, iifLength },
{ PADDING_BUFFER, iifPadding },
{ &fraOifName, oif != OIF_NONE ? sizeof(fraOifName) : 0 },
{ oifName, oifLength },
{ PADDING_BUFFER, oifPadding },
};
uint16_t flags = (action == RTM_NEWRULE) ? NETLINK_RULE_CREATE_FLAGS : NETLINK_REQUEST_FLAGS;
for (size_t i = 0; i < ARRAY_SIZE(AF_FAMILIES); ++i) {
rule.family = AF_FAMILIES[i];
if (int ret = sendNetlinkRequest(action, flags, iov, ARRAY_SIZE(iov), nullptr)) {
if (!(action == RTM_DELRULE && ret == -ENOENT && priority == RULE_PRIORITY_TETHERING)) {
// Don't log when deleting a tethering rule that's not there. This matches the
// behaviour of clearTetheringRules, which ignores ENOENT in this case.
ALOGE("Error %s %s rule: %s", actionName(action), familyName(rule.family),
strerror(-ret));
}
return ret;
}
}
return 0;
}
WARN_UNUSED_RESULT int modifyIpRule(uint16_t action, uint32_t priority, uint32_t table,
uint32_t fwmark, uint32_t mask, const char* iif,
const char* oif, uid_t uidStart, uid_t uidEnd) {
return modifyIpRule(action, priority, FR_ACT_TO_TBL, table, fwmark, mask, iif, oif, uidStart,
uidEnd);
}
WARN_UNUSED_RESULT int modifyIpRule(uint16_t action, uint32_t priority, uint32_t table,
uint32_t fwmark, uint32_t mask) {
return modifyIpRule(action, priority, table, fwmark, mask, IIF_NONE, OIF_NONE, INVALID_UID,
INVALID_UID);
}
// Adds or deletes an IPv4 or IPv6 route.
// Returns 0 on success or negative errno on failure.
WARN_UNUSED_RESULT int modifyIpRoute(uint16_t action, uint32_t table, const char* interface,
const char* destination, const char* nexthop) {
// At least the destination must be non-null.
if (!destination) {
ALOGE("null destination");
return -EFAULT;
}
// Parse the prefix.
uint8_t rawAddress[sizeof(in6_addr)];
uint8_t family;
uint8_t prefixLength;
int rawLength = parsePrefix(destination, &family, rawAddress, sizeof(rawAddress),
&prefixLength);
if (rawLength < 0) {
ALOGE("parsePrefix failed for destination %s (%s)", destination, strerror(-rawLength));
return rawLength;
}
if (static_cast<size_t>(rawLength) > sizeof(rawAddress)) {
ALOGE("impossible! address too long (%d vs %zu)", rawLength, sizeof(rawAddress));
return -ENOBUFS; // Cannot happen; parsePrefix only supports IPv4 and IPv6.
}
uint8_t type = RTN_UNICAST;
uint32_t ifindex;
uint8_t rawNexthop[sizeof(in6_addr)];
if (nexthop && !strcmp(nexthop, "unreachable")) {
type = RTN_UNREACHABLE;
// 'interface' is likely non-NULL, as the caller (modifyRoute()) likely used it to lookup
// the table number. But it's an error to specify an interface ("dev ...") or a nexthop for
// unreachable routes, so nuke them. (IPv6 allows them to be specified; IPv4 doesn't.)
interface = OIF_NONE;
nexthop = NULL;
} else if (nexthop && !strcmp(nexthop, "throw")) {
type = RTN_THROW;
interface = OIF_NONE;
nexthop = NULL;
} else {
// If an interface was specified, find the ifindex.
if (interface != OIF_NONE) {
ifindex = if_nametoindex(interface);
if (!ifindex) {
ALOGE("cannot find interface %s", interface);
return -ENODEV;
}
}
// If a nexthop was specified, parse it as the same family as the prefix.
