// Copyright 2016 syzkaller project authors. All rights reserved. // Use of this source code is governed by Apache 2 LICENSE that can be found in the LICENSE file. // This file is shared between executor and csource package. #include #include #include #include #if SYZ_EXECUTOR const int kExtraCoverSize = 1024 << 10; struct cover_t; static void cover_reset(cover_t* cov); #endif #if SYZ_EXECUTOR || SYZ_THREADED #include #include typedef struct { int state; } event_t; static void event_init(event_t* ev) { ev->state = 0; } static void event_reset(event_t* ev) { ev->state = 0; } static void event_set(event_t* ev) { if (ev->state) exitf("event already set"); __atomic_store_n(&ev->state, 1, __ATOMIC_RELEASE); syscall(SYS_futex, &ev->state, FUTEX_WAKE | FUTEX_PRIVATE_FLAG, 1000000); } static void event_wait(event_t* ev) { while (!__atomic_load_n(&ev->state, __ATOMIC_ACQUIRE)) syscall(SYS_futex, &ev->state, FUTEX_WAIT | FUTEX_PRIVATE_FLAG, 0, 0); } static int event_isset(event_t* ev) { return __atomic_load_n(&ev->state, __ATOMIC_ACQUIRE); } static int event_timedwait(event_t* ev, uint64 timeout) { uint64 start = current_time_ms(); uint64 now = start; for (;;) { uint64 remain = timeout - (now - start); struct timespec ts; ts.tv_sec = remain / 1000; ts.tv_nsec = (remain % 1000) * 1000 * 1000; syscall(SYS_futex, &ev->state, FUTEX_WAIT | FUTEX_PRIVATE_FLAG, 0, &ts); if (__atomic_load_n(&ev->state, __ATOMIC_ACQUIRE)) return 1; now = current_time_ms(); if (now - start > timeout) return 0; } } #endif #if SYZ_EXECUTOR || SYZ_REPEAT || SYZ_NET_INJECTION || SYZ_FAULT || SYZ_SANDBOX_NONE || \ SYZ_SANDBOX_SETUID || SYZ_SANDBOX_NAMESPACE || SYZ_SANDBOX_ANDROID || \ SYZ_FAULT || SYZ_LEAK || SYZ_BINFMT_MISC || SYZ_SYSCTL || \ ((__NR_syz_usb_connect || __NR_syz_usb_connect_ath9k) && USB_DEBUG) || \ __NR_syz_usbip_server_init #include #include #include #include #include #include #include static bool write_file(const char* file, const char* what, ...) { char buf[1024]; va_list args; va_start(args, what); vsnprintf(buf, sizeof(buf), what, args); va_end(args); buf[sizeof(buf) - 1] = 0; int len = strlen(buf); int fd = open(file, O_WRONLY | O_CLOEXEC); if (fd == -1) return false; if (write(fd, buf, len) != len) { int err = errno; close(fd); debug("write(%s) failed: %d\n", file, err); errno = err; return false; } close(fd); return true; } #endif #if SYZ_EXECUTOR || SYZ_NET_DEVICES || SYZ_NET_INJECTION || SYZ_DEVLINK_PCI || SYZ_WIFI || SYZ_802154 || \ __NR_syz_genetlink_get_family_id || __NR_syz_80211_inject_frame || __NR_syz_80211_join_ibss || SYZ_NIC_VF #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include struct nlmsg { char* pos; int nesting; struct nlattr* nested[8]; char buf[4096]; }; static void netlink_init(struct nlmsg* nlmsg, int typ, int flags, const void* data, int size) { memset(nlmsg, 0, sizeof(*nlmsg)); struct nlmsghdr* hdr = (struct nlmsghdr*)nlmsg->buf; hdr->nlmsg_type = typ; hdr->nlmsg_flags = NLM_F_REQUEST | NLM_F_ACK | flags; memcpy(hdr + 1, data, size); nlmsg->pos = (char*)(hdr + 1) + NLMSG_ALIGN(size); } static void netlink_attr(struct nlmsg* nlmsg, int typ, const void* data, int size) { struct nlattr* attr = (struct nlattr*)nlmsg->pos; attr->nla_len = sizeof(*attr) + size; attr->nla_type = typ; if (size > 0) memcpy(attr + 1, data, size); nlmsg->pos += NLMSG_ALIGN(attr->nla_len); } #if SYZ_EXECUTOR || SYZ_NET_DEVICES || SYZ_802154 static void netlink_nest(struct nlmsg* nlmsg, int typ) { struct nlattr* attr = (struct nlattr*)nlmsg->pos; attr->nla_type = typ; nlmsg->pos += sizeof(*attr); nlmsg->nested[nlmsg->nesting++] = attr; } static void netlink_done(struct nlmsg* nlmsg) { struct nlattr* attr = nlmsg->nested[--nlmsg->nesting]; attr->nla_len = nlmsg->pos - (char*)attr; } #if SYZ_EXECUTOR || SYZ_NIC_VF #include #include #include #include struct vf_intf { char pass_thru_intf[IFNAMSIZ]; int ppid; // used by Child }; static struct vf_intf vf_intf; static void find_vf_interface(void) { #if SYZ_EXECUTOR if (!flag_nic_vf) return; #endif struct ifaddrs* addresses = NULL; int pid = getpid(); int ret = 0; memset(&vf_intf, 0, sizeof(struct vf_intf)); debug("Checking for VF pass-thru interface.\n"); if (getifaddrs(&addresses) == -1) { debug("%s: getifaddrs() failed.\n", __func__); return; } int fd = socket(AF_INET, SOCK_DGRAM, IPPROTO_IP); if (fd < 0) { debug("%s: socket() failed.\n", __func__); return; } struct ifreq ifr; struct ethtool_drvinfo drvinfo; struct ifaddrs* address = addresses; while (address) { debug("ifa_name: %s\n", address->ifa_name); memset(&ifr, 0, sizeof(struct ifreq)); strcpy(ifr.ifr_name, address->ifa_name); memset(&drvinfo, 0, sizeof(struct ethtool_drvinfo)); drvinfo.cmd = ETHTOOL_GDRVINFO; ifr.ifr_data = (caddr_t)&drvinfo; ret = ioctl(fd, SIOCETHTOOL, &ifr); if (ret < 0) { debug("%s: ioctl() failed.\n", __func__); } else if (strlen(drvinfo.bus_info)) { debug("bus_info: %s, strlen(drvinfo.bus_info)=%zu\n", drvinfo.bus_info, strlen(drvinfo.bus_info)); if (strcmp(drvinfo.bus_info, "0000:00:11.0") == 0) { if (strlen(address->ifa_name) < IFNAMSIZ) { strncpy(vf_intf.pass_thru_intf, address->ifa_name, IFNAMSIZ); vf_intf.ppid = pid; } else { debug("%s: %d strlen(%s) >= IFNAMSIZ.\n", __func__, pid, address->ifa_name); } break; } } address = address->ifa_next; } freeifaddrs(addresses); if (!vf_intf.ppid) { memset(&vf_intf, 0, sizeof(struct vf_intf)); debug("%s: %d could not find VF pass-thru interface.\n", __func__, pid); return; } debug("%s: %d found VF pass-thru interface %s\n", __func__, pid, vf_intf.pass_thru_intf); } #endif // SYZ_NIC_VF #endif static int netlink_send_ext(struct nlmsg* nlmsg, int sock, uint16 reply_type, int* reply_len, bool dofail) { #if SYZ_EXECUTOR if (in_execute_one && dofail) { // We can expect different sorts of breakages during fuzzing, // we should not kill the whole process because of them. failmsg("invalid netlink_send_ext arguments", "dofail is true during syscall execution"); } #endif if (nlmsg->pos > nlmsg->buf + sizeof(nlmsg->buf) || nlmsg->nesting) fail("nlmsg overflow/bad nesting"); struct nlmsghdr* hdr = (struct nlmsghdr*)nlmsg->buf; hdr->nlmsg_len = nlmsg->pos - nlmsg->buf; struct sockaddr_nl addr; memset(&addr, 0, sizeof(addr)); addr.nl_family = AF_NETLINK; ssize_t n = sendto(sock, nlmsg->buf, hdr->nlmsg_len, 0, (struct sockaddr*)&addr, sizeof(addr)); if (n != (ssize_t)hdr->nlmsg_len) { if (dofail) failmsg("netlink_send_ext: short netlink write", "wrote=%zd, want=%d", n, hdr->nlmsg_len); debug("netlink_send_ext: short netlink write: %zd/%d errno=%d\n", n, hdr->nlmsg_len, errno); return -1; } n = recv(sock, nlmsg->buf, sizeof(nlmsg->buf), 0); if (reply_len) *reply_len = 0; if (n < 0) { if (dofail) fail("netlink_send_ext: netlink read failed"); debug("netlink_send_ext: netlink read failed: errno=%d\n", errno); return -1; } if (n < (ssize_t)sizeof(struct nlmsghdr)) { errno = EINVAL; if (dofail) failmsg("netlink_send_ext: short netlink read", "read=%zd", n); debug("netlink_send_ext: short netlink read: %zd\n", n); return -1; } if (hdr->nlmsg_type == NLMSG_DONE) return 0; if (reply_len && hdr->nlmsg_type == reply_type) { *reply_len = n; return 0; } if (n < (ssize_t)(sizeof(struct nlmsghdr) + sizeof(struct nlmsgerr))) { errno = EINVAL; if (dofail) failmsg("netlink_send_ext: short netlink read", "read=%zd", n); debug("netlink_send_ext: short netlink read: %zd\n", n); return -1; } if (hdr->nlmsg_type != NLMSG_ERROR) { errno = EINVAL; if (dofail) failmsg("netlink_send_ext: bad netlink ack type", "type=%d", hdr->nlmsg_type); debug("netlink_send_ext: short netlink ack: %d\n", hdr->nlmsg_type); return -1; } errno = -((struct nlmsgerr*)(hdr + 1))->error; return -errno; } #if SYZ_EXECUTOR || SYZ_NET_DEVICES || SYZ_NET_INJECTION || SYZ_DEVLINK_PCI || SYZ_WIFI || SYZ_802154 static int netlink_send(struct nlmsg* nlmsg, int sock) { return netlink_send_ext(nlmsg, sock, 0, NULL, true); } #endif static int netlink_query_family_id(struct nlmsg* nlmsg, int sock, const char* family_name, bool dofail) { struct genlmsghdr genlhdr; memset(&genlhdr, 0, sizeof(genlhdr)); genlhdr.cmd = CTRL_CMD_GETFAMILY; netlink_init(nlmsg, GENL_ID_CTRL, 0, &genlhdr, sizeof(genlhdr)); netlink_attr(nlmsg, CTRL_ATTR_FAMILY_NAME, family_name, strnlen(family_name, GENL_NAMSIZ - 1) + 1); int n = 0; int err = netlink_send_ext(nlmsg, sock, GENL_ID_CTRL, &n, dofail); if (err < 0) { debug("netlink: failed to get family id for %.*s: %s\n", GENL_NAMSIZ, family_name, strerror(errno)); return -1; } uint16 id = 0; struct nlattr* attr = (struct nlattr*)(nlmsg->buf + NLMSG_HDRLEN + NLMSG_ALIGN(sizeof(genlhdr))); for (; (char*)attr < nlmsg->buf + n; attr = (struct nlattr*)((char*)attr + NLMSG_ALIGN(attr->nla_len))) { if (attr->nla_type == CTRL_ATTR_FAMILY_ID) { id = *(uint16*)(attr + 1); break; } } if (!id) { debug("netlink: failed to parse family id for %.*s\n", GENL_NAMSIZ, family_name); errno = EINVAL; return -1; } recv(sock, nlmsg->buf, sizeof(nlmsg->buf), 0); // recv ack return id; } #if SYZ_EXECUTOR || SYZ_NET_DEVICES || SYZ_DEVLINK_PCI static int netlink_next_msg(struct nlmsg* nlmsg, unsigned int offset, unsigned int total_len) { struct nlmsghdr* hdr = (struct nlmsghdr*)(nlmsg->buf + offset); if (offset == total_len || offset + hdr->nlmsg_len > total_len) return -1; return hdr->nlmsg_len; } #endif #if SYZ_EXECUTOR || SYZ_NET_DEVICES || SYZ_802154 // Force few TX and RX queues per interface to avoid creating 2 sysfs entries // per CPU per interface which takes a long time on machines with many cores. static unsigned int queue_count = 2; static void netlink_add_device_impl(struct nlmsg* nlmsg, const char* type, const char* name, bool up) { struct ifinfomsg hdr; memset(&hdr, 0, sizeof(hdr)); if (up) hdr.ifi_flags = hdr.ifi_change = IFF_UP; netlink_init(nlmsg, RTM_NEWLINK, NLM_F_EXCL | NLM_F_CREATE, &hdr, sizeof(hdr)); if (name) netlink_attr(nlmsg, IFLA_IFNAME, name, strlen(name)); netlink_attr(nlmsg, IFLA_NUM_TX_QUEUES, &queue_count, sizeof(queue_count)); netlink_attr(nlmsg, IFLA_NUM_RX_QUEUES, &queue_count, sizeof(queue_count)); netlink_nest(nlmsg, IFLA_LINKINFO); netlink_attr(nlmsg, IFLA_INFO_KIND, type, strlen(type)); } #endif #if SYZ_EXECUTOR || SYZ_NET_DEVICES static void netlink_add_device(struct nlmsg* nlmsg, int sock, const char* type, const char* name) { netlink_add_device_impl(nlmsg, type, name, false); netlink_done(nlmsg); int err = netlink_send(nlmsg, sock); if (err < 0) { debug("netlink: adding device %s type %s: %s\n", name, type, strerror(errno)); } } static void netlink_add_veth(struct nlmsg* nlmsg, int sock, const char* name, const char* peer) { netlink_add_device_impl(nlmsg, "veth", name, false); netlink_nest(nlmsg, IFLA_INFO_DATA); netlink_nest(nlmsg, VETH_INFO_PEER); nlmsg->pos += sizeof(struct ifinfomsg); netlink_attr(nlmsg, IFLA_IFNAME, peer, strlen(peer)); netlink_attr(nlmsg, IFLA_NUM_TX_QUEUES, &queue_count, sizeof(queue_count)); netlink_attr(nlmsg, IFLA_NUM_RX_QUEUES, &queue_count, sizeof(queue_count)); netlink_done(nlmsg); netlink_done(nlmsg); netlink_done(nlmsg); int err = netlink_send(nlmsg, sock); if (err < 0) { debug("netlink: adding device %s type veth peer %s: %s\n", name, peer, strerror(errno)); } } static void netlink_add_xfrm(struct nlmsg* nlmsg, int sock, const char* name) { netlink_add_device_impl(nlmsg, "xfrm", name, true); netlink_nest(nlmsg, IFLA_INFO_DATA); int if_id = 1; // This is IFLA_XFRM_IF_ID attr which is not present in older kernel headers. netlink_attr(nlmsg, 2, &if_id, sizeof(if_id)); netlink_done(nlmsg); netlink_done(nlmsg); int err = netlink_send(nlmsg, sock); if (err < 0) { debug("netlink: adding device %s type xfrm if_id %d: %s\n", name, if_id, strerror(errno)); } } static void netlink_add_hsr(struct nlmsg* nlmsg, int sock, const char* name, const char* slave1, const char* slave2) { netlink_add_device_impl(nlmsg, "hsr", name, false); netlink_nest(nlmsg, IFLA_INFO_DATA); int ifindex1 = if_nametoindex(slave1); netlink_attr(nlmsg, IFLA_HSR_SLAVE1, &ifindex1, sizeof(ifindex1)); int ifindex2 = if_nametoindex(slave2); netlink_attr(nlmsg, IFLA_HSR_SLAVE2, &ifindex2, sizeof(ifindex2)); netlink_done(nlmsg); netlink_done(nlmsg); int err = netlink_send(nlmsg, sock); if (err < 0) { debug("netlink: adding device %s type hsr slave1 %s slave2 %s: %s\n", name, slave1, slave2, strerror(errno)); } } static void netlink_add_linked(struct nlmsg* nlmsg, int sock, const char* type, const char* name, const char* link) { netlink_add_device_impl(nlmsg, type, name, false); netlink_done(nlmsg); int ifindex = if_nametoindex(link); netlink_attr(nlmsg, IFLA_LINK, &ifindex, sizeof(ifindex)); int err = netlink_send(nlmsg, sock); if (err < 0) { debug("netlink: adding device %s type %s link %s: %s\n", name, type, link, strerror(errno)); } } static void netlink_add_vlan(struct nlmsg* nlmsg, int sock, const char* name, const char* link, uint16 id, uint16 proto) { netlink_add_device_impl(nlmsg, "vlan", name, false); netlink_nest(nlmsg, IFLA_INFO_DATA); netlink_attr(nlmsg, IFLA_VLAN_ID, &id, sizeof(id)); netlink_attr(nlmsg, IFLA_VLAN_PROTOCOL, &proto, sizeof(proto)); netlink_done(nlmsg); netlink_done(nlmsg); int ifindex = if_nametoindex(link); netlink_attr(nlmsg, IFLA_LINK, &ifindex, sizeof(ifindex)); int err = netlink_send(nlmsg, sock); if (err < 0) { debug("netlink: add %s type vlan link %s id %d: %s\n", name, link, id, strerror(errno)); } } static void netlink_add_macvlan(struct nlmsg* nlmsg, int sock, const char* name, const char* link) { netlink_add_device_impl(nlmsg, "macvlan", name, false); netlink_nest(nlmsg, IFLA_INFO_DATA); uint32 mode = MACVLAN_MODE_BRIDGE; netlink_attr(nlmsg, IFLA_MACVLAN_MODE, &mode, sizeof(mode)); netlink_done(nlmsg); netlink_done(nlmsg); int ifindex = if_nametoindex(link); netlink_attr(nlmsg, IFLA_LINK, &ifindex, sizeof(ifindex)); int err = netlink_send(nlmsg, sock); if (err < 0) { debug("netlink: add %s type macvlan link %s mode %d: %s\n", name, link, mode, strerror(errno)); } } static void netlink_add_geneve(struct nlmsg* nlmsg, int sock, const char* name, uint32 vni, struct in_addr* addr4, struct in6_addr* addr6) { netlink_add_device_impl(nlmsg, "geneve", name, false); netlink_nest(nlmsg, IFLA_INFO_DATA); netlink_attr(nlmsg, IFLA_GENEVE_ID, &vni, sizeof(vni)); if (addr4) netlink_attr(nlmsg, IFLA_GENEVE_REMOTE, addr4, sizeof(*addr4)); if (addr6) netlink_attr(nlmsg, IFLA_GENEVE_REMOTE6, addr6, sizeof(*addr6)); netlink_done(nlmsg); netlink_done(nlmsg); int err = netlink_send(nlmsg, sock); if (err < 0) { debug("netlink: add %s type geneve vni %u: %s\n", name, vni, strerror(errno)); } } #define IFLA_IPVLAN_FLAGS 2 #define IPVLAN_MODE_L3S 2 #undef IPVLAN_F_VEPA #define IPVLAN_F_VEPA 2 static void netlink_add_ipvlan(struct nlmsg* nlmsg, int sock, const char* name, const char* link, uint16 mode, uint16 flags) { netlink_add_device_impl(nlmsg, "ipvlan", name, false); netlink_nest(nlmsg, IFLA_INFO_DATA); netlink_attr(nlmsg, IFLA_IPVLAN_MODE, &mode, sizeof(mode)); netlink_attr(nlmsg, IFLA_IPVLAN_FLAGS, &flags, sizeof(flags)); netlink_done(nlmsg); netlink_done(nlmsg); int ifindex = if_nametoindex(link); netlink_attr(nlmsg, IFLA_LINK, &ifindex, sizeof(ifindex)); int err = netlink_send(nlmsg, sock); if (err < 0) { debug("netlink: add %s type ipvlan link %s mode %d: %s\n", name, link, mode, strerror(errno)); } } #endif #if SYZ_EXECUTOR || SYZ_NET_DEVICES || SYZ_NET_INJECTION || SYZ_DEVLINK_PCI || SYZ_802154 static void netlink_device_change(struct nlmsg* nlmsg, int sock, const char* name, bool up, const char* master, const void* mac, int macsize, const char* new_name) { struct ifinfomsg hdr; memset(&hdr, 0, sizeof(hdr)); if (up) hdr.ifi_flags = hdr.ifi_change = IFF_UP; hdr.ifi_index = if_nametoindex(name); netlink_init(nlmsg, RTM_NEWLINK, 0, &hdr, sizeof(hdr)); if (new_name) netlink_attr(nlmsg, IFLA_IFNAME, new_name, strlen(new_name)); if (master) { int ifindex = if_nametoindex(master); netlink_attr(nlmsg, IFLA_MASTER, &ifindex, sizeof(ifindex)); } if (macsize) netlink_attr(nlmsg, IFLA_ADDRESS, mac, macsize); int err = netlink_send(nlmsg, sock); if (err < 0) { debug("netlink: device %s up master %s: %s\n", name, master ? master : "NULL", strerror(errno)); } } #endif #if SYZ_EXECUTOR || SYZ_NET_DEVICES || SYZ_NET_INJECTION static int netlink_add_addr(struct nlmsg* nlmsg, int sock, const char* dev, const void* addr, int addrsize) { struct ifaddrmsg hdr; memset(&hdr, 0, sizeof(hdr)); hdr.ifa_family = addrsize == 4 ? AF_INET : AF_INET6; hdr.ifa_prefixlen = addrsize == 4 ? 24 : 120; hdr.ifa_scope = RT_SCOPE_UNIVERSE; hdr.ifa_index = if_nametoindex(dev); netlink_init(nlmsg, RTM_NEWADDR, NLM_F_CREATE | NLM_F_REPLACE, &hdr, sizeof(hdr)); netlink_attr(nlmsg, IFA_LOCAL, addr, addrsize); netlink_attr(nlmsg, IFA_ADDRESS, addr, addrsize); return netlink_send(nlmsg, sock); } static void netlink_add_addr4(struct nlmsg* nlmsg, int sock, const char* dev, const char* addr) { struct in_addr in_addr; inet_pton(AF_INET, addr, &in_addr); int err = netlink_add_addr(nlmsg, sock, dev, &in_addr, sizeof(in_addr)); if (err < 0) { debug("netlink: add addr %s dev %s: %s\n", addr, dev, strerror(errno)); } } static void netlink_add_addr6(struct nlmsg* nlmsg, int sock, const char* dev, const char* addr) { struct in6_addr in6_addr; inet_pton(AF_INET6, addr, &in6_addr); int err = netlink_add_addr(nlmsg, sock, dev, &in6_addr, sizeof(in6_addr)); if (err < 0) { debug("netlink: add addr %s dev %s: %s\n", addr, dev, strerror(errno)); } } #endif #if SYZ_EXECUTOR || SYZ_NET_INJECTION static void netlink_add_neigh(struct nlmsg* nlmsg, int sock, const char* name, const void* addr, int addrsize, const void* mac, int macsize) { struct ndmsg hdr; memset(&hdr, 0, sizeof(hdr)); hdr.ndm_family = addrsize == 4 ? AF_INET : AF_INET6; hdr.ndm_ifindex = if_nametoindex(name); hdr.ndm_state = NUD_PERMANENT; netlink_init(nlmsg, RTM_NEWNEIGH, NLM_F_EXCL | NLM_F_CREATE, &hdr, sizeof(hdr)); netlink_attr(nlmsg, NDA_DST, addr, addrsize); netlink_attr(nlmsg, NDA_LLADDR, mac, macsize); int err = netlink_send(nlmsg, sock); if (err < 0) { debug("netlink: add neigh %s addr %d lladdr %d: %s\n", name, addrsize, macsize, strerror(errno)); } } #endif #endif #if SYZ_EXECUTOR || SYZ_NET_DEVICES || SYZ_NET_INJECTION || SYZ_DEVLINK_PCI || SYZ_WIFI || SYZ_802154 static struct nlmsg nlmsg; #endif #if SYZ_EXECUTOR || SYZ_NET_INJECTION #include #include #include #include #include #include #include #include #include #include #include #include #include static int tunfd = -1; #define TUN_IFACE "syz_tun" #define LOCAL_MAC 0xaaaaaaaaaaaa #define REMOTE_MAC 0xaaaaaaaaaabb #define LOCAL_IPV4 "172.20.20.170" #define REMOTE_IPV4 "172.20.20.187" #define LOCAL_IPV6 "fe80::aa" #define REMOTE_IPV6 "fe80::bb" #ifndef IFF_NAPI #define IFF_NAPI 0x0010 #endif #if ENABLE_NAPI_FRAGS static int tun_frags_enabled; #ifndef IFF_NAPI_FRAGS #define IFF_NAPI_FRAGS 0x0020 #endif #endif static void initialize_tun(void) { #if SYZ_EXECUTOR if (!flag_net_injection) return; #endif tunfd = open("/dev/net/tun", O_RDWR | O_NONBLOCK); if (tunfd == -1) { #if SYZ_EXECUTOR fail("tun: can't open /dev/net/tun"); #else printf("tun: can't open /dev/net/tun: please enable CONFIG_TUN=y\n"); printf("otherwise fuzzing or reproducing might not work as intended\n"); return; #endif } // Remap tun onto higher fd number to hide it from fuzzer and to keep // fd numbers stable regardless of whether tun is opened or not (also see kMaxFd). const int kTunFd = 200; if (dup2(tunfd, kTunFd) < 0) fail("dup2(tunfd, kTunFd) failed"); close(tunfd); tunfd = kTunFd; struct ifreq ifr; memset(&ifr, 0, sizeof(ifr)); strncpy(ifr.ifr_name, TUN_IFACE, IFNAMSIZ); ifr.ifr_flags = IFF_TAP | IFF_NO_PI; // Note: SYZ_ENABLE_NAPI_FRAGS is never enabled. This is code is only