if (nexthop && inet_pton(family, nexthop, rawNexthop) <= 0) {
ALOGE("inet_pton failed for nexthop %s", nexthop);
return -EINVAL;
}
}
bool isDefaultThrowRoute = (type == RTN_THROW && prefixLength == 0);
// Assemble a rtmsg and put it in an array of iovec structures.
rtmsg route = {
.rtm_protocol = RTPROT_STATIC,
.rtm_type = type,
.rtm_family = family,
.rtm_dst_len = prefixLength,
.rtm_scope = static_cast<uint8_t>(nexthop ? RT_SCOPE_UNIVERSE : RT_SCOPE_LINK),
};
rtattr rtaDst = { U16_RTA_LENGTH(rawLength), RTA_DST };
rtattr rtaGateway = { U16_RTA_LENGTH(rawLength), RTA_GATEWAY };
iovec iov[] = {
{ NULL, 0 },
{ &route, sizeof(route) },
{ &RTATTR_TABLE, sizeof(RTATTR_TABLE) },
{ &table, sizeof(table) },
{ &rtaDst, sizeof(rtaDst) },
{ rawAddress, static_cast<size_t>(rawLength) },
{ &RTATTR_OIF, interface != OIF_NONE ? sizeof(RTATTR_OIF) : 0 },
{ &ifindex, interface != OIF_NONE ? sizeof(ifindex) : 0 },
{ &rtaGateway, nexthop ? sizeof(rtaGateway) : 0 },
{ rawNexthop, nexthop ? static_cast<size_t>(rawLength) : 0 },
{ &RTATTR_PRIO, isDefaultThrowRoute ? sizeof(RTATTR_PRIO) : 0 },
{ &PRIO_THROW, isDefaultThrowRoute ? sizeof(PRIO_THROW) : 0 },
};
uint16_t flags = (action == RTM_NEWROUTE) ? NETLINK_ROUTE_CREATE_FLAGS : NETLINK_REQUEST_FLAGS;
int ret = sendNetlinkRequest(action, flags, iov, ARRAY_SIZE(iov), nullptr);
if (ret) {
ALOGE("Error %s route %s -> %s %s to table %u: %s",
actionName(action), destination, nexthop, interface, table, strerror(-ret));
}
return ret;
}
// An iptables rule to mark incoming packets on a network with the netId of the network.
//
// This is so that the kernel can:
// + Use the right fwmark for (and thus correctly route) replies (e.g.: TCP RST, ICMP errors, ping
// replies, SYN-ACKs, etc).
// + Mark sockets that accept connections from this interface so that the connection stays on the
// same interface.
WARN_UNUSED_RESULT int modifyIncomingPacketMark(unsigned netId, const char* interface,
Permission permission, bool add) {
Fwmark fwmark;
fwmark.netId = netId;
fwmark.explicitlySelected = true;
fwmark.protectedFromVpn = true;
fwmark.permission = permission;
const uint32_t mask = ~Fwmark::getUidBillingMask();
std::string cmd = StringPrintf(
"%s %s -i %s -j MARK --set-mark 0x%x/0x%x", add ? "-A" : "-D",
RouteController::LOCAL_MANGLE_INPUT, interface, fwmark.intValue, mask);
if (RouteController::iptablesRestoreCommandFunction(V4V6, "mangle", cmd, nullptr) != 0) {
ALOGE("failed to change iptables rule that sets incoming packet mark");
return -EREMOTEIO;
}
return 0;
}
// A rule to route responses to the local network forwarded via the VPN.
//
// When a VPN is in effect, packets from the local network to upstream networks are forwarded into
// the VPN's tunnel interface. When the VPN forwards the responses, they emerge out of the tunnel.
WARN_UNUSED_RESULT int modifyVpnOutputToLocalRule(const char* vpnInterface, bool add) {
return modifyIpRule(add ? RTM_NEWRULE : RTM_DELRULE, RULE_PRIORITY_VPN_OUTPUT_TO_LOCAL,
ROUTE_TABLE_LOCAL_NETWORK, MARK_UNSET, MARK_UNSET, vpnInterface, OIF_NONE,
INVALID_UID, INVALID_UID);
}
// A rule to route all traffic from a given set of UIDs to go over the VPN.
//
// Notice that this rule doesn't use the netId. I.e., no matter what netId the user's socket may
// have, if they are subject to this VPN, their traffic has to go through it. Allows the traffic to
// bypass the VPN if the protectedFromVpn bit is set.
WARN_UNUSED_RESULT int modifyVpnUidRangeRule(uint32_t table, uid_t uidStart, uid_t uidEnd,
bool secure, bool add) {
Fwmark fwmark;
Fwmark mask;
fwmark.protectedFromVpn = false;
mask.protectedFromVpn = true;
uint32_t priority;
if (secure) {
priority = RULE_PRIORITY_SECURE_VPN;
} else {
priority = RULE_PRIORITY_BYPASSABLE_VPN;
fwmark.explicitlySelected = false;
mask.explicitlySelected = true;
}
return modifyIpRule(add ? RTM_NEWRULE : RTM_DELRULE, priority, table, fwmark.intValue,
mask.intValue, IIF_LOOPBACK, OIF_NONE, uidStart, uidEnd);
}
// A rule to allow system apps to send traffic over this VPN even if they are not part of the target
// set of UIDs.
//
// This is needed for DnsProxyListener to correctly resolve a request for a user who is in the
// target set, but where the DnsProxyListener itself is not.
WARN_UNUSED_RESULT int modifyVpnSystemPermissionRule(unsigned netId, uint32_t table, bool secure,
bool add) {
Fwmark fwmark;
Fwmark mask;
fwmark.netId = netId;
mask.netId = FWMARK_NET_ID_MASK;
fwmark.permission = PERMISSION_SYSTEM;
mask.permission = PERMISSION_SYSTEM;
uint32_t priority = secure ? RULE_PRIORITY_SECURE_VPN : RULE_PRIORITY_BYPASSABLE_VPN;
return modifyIpRule(add ? RTM_NEWRULE : RTM_DELRULE, priority, table, fwmark.intValue,
mask.intValue);
}
// A rule to route traffic based on an explicitly chosen network.
//
// Supports apps that use the multinetwork APIs to restrict their traffic to a network.
//
// Even though we check permissions at the time we set a netId into the fwmark of a socket, we need
// to check it again in the rules here, because a network's permissions may have been updated via
// modifyNetworkPermission().
WARN_UNUSED_RESULT int modifyExplicitNetworkRule(unsigned netId, uint32_t table,
Permission permission, uid_t uidStart,
uid_t uidEnd, bool add) {
Fwmark fwmark;
Fwmark mask;
fwmark.netId = netId;
mask.netId = FWMARK_NET_ID_MASK;
fwmark.explicitlySelected = true;
mask.explicitlySelected = true;
fwmark.permission = permission;
mask.permission = permission;
return modifyIpRule(add ? RTM_NEWRULE : RTM_DELRULE, RULE_PRIORITY_EXPLICIT_NETWORK, table,
fwmark.intValue, mask.intValue, IIF_LOOPBACK, OIF_NONE, uidStart, uidEnd);
}
// A rule to route traffic based on a chosen outgoing interface.
//
// Supports apps that use SO_BINDTODEVICE or IP_PKTINFO options and the kernel that already knows
// the outgoing interface (typically for link-local communications).
WARN_UNUSED_RESULT int modifyOutputInterfaceRules(const char* interface, uint32_t table,
Permission permission, uid_t uidStart,
uid_t uidEnd, bool add) {
Fwmark fwmark;
Fwmark mask;
fwmark.permission = permission;
mask.permission = permission;
// If this rule does not specify a UID range, then also add a corresponding high-priority rule
// for root. This covers kernel-originated packets, TEEd packets and any local daemons that open
// sockets as root.
if (uidStart == INVALID_UID && uidEnd == INVALID_UID) {
if (int ret = modifyIpRule(add ? RTM_NEWRULE : RTM_DELRULE, RULE_PRIORITY_VPN_OVERRIDE_OIF,
table, FWMARK_NONE, MASK_NONE, IIF_LOOPBACK, interface,
UID_ROOT, UID_ROOT)) {
return ret;
}
}
return modifyIpRule(add ? RTM_NEWRULE : RTM_DELRULE, RULE_PRIORITY_OUTPUT_INTERFACE, table,
fwmark.intValue, mask.intValue, IIF_LOOPBACK, interface, uidStart, uidEnd);
}
// A rule to route traffic based on the chosen network.