for reference // in case we figure out how IFF_NAPI_FRAGS works. With IFF_NAPI_FRAGS packets // don't reach destinations and bail out in udp_gro_receive (see #1594). // Also IFF_NAPI_FRAGS does not work with sandbox_namespace (see comment there). #if ENABLE_NAPI_FRAGS ifr.ifr_flags |= IFF_NAPI | IFF_NAPI_FRAGS; #endif if (ioctl(tunfd, TUNSETIFF, (void*)&ifr) < 0) { #if ENABLE_NAPI_FRAGS // IFF_NAPI_FRAGS requires root, so try without it. ifr.ifr_flags = IFF_TAP | IFF_NO_PI; if (ioctl(tunfd, TUNSETIFF, (void*)&ifr) < 0) #endif fail("tun: ioctl(TUNSETIFF) failed"); } #if ENABLE_NAPI_FRAGS // If IFF_NAPI_FRAGS is not supported it will be silently dropped, // so query the effective flags. if (ioctl(tunfd, TUNGETIFF, (void*)&ifr) < 0) fail("tun: ioctl(TUNGETIFF) failed"); tun_frags_enabled = (ifr.ifr_flags & IFF_NAPI_FRAGS) != 0; debug("tun_frags_enabled=%d\n", tun_frags_enabled); #endif // Disable IPv6 DAD, otherwise the address remains unusable until DAD completes. // Don't panic because this is an optional config. char sysctl[64]; sprintf(sysctl, "/proc/sys/net/ipv6/conf/%s/accept_dad", TUN_IFACE); write_file(sysctl, "0"); // Disable IPv6 router solicitation to prevent IPv6 spam. // Don't panic because this is an optional config. sprintf(sysctl, "/proc/sys/net/ipv6/conf/%s/router_solicitations", TUN_IFACE); write_file(sysctl, "0"); // There seems to be no way to disable IPv6 MTD to prevent more IPv6 spam. int sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_ROUTE); if (sock == -1) fail("socket(AF_NETLINK) failed"); netlink_add_addr4(&nlmsg, sock, TUN_IFACE, LOCAL_IPV4); netlink_add_addr6(&nlmsg, sock, TUN_IFACE, LOCAL_IPV6); uint64 macaddr = REMOTE_MAC; struct in_addr in_addr; inet_pton(AF_INET, REMOTE_IPV4, &in_addr); netlink_add_neigh(&nlmsg, sock, TUN_IFACE, &in_addr, sizeof(in_addr), &macaddr, ETH_ALEN); struct in6_addr in6_addr; inet_pton(AF_INET6, REMOTE_IPV6, &in6_addr); netlink_add_neigh(&nlmsg, sock, TUN_IFACE, &in6_addr, sizeof(in6_addr), &macaddr, ETH_ALEN); macaddr = LOCAL_MAC; netlink_device_change(&nlmsg, sock, TUN_IFACE, true, 0, &macaddr, ETH_ALEN, NULL); close(sock); } #endif #if SYZ_EXECUTOR || __NR_syz_init_net_socket || SYZ_DEVLINK_PCI || __NR_syz_socket_connect_nvme_tcp const int kInitNetNsFd = 201; // see kMaxFd #endif #if SYZ_EXECUTOR || SYZ_DEVLINK_PCI || SYZ_NET_DEVICES #include #include #define DEVLINK_FAMILY_NAME "devlink" #define DEVLINK_CMD_PORT_GET 5 #if SYZ_EXECUTOR || SYZ_DEVLINK_PCI #define DEVLINK_CMD_RELOAD 37 #endif #define DEVLINK_ATTR_BUS_NAME 1 #define DEVLINK_ATTR_DEV_NAME 2 #define DEVLINK_ATTR_NETDEV_NAME 7 #if SYZ_EXECUTOR || SYZ_DEVLINK_PCI #define DEVLINK_ATTR_NETNS_FD 138 #endif #if SYZ_EXECUTOR || SYZ_DEVLINK_PCI static void netlink_devlink_netns_move(const char* bus_name, const char* dev_name, int netns_fd) { struct genlmsghdr genlhdr; int sock; int id, err; sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_GENERIC); if (sock == -1) fail("socket(AF_NETLINK) failed"); id = netlink_query_family_id(&nlmsg, sock, DEVLINK_FAMILY_NAME, true); if (id == -1) goto error; memset(&genlhdr, 0, sizeof(genlhdr)); genlhdr.cmd = DEVLINK_CMD_RELOAD; netlink_init(&nlmsg, id, 0, &genlhdr, sizeof(genlhdr)); netlink_attr(&nlmsg, DEVLINK_ATTR_BUS_NAME, bus_name, strlen(bus_name) + 1); netlink_attr(&nlmsg, DEVLINK_ATTR_DEV_NAME, dev_name, strlen(dev_name) + 1); netlink_attr(&nlmsg, DEVLINK_ATTR_NETNS_FD, &netns_fd, sizeof(netns_fd)); err = netlink_send(&nlmsg, sock); if (err < 0) { debug("netlink: failed to move devlink instance %s/%s into network namespace: %s\n", bus_name, dev_name, strerror(errno)); } error: close(sock); } #endif static struct nlmsg nlmsg2; static void initialize_devlink_ports(const char* bus_name, const char* dev_name, const char* netdev_prefix) { struct genlmsghdr genlhdr; int len, total_len, id, err, offset; uint16 netdev_index; int sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_GENERIC); if (sock == -1) fail("socket(AF_NETLINK) failed"); int rtsock = socket(AF_NETLINK, SOCK_RAW, NETLINK_ROUTE); if (rtsock == -1) fail("socket(AF_NETLINK) failed"); id = netlink_query_family_id(&nlmsg, sock, DEVLINK_FAMILY_NAME, true); if (id == -1) goto error; memset(&genlhdr, 0, sizeof(genlhdr)); genlhdr.cmd = DEVLINK_CMD_PORT_GET; netlink_init(&nlmsg, id, NLM_F_DUMP, &genlhdr, sizeof(genlhdr)); netlink_attr(&nlmsg, DEVLINK_ATTR_BUS_NAME, bus_name, strlen(bus_name) + 1); netlink_attr(&nlmsg, DEVLINK_ATTR_DEV_NAME, dev_name, strlen(dev_name) + 1); err = netlink_send_ext(&nlmsg, sock, id, &total_len, true); if (err < 0) { debug("netlink: failed to get port get reply: %s\n", strerror(errno)); goto error; } offset = 0; netdev_index = 0; while ((len = netlink_next_msg(&nlmsg, offset, total_len)) != -1) { struct nlattr* attr = (struct nlattr*)(nlmsg.buf + offset + NLMSG_HDRLEN + NLMSG_ALIGN(sizeof(genlhdr))); for (; (char*)attr < nlmsg.buf + offset + len; attr = (struct nlattr*)((char*)attr + NLMSG_ALIGN(attr->nla_len))) { if (attr->nla_type == DEVLINK_ATTR_NETDEV_NAME) { char* port_name; char netdev_name[IFNAMSIZ]; port_name = (char*)(attr + 1); snprintf(netdev_name, sizeof(netdev_name), "%s%d", netdev_prefix, netdev_index); netlink_device_change(&nlmsg2, rtsock, port_name, true, 0, 0, 0, netdev_name); break; } } offset += len; netdev_index++; } error: close(rtsock); close(sock); } #if SYZ_EXECUTOR || SYZ_DEVLINK_PCI #include #include static void initialize_devlink_pci(void) { #if SYZ_EXECUTOR if (!flag_devlink_pci) return; #endif int netns = open("/proc/self/ns/net", O_RDONLY); if (netns == -1) fail("open(/proc/self/ns/net) failed"); int ret = setns(kInitNetNsFd, 0); if (ret == -1) fail("set_ns(init_netns_fd) failed"); netlink_devlink_netns_move("pci", "0000:00:10.0", netns); ret = setns(netns, 0); if (ret == -1) fail("set_ns(this_netns_fd) failed"); close(netns); initialize_devlink_ports("pci", "0000:00:10.0", "netpci"); } #endif #endif #if SYZ_EXECUTOR || SYZ_WIFI || __NR_syz_80211_inject_frame || __NR_syz_80211_join_ibss #define WIFI_INITIAL_DEVICE_COUNT 2 #define WIFI_MAC_BASE \ { \ 0x08, 0x02, 0x11, 0x00, 0x00, 0x00} #define WIFI_IBSS_BSSID \ { \ 0x50, 0x50, 0x50, 0x50, 0x50, 0x50} #define WIFI_IBSS_SSID \ { \ 0x10, 0x10, 0x10, 0x10, 0x10, 0x10} #define WIFI_DEFAULT_FREQUENCY 2412 #define WIFI_DEFAULT_SIGNAL 0 #define WIFI_DEFAULT_RX_RATE 1 // consts from drivers/net/wireless/mac80211_hwsim.h #define HWSIM_CMD_REGISTER 1 #define HWSIM_CMD_FRAME 2 #define HWSIM_CMD_NEW_RADIO 4 #define HWSIM_ATTR_SUPPORT_P2P_DEVICE 14 #define HWSIM_ATTR_PERM_ADDR 22 #endif #if SYZ_EXECUTOR || SYZ_WIFI || __NR_syz_80211_join_ibss #include #include #include #include #include #include #include // From linux/if.h, but we cannot include the file as it conflicts with net/if.h #define IF_OPER_UP 6 // IBSS parameters for nl80211_join_ibss struct join_ibss_props { int wiphy_freq; bool wiphy_freq_fixed; uint8* mac; uint8* ssid; int ssid_len; }; static int set_interface_state(const char* interface_name, int on) { struct ifreq ifr; int sock = socket(AF_INET, SOCK_DGRAM, 0); if (sock < 0) { debug("set_interface_state: failed to open socket, errno %d\n", errno); return -1; } memset(&ifr, 0, sizeof(ifr)); strcpy(ifr.ifr_name, interface_name); int ret = ioctl(sock, SIOCGIFFLAGS, &ifr); if (ret < 0) { debug("set_interface_state: failed to execute SIOCGIFFLAGS, ret %d\n", ret); close(sock); return -1; } if (on) ifr.ifr_flags |= IFF_UP; else ifr.ifr_flags &= ~IFF_UP; ret = ioctl(sock, SIOCSIFFLAGS, &ifr); close(sock); if (ret < 0) { debug("set_interface_state: failed to execute SIOCSIFFLAGS, ret %d\n", ret); return -1; } return 0; } static int nl80211_set_interface(struct nlmsg* nlmsg, int sock, int nl80211_family, uint32 ifindex, uint32 iftype, bool dofail) { struct genlmsghdr genlhdr; memset(&genlhdr, 0, sizeof(genlhdr)); genlhdr.cmd = NL80211_CMD_SET_INTERFACE; netlink_init(nlmsg, nl80211_family, 0, &genlhdr, sizeof(genlhdr)); netlink_attr(nlmsg, NL80211_ATTR_IFINDEX, &ifindex, sizeof(ifindex)); netlink_attr(nlmsg, NL80211_ATTR_IFTYPE, &iftype, sizeof(iftype)); int err = netlink_send_ext(nlmsg, sock, 0, NULL, dofail); if (err < 0) { debug("nl80211_set_interface failed: %s\n", strerror(errno)); } return err; } static int nl80211_join_ibss(struct nlmsg* nlmsg, int sock, int nl80211_family, uint32 ifindex, struct join_ibss_props* props, bool dofail) { struct genlmsghdr genlhdr; memset(&genlhdr, 0, sizeof(genlhdr)); genlhdr.cmd = NL80211_CMD_JOIN_IBSS; netlink_init(nlmsg, nl80211_family, 0, &genlhdr, sizeof(genlhdr)); netlink_attr(nlmsg, NL80211_ATTR_IFINDEX, &ifindex, sizeof(ifindex)); netlink_attr(nlmsg, NL80211_ATTR_SSID, props->ssid, props->ssid_len); netlink_attr(nlmsg, NL80211_ATTR_WIPHY_FREQ, &(props->wiphy_freq), sizeof(props->wiphy_freq)); if (props->mac) netlink_attr(nlmsg, NL80211_ATTR_MAC, props->mac, ETH_ALEN); if (props->wiphy_freq_fixed) netlink_attr(nlmsg, NL80211_ATTR_FREQ_FIXED, NULL, 0); int err = netlink_send_ext(nlmsg, sock, 0, NULL, dofail); if (err < 0) { debug("nl80211_join_ibss failed: %s\n", strerror(errno)); } return err; } static int get_ifla_operstate(struct nlmsg* nlmsg, int ifindex, bool dofail) { struct ifinfomsg info; memset(&info, 0, sizeof(info)); info.ifi_family = AF_UNSPEC; info.ifi_index = ifindex; int sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_ROUTE); if (sock == -1) { debug("get_ifla_operstate: socket failed: %d\n", errno); return -1; } netlink_init(nlmsg, RTM_GETLINK, 0, &info, sizeof(info)); int n; int err = netlink_send_ext(nlmsg, sock, RTM_NEWLINK, &n, dofail); close(sock); if (err) { debug("get_ifla_operstate: failed to query: %s\n", strerror(errno)); return -1; } struct rtattr* attr = IFLA_RTA(NLMSG_DATA(nlmsg->buf)); for (; RTA_OK(attr, n); attr = RTA_NEXT(attr, n)) { if (attr->rta_type == IFLA_OPERSTATE) return *((int32_t*)RTA_DATA(attr)); } return -1; } static int await_ifla_operstate(struct nlmsg* nlmsg, char* interface, int operstate, bool dofail) { int ifindex = if_nametoindex(interface); while (true) { usleep(1000); // 1 ms int ret = get_ifla_operstate(nlmsg, ifindex, dofail); if (ret < 0) return ret; if (ret == operstate) return 0; } return 0; } static int nl80211_setup_ibss_interface(struct nlmsg* nlmsg, int sock, int nl80211_family_id, char* interface, struct join_ibss_props* ibss_props, bool dofail) { int ifindex = if_nametoindex(interface); if (ifindex == 0) { debug("nl80211_setup_ibss_interface: if_nametoindex failed for %.32s, ret 0\n", interface); return -1; } int ret = nl80211_set_interface(nlmsg, sock, nl80211_family_id, ifindex, NL80211_IFTYPE_ADHOC, dofail); if (ret < 0) { debug("nl80211_setup_ibss_interface: nl80211_set_interface failed for %.32s, ret %d\n", interface, ret); return -1; } ret = set_interface_state(interface, 1); if (ret < 0) { debug("nl80211_setup_ibss_interface: set_interface_state failed for %.32s, ret %d\n", interface, ret); return -1; } ret = nl80211_join_ibss(nlmsg, sock, nl80211_family_id, ifindex, ibss_props, dofail); if (ret < 0) { debug("nl80211_setup_ibss_interface: nl80211_join_ibss failed for %.32s, ret %d\n", interface, ret); return -1; } return 0; } #endif #if SYZ_EXECUTOR || SYZ_WIFI #include #include #include #include static int hwsim80211_create_device(struct nlmsg* nlmsg, int sock, int hwsim_family, uint8 mac_addr[ETH_ALEN]) { struct genlmsghdr genlhdr; memset(&genlhdr, 0, sizeof(genlhdr)); genlhdr.cmd = HWSIM_CMD_NEW_RADIO; netlink_init(nlmsg, hwsim_family, 0, &genlhdr, sizeof(genlhdr)); netlink_attr(nlmsg, HWSIM_ATTR_SUPPORT_P2P_DEVICE, NULL, 0); netlink_attr(nlmsg, HWSIM_ATTR_PERM_ADDR, mac_addr, ETH_ALEN); int err = netlink_send(nlmsg, sock); if (err < 0) { debug("hwsim80211_create_device failed: %s\n", strerror(errno)); } return err; } static void initialize_wifi_devices(void) { // Set up virtual wifi devices and join them into an IBSS network. // An IBSS network is created here in order to put these devices in an operable state right from // the beginning. It has the following positive effects. // 1. Frame injection becomes possible from the very start. // 2. A number of nl80211 commands expect their target wireless interface to be in an operable state. // 3. Simplification of reproducer generation - in many cases the reproducer will not have to spend time // selecting system calls that set up the environment. // // IBSS network was chosen as the simplest network type to begin with. #if SYZ_EXECUTOR if (!flag_wifi) return; #endif int rfkill = open("/dev/rfkill", O_RDWR); if (rfkill == -1) fail("open(/dev/rfkill) failed"); struct rfkill_event event = {0}; event.type = RFKILL_TYPE_ALL; event.op = RFKILL_OP_CHANGE_ALL; if (write(rfkill, &event, sizeof(event)) != (ssize_t)(sizeof(event))) fail("write(/dev/rfkill) failed"); close(rfkill); uint8 mac_addr[6] = WIFI_MAC_BASE; int sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_GENERIC); if (sock < 0) fail("initialize_wifi_devices: failed to create socket"); int hwsim_family_id = netlink_query_family_id(&nlmsg, sock, "MAC80211_HWSIM", true); int nl80211_family_id = netlink_query_family_id(&nlmsg, sock, "nl80211", true); if (hwsim_family_id < 0 || nl80211_family_id < 0) fail("netlink_query_family_id failed"); uint8 ssid[] = WIFI_IBSS_SSID; uint8 bssid[] = WIFI_IBSS_BSSID; struct join_ibss_props ibss_props = { .wiphy_freq = WIFI_DEFAULT_FREQUENCY, .wiphy_freq_fixed = true, .mac = bssid, .ssid = ssid, .ssid_len = sizeof(ssid)}; for (int device_id = 0; device_id < WIFI_INITIAL_DEVICE_COUNT; device_id++) { // Virtual wifi devices will have consequtive mac addresses mac_addr[5] = device_id; int ret = hwsim80211_create_device(&nlmsg, sock, hwsim_family_id, mac_addr); if (ret < 0) failmsg("initialize_wifi_devices: failed to create device", "device=%d", device_id); // For each device, unless HWSIM_ATTR_NO_VIF is passed, a network interface is created // automatically. Such interfaces are named "wlan0", "wlan1" and so on. char interface[6] = "wlan0"; interface[4] += device_id; if (nl80211_setup_ibss_interface(&nlmsg, sock, nl80211_family_id, interface, &ibss_props, true) < 0) failmsg("initialize_wifi_devices: failed set up IBSS network", "device=%d", device_id); } // Wait for all devices to join the IBSS network for (int device_id = 0; device_id < WIFI_INITIAL_DEVICE_COUNT; device_id++) { char interface[6] = "wlan0"; interface[4] += device_id; int ret = await_ifla_operstate(&nlmsg, interface, IF_OPER_UP, true); if (ret < 0) failmsg("initialize_wifi_devices: get_ifla_operstate failed", "device=%d, ret=%d", device_id, ret); } close(sock); } #endif #if SYZ_EXECUTOR || (SYZ_NET_DEVICES && SYZ_NIC_VF) || SYZ_SWAP static int runcmdline(char* cmdline) { debug("%s\n", cmdline); int ret = system(cmdline); if (ret) { debug("FAIL: %s\n", cmdline); } return ret; } #endif #if SYZ_EXECUTOR || SYZ_NET_DEVICES #include #include #include #include #include #include #include #include #include #include #include #include #include #include // Addresses are chosen to be in the same subnet as tun addresses. #define DEV_IPV4 "172.20.20.%d" #define DEV_IPV6 "fe80::%02x" #define DEV_MAC 0x00aaaaaaaaaa static void netdevsim_add(unsigned int addr, unsigned int port_count) { // These devices are sticky and are not deleted on net namespace destruction. // So try to delete the previous version of the device. write_file("/sys/bus/netdevsim/del_device", "%u", addr); if (write_file("/sys/bus/netdevsim/new_device", "%u %u", addr, port_count)) { char buf[32]; snprintf(buf, sizeof(buf), "netdevsim%d", addr); initialize_devlink_ports("netdevsim", buf, "netdevsim"); } } #define WG_GENL_NAME "wireguard" enum wg_cmd { WG_CMD_GET_DEVICE, WG_CMD_SET_DEVICE, }; enum wgdevice_attribute { WGDEVICE_A_UNSPEC, WGDEVICE_A_IFINDEX, WGDEVICE_A_IFNAME, WGDEVICE_A_PRIVATE_KEY, WGDEVICE_A_PUBLIC_KEY, WGDEVICE_A_FLAGS, WGDEVICE_A_LISTEN_PORT, WGDEVICE_A_FWMARK, WGDEVICE_A_PEERS, }; enum wgpeer_attribute { WGPEER_A_UNSPEC, WGPEER_A_PUBLIC_KEY, WGPEER_A_PRESHARED_KEY, WGPEER_A_FLAGS, WGPEER_A_ENDPOINT, WGPEER_A_PERSISTENT_KEEPALIVE_INTERVAL, WGPEER_A_LAST_HANDSHAKE_TIME, WGPEER_A_RX_BYTES, WGPEER_A_TX_BYTES, WGPEER_A_ALLOWEDIPS, WGPEER_A_PROTOCOL_VERSION, }; enum wgallowedip_attribute { WGALLOWEDIP_A_UNSPEC, WGALLOWEDIP_A_FAMILY, WGALLOWEDIP_A_IPADDR, WGALLOWEDIP_A_CIDR_MASK, }; static void netlink_wireguard_setup(void) { const char ifname_a[] = "wg0"; const char ifname_b[] = "wg1"; const char ifname_c[] = "wg2"; const char private_a[] = "\xa0\x5c\xa8\x4f\x6c\x9c\x8e\x38\x53\xe2\xfd\x7a\x70\xae\x0f\xb2\x0f\xa1\x52\x60\x0c\xb0\x08\x45\x17\x4f\x08\x07\x6f\x8d\x78\x43"; const char private_b[] = "\xb0\x80\x73\xe8\xd4\x4e\x91\xe3\xda\x92\x2c\x22\x43\x82\x44\xbb\x88\x5c\x69\xe2\x69\xc8\xe9\xd8\x35\xb1\x14\x29\x3a\x4d\xdc\x6e"; const char private_c[] = "\xa0\xcb\x87\x9a\x47\xf5\xbc\x64\x4c\x0e\x69\x3f\xa6\xd0\x31\xc7\x4a\x15\x53\xb6\xe9\x01\xb9\xff\x2f\x51\x8c\x78\x04\x2f\xb5\x42"; const char public_a[] = "\x97\x5c\x9d\x81\xc9\x83\xc8\x20\x9e\xe7\x81\x25\x4b\x89\x9f\x8e\xd9\x25\xae\x9f\x09\x23\xc2\x3c\x62\xf5\x3c\x57\xcd\xbf\x69\x1c"; const char public_b[] = "\xd1\x73\x28\x99\xf6\x11\xcd\x89\x94\x03\x4d\x7f\x41\x3d\xc9\x57\x63\x0e\x54\x93\xc2\x85\xac\xa4\x00\x65\xcb\x63\x11\xbe\x69\x6b"; const char public_c[] = "\xf4\x4d\xa3\x67\xa8\x8e\xe6\x56\x4f\x02\x02\x11\x45\x67\x27\x08\x2f\x5c\xeb\xee\x8b\x1b\xf5\xeb\x73\x37\x34\x1b\x45\x9b\x39\x22"; const uint16 listen_a = 20001; const uint16 listen_b = 20002; const uint16 listen_c = 20003; const uint16 af_inet = AF_INET; const uint16 af_inet6 = AF_INET6; // Unused, but useful in case we change this: // const struct sockaddr_in endpoint_a_v4 = { // .sin_family = AF_INET, // .sin_port = htons(listen_a), // .sin_addr = {htonl(INADDR_LOOPBACK)}}; const struct sockaddr_in endpoint_b_v4 = { .sin_family = AF_INET, .sin_port = htons(listen_b), .sin_addr = {htonl(INADDR_LOOPBACK)}}; const struct sockaddr_in endpoint_c_v4 = { .sin_family = AF_INET, .sin_port = htons(listen_c), .sin_addr = {htonl(INADDR_LOOPBACK)}}; struct sockaddr_in6 endpoint_a_v6 = { .sin6_family = AF_INET6, .sin6_port = htons(listen_a)}; endpoint_a_v6.sin6_addr = in6addr_loopback; // Unused, but useful in case we change this: // const struct sockaddr_in6 endpoint_b_v6 = { // .sin6_family = AF_INET6, // .sin6_port = htons(listen_b)}; // endpoint_b_v6.sin6_addr = in6addr_loopback; struct sockaddr_in6 endpoint_c_v6 = { .sin6_family = AF_INET6, .sin6_port = htons(listen_c)}; endpoint_c_v6.sin6_addr = in6addr_loopback; const struct in_addr first_half_v4 = {0}; const struct in_addr second_half_v4 = {(uint32)htonl(128 << 24)}; const struct in6_addr first_half_v6 = {{{0}}}; const struct in6_addr second_half_v6 = {{{0x80}}}; const uint8 half_cidr = 1; const uint16 persistent_keepalives[] = {1, 3, 7, 9, 14, 19}; struct genlmsghdr genlhdr = { .cmd = WG_CMD_SET_DEVICE, .version = 1}; int sock; int id, err; sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_GENERIC); if (sock == -1) { debug("socket(AF_NETLINK) failed: %s\n", strerror(errno)); return; } id = netlink_query_family_id(&nlmsg, sock, WG_GENL_NAME, true); if (id == -1) goto error; netlink_init(&nlmsg, id, 0, &genlhdr, sizeof(genlhdr)); netlink_attr(&nlmsg, WGDEVICE_A_IFNAME, ifname_a, strlen(ifname_a) + 1); netlink_attr(&nlmsg, WGDEVICE_A_PRIVATE_KEY, private_a, 32); netlink_attr(&nlmsg, WGDEVICE_A_LISTEN_PORT, &listen_a, 2); netlink_nest(&nlmsg, NLA_F_NESTED | WGDEVICE_A_PEERS); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGPEER_A_PUBLIC_KEY, public_b, 32); netlink_attr(&nlmsg, WGPEER_A_ENDPOINT, &endpoint_b_v4, sizeof(endpoint_b_v4)); netlink_attr(&nlmsg, WGPEER_A_PERSISTENT_KEEPALIVE_INTERVAL, &persistent_keepalives[0], 2); netlink_nest(&nlmsg, NLA_F_NESTED | WGPEER_A_ALLOWEDIPS); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGALLOWEDIP_A_FAMILY, &af_inet, 2); netlink_attr(&nlmsg, WGALLOWEDIP_A_IPADDR, &first_half_v4, sizeof(first_half_v4)); netlink_attr(&nlmsg, WGALLOWEDIP_A_CIDR_MASK, &half_cidr, 1); netlink_done(&nlmsg); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGALLOWEDIP_A_FAMILY, &af_inet6, 2); netlink_attr(&nlmsg, WGALLOWEDIP_A_IPADDR, &first_half_v6, sizeof(first_half_v6)); netlink_attr(&nlmsg, WGALLOWEDIP_A_CIDR_MASK, &half_cidr, 1); netlink_done(&nlmsg); netlink_done(&nlmsg); netlink_done(&nlmsg); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGPEER_A_PUBLIC_KEY, public_c, 32); netlink_attr(&nlmsg, WGPEER_A_ENDPOINT, &endpoint_c_v6, sizeof(endpoint_c_v6)); netlink_attr(&nlmsg, WGPEER_A_PERSISTENT_KEEPALIVE_INTERVAL, &persistent_keepalives[1], 2); netlink_nest(&nlmsg, NLA_F_NESTED | WGPEER_A_ALLOWEDIPS); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGALLOWEDIP_A_FAMILY, &af_inet, 2); netlink_attr(&nlmsg, WGALLOWEDIP_A_IPADDR, &second_half_v4, sizeof(second_half_v4)); netlink_attr(&nlmsg, WGALLOWEDIP_A_CIDR_MASK, &half_cidr, 1); netlink_done(&nlmsg); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGALLOWEDIP_A_FAMILY, &af_inet6, 2); netlink_attr(&nlmsg, WGALLOWEDIP_A_IPADDR, &second_half_v6, sizeof(second_half_v6)); netlink_attr(&nlmsg, WGALLOWEDIP_A_CIDR_MASK, &half_cidr, 1); netlink_done(&nlmsg); netlink_done(&nlmsg); netlink_done(&nlmsg); netlink_done(&nlmsg); err = netlink_send(&nlmsg, sock); if (err < 0) { debug("netlink: failed to setup wireguard instance: %s\n", strerror(errno)); } netlink_init(&nlmsg, id, 0, &genlhdr, sizeof(genlhdr)); netlink_attr(&nlmsg, WGDEVICE_A_IFNAME, ifname_b, strlen(ifname_b) + 1); netlink_attr(&nlmsg, WGDEVICE_A_PRIVATE_KEY, private_b, 32); netlink_attr(&nlmsg, WGDEVICE_A_LISTEN_PORT, &listen_b, 2); netlink_nest(&nlmsg, NLA_F_NESTED | WGDEVICE_A_PEERS); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGPEER_A_PUBLIC_KEY, public_a, 32); netlink_attr(&nlmsg, WGPEER_A_ENDPOINT, &endpoint_a_v6, sizeof(endpoint_a_v6)); netlink_attr(&nlmsg, WGPEER_A_PERSISTENT_KEEPALIVE_INTERVAL, &persistent_keepalives[2], 2); netlink_nest(&nlmsg, NLA_F_NESTED | WGPEER_A_ALLOWEDIPS); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGALLOWEDIP_A_FAMILY, &af_inet, 2); netlink_attr(&nlmsg, WGALLOWEDIP_A_IPADDR, &first_half_v4, sizeof(first_half_v4)); netlink_attr(&nlmsg, WGALLOWEDIP_A_CIDR_MASK, &half_cidr, 1); netlink_done(&nlmsg); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGALLOWEDIP_A_FAMILY, &af_inet6, 2); netlink_attr(&nlmsg, WGALLOWEDIP_A_IPADDR, &first_half_v6, sizeof(first_half_v6)); netlink_attr(&nlmsg, WGALLOWEDIP_A_CIDR_MASK, &half_cidr, 1); netlink_done(&nlmsg); netlink_done(&nlmsg); netlink_done(&nlmsg); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGPEER_A_PUBLIC_KEY, public_c, 32); netlink_attr(&nlmsg, WGPEER_A_ENDPOINT, &endpoint_c_v4, sizeof(endpoint_c_v4)); netlink_attr(&nlmsg, WGPEER_A_PERSISTENT_KEEPALIVE_INTERVAL, &persistent_keepalives[3], 2); netlink_nest(&nlmsg, NLA_F_NESTED | WGPEER_A_ALLOWEDIPS); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGALLOWEDIP_A_FAMILY, &af_inet, 2); netlink_attr(&nlmsg, WGALLOWEDIP_A_IPADDR, &second_half_v4, sizeof(second_half_v4)); netlink_attr(&nlmsg, WGALLOWEDIP_A_CIDR_MASK, &half_cidr, 1); netlink_done(&nlmsg); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGALLOWEDIP_A_FAMILY, &af_inet6, 2); netlink_attr(&nlmsg, WGALLOWEDIP_A_IPADDR, &second_half_v6, sizeof(second_half_v6)); netlink_attr(&nlmsg, WGALLOWEDIP_A_CIDR_MASK, &half_cidr, 1); netlink_done(&nlmsg); netlink_done(&nlmsg); netlink_done(&nlmsg); netlink_done(&nlmsg); err = netlink_send(&nlmsg, sock); if (err < 0) { debug("netlink: failed to setup wireguard instance: %s\n", strerror(errno)); } netlink_init(&nlmsg, id, 0, &genlhdr, sizeof(genlhdr)); netlink_attr(&nlmsg, WGDEVICE_A_IFNAME, ifname_c, strlen(ifname_c) + 1); netlink_attr(&nlmsg, WGDEVICE_A_PRIVATE_KEY, private_c, 32); netlink_attr(&nlmsg, WGDEVICE_A_LISTEN_PORT, &listen_c, 2); netlink_nest(&nlmsg, NLA_F_NESTED | WGDEVICE_A_PEERS); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGPEER_A_PUBLIC_KEY, public_a, 32); netlink_attr(&nlmsg, WGPEER_A_ENDPOINT, &endpoint_a_v6, sizeof(endpoint_a_v6)); netlink_attr(&nlmsg, WGPEER_A_PERSISTENT_KEEPALIVE_INTERVAL, &persistent_keepalives[4], 2); netlink_nest(&nlmsg, NLA_F_NESTED | WGPEER_A_ALLOWEDIPS); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGALLOWEDIP_A_FAMILY, &af_inet, 2); netlink_attr(&nlmsg, WGALLOWEDIP_A_IPADDR, &first_half_v4, sizeof(first_half_v4)); netlink_attr(&nlmsg, WGALLOWEDIP_A_CIDR_MASK, &half_cidr, 1); netlink_done(&nlmsg); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGALLOWEDIP_A_FAMILY, &af_inet6, 2); netlink_attr(&nlmsg, WGALLOWEDIP_A_IPADDR, &first_half_v6, sizeof(first_half_v6)); netlink_attr(&nlmsg, WGALLOWEDIP_A_CIDR_MASK, &half_cidr, 1); netlink_done(&nlmsg); netlink_done(&nlmsg); netlink_done(&nlmsg); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGPEER_A_PUBLIC_KEY, public_b, 32); netlink_attr(&nlmsg, WGPEER_A_ENDPOINT, &endpoint_b_v4, sizeof(endpoint_b_v4)); netlink_attr(&nlmsg, WGPEER_A_PERSISTENT_KEEPALIVE_INTERVAL, &persistent_keepalives[5], 2); netlink_nest(&nlmsg, NLA_F_NESTED | WGPEER_A_ALLOWEDIPS); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGALLOWEDIP_A_FAMILY, &af_inet, 2); netlink_attr(&nlmsg, WGALLOWEDIP_A_IPADDR, &second_half_v4, sizeof(second_half_v4)); netlink_attr(&nlmsg, WGALLOWEDIP_A_CIDR_MASK, &half_cidr, 1); netlink_done(&nlmsg); netlink_nest(&nlmsg, NLA_F_NESTED | 0); netlink_attr(&nlmsg, WGALLOWEDIP_A_FAMILY, &af_inet6, 2); netlink_attr(&nlmsg, WGALLOWEDIP_A_IPADDR, &second_half_v6, sizeof(second_half_v6)); netlink_attr(&nlmsg, WGALLOWEDIP_A_CIDR_MASK, &half_cidr, 1); netlink_done(&nlmsg); netlink_done(&nlmsg); netlink_done(&nlmsg); netlink_done(&nlmsg); err = netlink_send(&nlmsg, sock); if (err < 0) { debug("netlink: failed to setup wireguard instance: %s\n", strerror(errno)); } error: close(sock); } #if SYZ_EXECUTOR || SYZ_NIC_VF static void netlink_nicvf_setup(void) { char cmdline[256]; #if SYZ_EXECUTOR if (!flag_nic_vf) return; #endif if (!vf_intf.ppid) return; debug("ppid = %d, vf_intf.pass_thru_intf: %s\n", vf_intf.ppid, vf_intf.pass_thru_intf); sprintf(cmdline, "nsenter -t 1 -n ip link set %s netns %d", vf_intf.pass_thru_intf, getpid()); if (runcmdline(cmdline)) failmsg("failed to run command", "%s", cmdline); sprintf(cmdline, "ip a s %s", vf_intf.pass_thru_intf); if (runcmdline(cmdline)) failmsg("failed to run command", "%s", cmdline); sprintf(cmdline, "ip link set %s down", vf_intf.pass_thru_intf); if (runcmdline(cmdline)) failmsg("failed to run command", "%s", cmdline); sprintf(cmdline, "ip link set %s name nicvf0", vf_intf.pass_thru_intf); if (runcmdline(cmdline)) failmsg("failed to run command", "%s", cmdline); debug("nicvf0 VF pass-through setup complete.