//
// This is for sockets that have not explicitly requested a particular network, but have been
// bound to one when they called connect(). This ensures that sockets connected on a particular
// network stay on that network even if the default network changes.
WARN_UNUSED_RESULT int modifyImplicitNetworkRule(unsigned netId, uint32_t table, bool add) {
Fwmark fwmark;
Fwmark mask;
fwmark.netId = netId;
mask.netId = FWMARK_NET_ID_MASK;
fwmark.explicitlySelected = false;
mask.explicitlySelected = true;
fwmark.permission = PERMISSION_NONE;
mask.permission = PERMISSION_NONE;
return modifyIpRule(add ? RTM_NEWRULE : RTM_DELRULE, RULE_PRIORITY_IMPLICIT_NETWORK, table,
fwmark.intValue, mask.intValue, IIF_LOOPBACK, OIF_NONE, INVALID_UID,
INVALID_UID);
}
// A rule to enable split tunnel VPNs.
//
// If a packet with a VPN's netId doesn't find a route in the VPN's routing table, it's allowed to
// go over the default network, provided it wasn't explicitly restricted to the VPN and has the
// permissions required by the default network.
WARN_UNUSED_RESULT int RouteController::modifyVpnFallthroughRule(uint16_t action, unsigned vpnNetId,
const char* physicalInterface,
Permission permission) {
uint32_t table = getRouteTableForInterface(physicalInterface);
if (table == RT_TABLE_UNSPEC) {
return -ESRCH;
}
Fwmark fwmark;
Fwmark mask;
fwmark.netId = vpnNetId;
mask.netId = FWMARK_NET_ID_MASK;
fwmark.explicitlySelected = false;
mask.explicitlySelected = true;
fwmark.permission = permission;
mask.permission = permission;
return modifyIpRule(action, RULE_PRIORITY_VPN_FALLTHROUGH, table, fwmark.intValue,
mask.intValue);
}
// Add rules to allow legacy routes added through the requestRouteToHost() API.
WARN_UNUSED_RESULT int addLegacyRouteRules() {
Fwmark fwmark;
Fwmark mask;
fwmark.explicitlySelected = false;
mask.explicitlySelected = true;
// Rules to allow legacy routes to override the default network.
if (int ret = modifyIpRule(RTM_NEWRULE, RULE_PRIORITY_LEGACY_SYSTEM, ROUTE_TABLE_LEGACY_SYSTEM,
fwmark.intValue, mask.intValue)) {
return ret;
}
if (int ret = modifyIpRule(RTM_NEWRULE, RULE_PRIORITY_LEGACY_NETWORK,
ROUTE_TABLE_LEGACY_NETWORK, fwmark.intValue, mask.intValue)) {
return ret;
}
fwmark.permission = PERMISSION_SYSTEM;
mask.permission = PERMISSION_SYSTEM;
// A rule to allow legacy routes from system apps to override VPNs.
return modifyIpRule(RTM_NEWRULE, RULE_PRIORITY_VPN_OVERRIDE_SYSTEM, ROUTE_TABLE_LEGACY_SYSTEM,
fwmark.intValue, mask.intValue);
}
// Add rules to lookup the local network when specified explicitly or otherwise.
WARN_UNUSED_RESULT int addLocalNetworkRules(unsigned localNetId) {
if (int ret = modifyExplicitNetworkRule(localNetId, ROUTE_TABLE_LOCAL_NETWORK, PERMISSION_NONE,
INVALID_UID, INVALID_UID, ACTION_ADD)) {
return ret;
}
Fwmark fwmark;
Fwmark mask;
fwmark.explicitlySelected = false;
mask.explicitlySelected = true;
return modifyIpRule(RTM_NEWRULE, RULE_PRIORITY_LOCAL_NETWORK, ROUTE_TABLE_LOCAL_NETWORK,
fwmark.intValue, mask.intValue);
}
/* static */
int RouteController::configureDummyNetwork() {
const char *interface = DummyNetwork::INTERFACE_NAME;
uint32_t table = getRouteTableForInterface(interface);
if (table == RT_TABLE_UNSPEC) {
// getRouteTableForInterface has already looged an error.
return -ESRCH;
}
ifc_init();
int ret = ifc_up(interface);
ifc_close();
if (ret) {
ALOGE("Can't bring up %s: %s", interface, strerror(errno));
return -errno;
}
if ((ret = modifyOutputInterfaceRules(interface, table, PERMISSION_NONE,
INVALID_UID, INVALID_UID, ACTION_ADD))) {
ALOGE("Can't create oif rules for %s: %s", interface, strerror(-ret));
return ret;
}
if ((ret = modifyIpRoute(RTM_NEWROUTE, table, interface, "0.0.0.0/0", NULL))) {
return ret;
}
if ((ret = modifyIpRoute(RTM_NEWROUTE, table, interface, "::/0", NULL))) {
return ret;
}
return 0;
}
// Add an explicit unreachable rule close to the end of the prioriy list to make it clear that
// relying on the kernel-default "from all lookup main" rule at priority 32766 is not intended
// behaviour. We do flush the kernel-default rules at startup, but having an explicit unreachable
// rule will hopefully make things even clearer.
WARN_UNUSED_RESULT int addUnreachableRule() {
return modifyIpRule(RTM_NEWRULE, RULE_PRIORITY_UNREACHABLE, FR_ACT_UNREACHABLE, RT_TABLE_UNSPEC,
MARK_UNSET, MARK_UNSET, IIF_NONE, OIF_NONE, INVALID_UID, INVALID_UID);
}
WARN_UNUSED_RESULT int modifyLocalNetwork(unsigned netId, const char* interface, bool add) {
if (int ret = modifyIncomingPacketMark(netId, interface, PERMISSION_NONE, add)) {
return ret;
}
return modifyOutputInterfaceRules(interface, ROUTE_TABLE_LOCAL_NETWORK, PERMISSION_NONE,
INVALID_UID, INVALID_UID, add);
}
/* static */
WARN_UNUSED_RESULT int RouteController::modifyPhysicalNetwork(unsigned netId, const char* interface,
Permission permission, bool add) {
uint32_t table = getRouteTableForInterface(interface);
if (table == RT_TABLE_UNSPEC) {
return -ESRCH;
}
if (int ret = modifyIncomingPacketMark(netId, interface, permission, add)) {
return ret;
}
if (int ret = modifyExplicitNetworkRule(netId, table, permission, INVALID_UID, INVALID_UID,
add)) {
return ret;
}
if (int ret = modifyOutputInterfaceRules(interface, table, permission, INVALID_UID, INVALID_UID,
add)) {
return ret;
}
// Only set implicit rules for networks that don't require permissions.
//
// This is so that if the default network ceases to be the default network and then switches
// from requiring no permissions to requiring permissions, we ensure that apps only use the
// network if they explicitly select it. This is consistent with destroySocketsLackingPermission
// - it closes all sockets on the network except sockets that are explicitly selected.
//
// The lack of this rule only affects the special case above, because:
// - The only cases where we implicitly bind a socket to a network are the default network and
// the bypassable VPN that applies to the app, if any.
// - This rule doesn't affect VPNs because they don't support permissions at all.
// - The default network doesn't require permissions. While we support doing this, the framework
// never does it (partly because we'd end up in the situation where we tell apps that there is
// a default network, but they can't use it).
// - If the network is still the default network, the presence or absence of this rule does not
// matter.
//
// Therefore, for the lack of this rule to affect a socket, the socket has to have been
// implicitly bound to a network because at the time of connect() it was the default, and that
// network must no longer be the default, and must now require permissions.
if (permission == PERMISSION_NONE) {
return modifyImplicitNetworkRule(netId, table, add);
}
return 0;
}
WARN_UNUSED_RESULT int modifyRejectNonSecureNetworkRule(const UidRanges& uidRanges, bool add) {
Fwmark fwmark;
Fwmark mask;
fwmark.protectedFromVpn = false;
mask.protectedFromVpn = true;
for (const UidRange& range : uidRanges.getRanges()) {
if (int ret = modifyIpRule(add ? RTM_NEWRULE : RTM_DELRULE,
RULE_PRIORITY_PROHIBIT_NON_VPN, FR_ACT_PROHIBIT, RT_TABLE_UNSPEC,
fwmark.intValue, mask.intValue, IIF_LOOPBACK, OIF_NONE,
range.getStart(), range.getStop())) {
return ret;
}
}
return 0;
}
WARN_UNUSED_RESULT int RouteController::modifyVirtualNetwork(unsigned netId, const char* interface,
const UidRanges& uidRanges,
bool secure, bool add,
bool modifyNonUidBasedRules) {
uint32_t table = getRouteTableForInterface(interface);
if (table == RT_TABLE_UNSPEC) {
return -ESRCH;
}
for (const UidRange& range : uidRanges.getRanges()) {
if (int ret = modifyVpnUidRangeRule(table, range.getStart(), range.getStop(), secure, add))
{
return ret;
}
if (int ret = modifyExplicitNetworkRule(netId, table, PERMISSION_NONE, range.getStart(),
range.getStop(), add)) {
return ret;
}
if (int ret = modifyOutputInterfaceRules(interface, table, PERMISSION_NONE,
range.getStart(), range.getStop(), add)) {
return ret;
}
}
if (modifyNonUidBasedRules) {
if (int ret = modifyIncomingPacketMark(netId, interface, PERMISSION_NONE, add)) {
return ret;
}
if (int ret = modifyVpnOutputToLocalRule(interface, add)) {
return ret;
}
if (int ret = modifyVpnSystemPermissionRule(netId, table, secure, add)) {
return ret;
}
return modifyExplicitNetworkRule(netId, table, PERMISSION_NONE, UID_ROOT, UID_ROOT, add);
}
return 0;
}
WARN_UNUSED_RESULT int RouteController::modifyDefaultNetwork(uint16_t action, const char* interface,
Permission permission) {
uint32_t table = getRouteTableForInterface(interface);
if (table == RT_TABLE_UNSPEC) {
return -ESRCH;
}
Fwmark fwmark;
Fwmark mask;
fwmark.netId = NETID_UNSET;
mask.netId = FWMARK_NET_ID_MASK;
fwmark.permission = permission;
mask.permission = permission;
return modifyIpRule(action, RULE_PRIORITY_DEFAULT_NETWORK, table, fwmark.intValue,
mask.intValue, IIF_LOOPBACK, OIF_NONE, INVALID_UID, INVALID_UID);
}
WARN_UNUSED_RESULT int RouteController::modifyTetheredNetwork(uint16_t action,
const char* inputInterface,
const char* outputInterface) {
uint32_t table = getRouteTableForInterface(outputInterface);
if (table == RT_TABLE_UNSPEC) {
return -ESRCH;
}
return modifyIpRule(action, RULE_PRIORITY_TETHERING, table, MARK_UNSET, MARK_UNSET,
inputInterface, OIF_NONE, INVALID_UID, INVALID_UID);
}
// Adds or removes an IPv4 or IPv6 route to the specified table.
// Returns 0 on success or negative errno on failure.
WARN_UNUSED_RESULT int RouteController::modifyRoute(uint16_t action, const char* interface,
const char* destination, const char* nexthop,
TableType tableType) {
uint32_t table;
switch (tableType) {
case RouteController::INTERFACE: {
table = getRouteTableForInterface(interface);
if (table == RT_TABLE_UNSPEC) {
return -ESRCH;
}
break;
}
case RouteController::LOCAL_NETWORK: {
table = ROUTE_TABLE_LOCAL_NETWORK;
break;
}
case RouteController::LEGACY_NETWORK: {
table = ROUTE_TABLE_LEGACY_NETWORK;
break;
}
case RouteController::LEGACY_SYSTEM: {
table = ROUTE_TABLE_LEGACY_SYSTEM;
break;
}
}
int ret = modifyIpRoute(action, table, interface, destination, nexthop);
// Trying to add a route that already exists shouldn't cause an error.
if (ret && !(action == RTM_NEWROUTE && ret == -EEXIST)) {
return ret;
}
return 0;
}
WARN_UNUSED_RESULT int clearTetheringRules(const char* inputInterface) {
int ret = 0;
while (ret == 0) {
ret = modifyIpRule(RTM_DELRULE, RULE_PRIORITY_TETHERING, 0, MARK_UNSET, MARK_UNSET,
inputInterface, OIF_NONE, INVALID_UID, INVALID_UID);
}
if (ret == -ENOENT) {
return 0;
} else {
return ret;
}
}
uint32_t getRulePriority(const nlmsghdr *nlh) {
return getRtmU32Attribute(nlh, FRA_PRIORITY);
}
uint32_t getRouteTable(const nlmsghdr *nlh) {
return getRtmU32Attribute(nlh, RTA_TABLE);
}
WARN_UNUSED_RESULT int flushRules() {
NetlinkDumpFilter shouldDelete = [] (nlmsghdr *nlh) {
// Don't touch rules at priority 0 because by default they are used for local input.
return getRulePriority(nlh) != 0;
};
return rtNetlinkFlush(RTM_GETRULE, RTM_DELRULE, "rules", shouldDelete);
}
WARN_UNUSED_RESULT int RouteController::flushRoutes(uint32_t table) {
NetlinkDumpFilter shouldDelete = [table] (nlmsghdr *nlh) {
return getRouteTable(nlh) == table;
};
return rtNetlinkFlush(RTM_GETROUTE, RTM_DELROUTE, "routes", shouldDelete);
}
// Returns 0 on success or negative errno on failure.
WARN_UNUSED_RESULT int RouteController::flushRoutes(const char* interface) {
android::RWLock::AutoWLock lock(sInterfaceToTableLock);
uint32_t table = getRouteTableForInterfaceLocked(interface);
if (table == RT_TABLE_UNSPEC) {
return -ESRCH;
}
int ret = flushRoutes(table);
// If we failed to flush routes, the caller may elect to keep this interface around, so keep
// track of its name.
if (ret == 0) {
sInterfaceToTable.erase(interface);
}
return ret;
}
int RouteController::Init(unsigned localNetId) {
if (int ret = flushRules()) {
return ret;
}
if (int ret = addLegacyRouteRules()) {
return ret;
}
if (int ret = addLocalNetworkRules(localNetId)) {
return ret;
}
if (int ret = addUnreachableRule()) {
return ret;
}
// Don't complain if we can't add the dummy network, since not all devices support it.
configureDummyNetwork();
updateTableNamesFile();
return 0;
}
int RouteController::addInterfaceToLocalNetwork(unsigned netId, const char* interface) {
return modifyLocalNetwork(netId, interface, ACTION_ADD);
}
int RouteController::removeInterfaceFromLocalNetwork(unsigned netId, const char* interface) {
return modifyLocalNetwork(netId, interface, ACTION_DEL);
}
int RouteController::addInterfaceToPhysicalNetwork(unsigned netId, const char* interface,
Permission permission) {
if (int ret = modifyPhysicalNetwork(netId, interface, permission, ACTION_ADD)) {
return ret;
}
updateTableNamesFile();
return 0;
}
int RouteController::removeInterfaceFromPhysicalNetwork(unsigned netId, const char* interface,
Permission permission) {
if (int ret = modifyPhysicalNetwork(netId, interface, permission, ACTION_DEL)) {
return ret;
}
if (int ret = flushRoutes(interface)) {
return ret;
}
if (int ret = clearTetheringRules(interface)) {
return ret;
}
updateTableNamesFile();
return 0;
}
int RouteController::addInterfaceToVirtualNetwork(unsigned netId, const char* interface,
bool secure, const UidRanges& uidRanges) {
if (int ret = modifyVirtualNetwork(netId, interface, uidRanges, secure, ACTION_ADD,
MODIFY_NON_UID_BASED_RULES)) {
return ret;
}
updateTableNamesFile();
return 0;
}
int RouteController::removeInterfaceFromVirtualNetwork(unsigned netId, const char* interface,
bool secure, const UidRanges& uidRanges) {
if (int ret = modifyVirtualNetwork(netId, interface, uidRanges, secure, ACTION_DEL,
MODIFY_NON_UID_BASED_RULES)) {
return ret;
}
if (int ret = flushRoutes(interface)) {
return ret;
}
updateTableNamesFile();
return 0;
}
int RouteController::modifyPhysicalNetworkPermission(unsigned netId, const char* interface,
Permission oldPermission,
Permission newPermission) {
// Add the new rules before deleting the old ones, to avoid race conditions.
if (int ret = modifyPhysicalNetwork(netId, interface, newPermission, ACTION_ADD)) {
return ret;
}
return modifyPhysicalNetwork(netId, interface, oldPermission, ACTION_DEL);
}
int RouteController::addUsersToRejectNonSecureNetworkRule(const UidRanges& uidRanges) {
return modifyRejectNonSecureNetworkRule(uidRanges, true);
}
int RouteController::removeUsersFromRejectNonSecureNetworkRule(const UidRanges& uidRanges) {
return modifyRejectNonSecureNetworkRule(uidRanges, false);
}
int RouteController::addUsersToVirtualNetwork(unsigned netId, const char* interface, bool secure,
const UidRanges& uidRanges) {
return modifyVirtualNetwork(netId, interface, uidRanges, secure, ACTION_ADD,
!MODIFY_NON_UID_BASED_RULES);
}
int RouteController::removeUsersFromVirtualNetwork(unsigned netId, const char* interface,
bool secure, const UidRanges& uidRanges) {
return modifyVirtualNetwork(netId, interface, uidRanges, secure, ACTION_DEL,
!MODIFY_NON_UID_BASED_RULES);
}
int RouteController::addInterfaceToDefaultNetwork(const char* interface, Permission permission) {
return modifyDefaultNetwork(RTM_NEWRULE, interface, permission);
}
int RouteController::removeInterfaceFromDefaultNetwork(const char* interface,
Permission permission) {
return modifyDefaultNetwork(RTM_DELRULE, interface, permission);
}
int RouteController::addRoute(const char* interface, const char* destination, const char* nexthop,
TableType tableType) {
return modifyRoute(RTM_NEWROUTE, interface, destination, nexthop, tableType);
}
int RouteController::removeRoute(const char* interface, const char* destination,
const char* nexthop, TableType tableType) {
return modifyRoute(RTM_DELROUTE, interface, destination, nexthop, tableType);
}
int RouteController::enableTethering(const char* inputInterface, const char* outputInterface) {
return modifyTetheredNetwork(RTM_NEWRULE, inputInterface, outputInterface);
}
int RouteController::disableTethering(const char* inputInterface, const char* outputInterface) {
return modifyTetheredNetwork(RTM_DELRULE, inputInterface, outputInterface);
}
int RouteController::addVirtualNetworkFallthrough(unsigned vpnNetId, const char* physicalInterface,
Permission permission) {
return modifyVpnFallthroughRule(RTM_NEWRULE, vpnNetId, physicalInterface, permission);
}
int RouteController::removeVirtualNetworkFallthrough(unsigned vpnNetId,
const char* physicalInterface,
Permission permission) {
return modifyVpnFallthroughRule(RTM_DELRULE, vpnNetId, physicalInterface, permission);
}
// Protects sInterfaceToTable.
android::RWLock RouteController::sInterfaceToTableLock;
std::map<std::string, uint32_t> RouteController::sInterfaceToTable;
} // namespace net
} // namespace android