\n"); } #endif // SYZ_NIC_VF // We test in a separate namespace, which does not have any network devices initially (even lo). // Create/up as many as we can. static void initialize_netdevices(void) { #if SYZ_EXECUTOR if (!flag_net_devices) return; #endif // TODO: add the following devices: // - vxlan // - ipip // - lowpan (requires link to device of type IEEE802154, e.g. wpan0) // - ipoib (requires link to device of type ARPHRD_INFINIBAND) // - vrf // - rmnet // - openvswitch // Naive attempts to add devices of these types fail with various errors. // Also init namespace contains the following devices (which presumably can't be // created in non-init namespace), can we use them somehow? // - ifb0/1 // - teql0 // - eql // Note: netdevsim devices can't have the same name even in different namespaces. char netdevsim[16]; sprintf(netdevsim, "netdevsim%d", (int)procid); struct { const char* type; const char* dev; } devtypes[] = { // Note: ip6erspan device can't be added if ip6gretap exists in the same namespace. {"ip6gretap", "ip6gretap0"}, {"bridge", "bridge0"}, {"vcan", "vcan0"}, {"bond", "bond0"}, {"team", "team0"}, {"dummy", "dummy0"}, #if SYZ_EXECUTOR || SYZ_NIC_VF {"nicvf", "nicvf0"}, #endif {"nlmon", "nlmon0"}, {"caif", "caif0"}, {"batadv", "batadv0"}, // Note: this adds vxcan0/vxcan1 pair, similar to veth (creating vxcan0 would fail). {"vxcan", "vxcan1"}, // This adds connected veth0 and veth1 devices. {"veth", 0}, {"wireguard", "wg0"}, {"wireguard", "wg1"}, {"wireguard", "wg2"}, }; const char* devmasters[] = {"bridge", "bond", "team", "batadv"}; // If you extend this array, also update netdev_addr_id in vnet.txt // and devnames in socket.txt. struct { const char* name; int macsize; bool noipv6; } devices[] = { {"lo", ETH_ALEN}, {"sit0", 0}, {"bridge0", ETH_ALEN}, {"vcan0", 0, true}, {"tunl0", 0}, {"gre0", 0}, {"gretap0", ETH_ALEN}, {"ip_vti0", 0}, {"ip6_vti0", 0}, {"ip6tnl0", 0}, {"ip6gre0", 0}, {"ip6gretap0", ETH_ALEN}, {"erspan0", ETH_ALEN}, {"bond0", ETH_ALEN}, {"veth0", ETH_ALEN}, {"veth1", ETH_ALEN}, {"team0", ETH_ALEN}, {"veth0_to_bridge", ETH_ALEN}, {"veth1_to_bridge", ETH_ALEN}, {"veth0_to_bond", ETH_ALEN}, {"veth1_to_bond", ETH_ALEN}, {"veth0_to_team", ETH_ALEN}, {"veth1_to_team", ETH_ALEN}, {"veth0_to_hsr", ETH_ALEN}, {"veth1_to_hsr", ETH_ALEN}, {"hsr0", 0}, {"dummy0", ETH_ALEN}, #if SYZ_EXECUTOR || SYZ_NIC_VF {"nicvf0", 0, true}, #endif {"nlmon0", 0}, {"vxcan0", 0, true}, {"vxcan1", 0, true}, {"caif0", ETH_ALEN}, // TODO: up'ing caif fails with ENODEV {"batadv0", ETH_ALEN}, {netdevsim, ETH_ALEN}, {"xfrm0", ETH_ALEN}, {"veth0_virt_wifi", ETH_ALEN}, {"veth1_virt_wifi", ETH_ALEN}, {"virt_wifi0", ETH_ALEN}, {"veth0_vlan", ETH_ALEN}, {"veth1_vlan", ETH_ALEN}, {"vlan0", ETH_ALEN}, {"vlan1", ETH_ALEN}, {"macvlan0", ETH_ALEN}, {"macvlan1", ETH_ALEN}, {"ipvlan0", ETH_ALEN}, {"ipvlan1", ETH_ALEN}, {"veth0_macvtap", ETH_ALEN}, {"veth1_macvtap", ETH_ALEN}, {"macvtap0", ETH_ALEN}, {"macsec0", ETH_ALEN}, {"veth0_to_batadv", ETH_ALEN}, {"veth1_to_batadv", ETH_ALEN}, {"batadv_slave_0", ETH_ALEN}, {"batadv_slave_1", ETH_ALEN}, {"geneve0", ETH_ALEN}, {"geneve1", ETH_ALEN}, {"wg0", 0}, {"wg1", 0}, {"wg2", 0}, }; int sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_ROUTE); if (sock == -1) fail("socket(AF_NETLINK) failed"); unsigned i; for (i = 0; i < sizeof(devtypes) / sizeof(devtypes[0]); i++) netlink_add_device(&nlmsg, sock, devtypes[i].type, devtypes[i].dev); // This creates connected bridge/bond/team_slave devices of type veth, // and makes them slaves of bridge/bond/team devices, respectively. // Note: slave devices don't need MAC/IP addresses, only master devices. // veth0_to_* is not slave devices, which still need ip addresses. for (i = 0; i < sizeof(devmasters) / (sizeof(devmasters[0])); i++) { char master[32], slave0[32], veth0[32], slave1[32], veth1[32]; sprintf(slave0, "%s_slave_0", devmasters[i]); sprintf(veth0, "veth0_to_%s", devmasters[i]); netlink_add_veth(&nlmsg, sock, slave0, veth0); sprintf(slave1, "%s_slave_1", devmasters[i]); sprintf(veth1, "veth1_to_%s", devmasters[i]); netlink_add_veth(&nlmsg, sock, slave1, veth1); sprintf(master, "%s0", devmasters[i]); netlink_device_change(&nlmsg, sock, slave0, false, master, 0, 0, NULL); netlink_device_change(&nlmsg, sock, slave1, false, master, 0, 0, NULL); } netlink_add_xfrm(&nlmsg, sock, "xfrm0"); // bond/team_slave_* will set up automatically when set their master. // But bridge_slave_* need to set up manually. netlink_device_change(&nlmsg, sock, "bridge_slave_0", true, 0, 0, 0, NULL); netlink_device_change(&nlmsg, sock, "bridge_slave_1", true, 0, 0, 0, NULL); // Setup hsr device (slightly different from what we do for devmasters). netlink_add_veth(&nlmsg, sock, "hsr_slave_0", "veth0_to_hsr"); netlink_add_veth(&nlmsg, sock, "hsr_slave_1", "veth1_to_hsr"); netlink_add_hsr(&nlmsg, sock, "hsr0", "hsr_slave_0", "hsr_slave_1"); netlink_device_change(&nlmsg, sock, "hsr_slave_0", true, 0, 0, 0, NULL); netlink_device_change(&nlmsg, sock, "hsr_slave_1", true, 0, 0, 0, NULL); netlink_add_veth(&nlmsg, sock, "veth0_virt_wifi", "veth1_virt_wifi"); netlink_add_linked(&nlmsg, sock, "virt_wifi", "virt_wifi0", "veth1_virt_wifi"); netlink_add_veth(&nlmsg, sock, "veth0_vlan", "veth1_vlan"); netlink_add_vlan(&nlmsg, sock, "vlan0", "veth0_vlan", 0, htons(ETH_P_8021Q)); netlink_add_vlan(&nlmsg, sock, "vlan1", "veth0_vlan", 1, htons(ETH_P_8021AD)); netlink_add_macvlan(&nlmsg, sock, "macvlan0", "veth1_vlan"); netlink_add_macvlan(&nlmsg, sock, "macvlan1", "veth1_vlan"); netlink_add_ipvlan(&nlmsg, sock, "ipvlan0", "veth0_vlan", IPVLAN_MODE_L2, 0); netlink_add_ipvlan(&nlmsg, sock, "ipvlan1", "veth0_vlan", IPVLAN_MODE_L3S, IPVLAN_F_VEPA); netlink_add_veth(&nlmsg, sock, "veth0_macvtap", "veth1_macvtap"); netlink_add_linked(&nlmsg, sock, "macvtap", "macvtap0", "veth0_macvtap"); netlink_add_linked(&nlmsg, sock, "macsec", "macsec0", "veth1_macvtap"); char addr[32]; sprintf(addr, DEV_IPV4, 14 + 10); // should point to veth0 struct in_addr geneve_addr4; if (inet_pton(AF_INET, addr, &geneve_addr4) <= 0) fail("geneve0 inet_pton failed"); struct in6_addr geneve_addr6; // Must not be link local (our device addresses are link local). if (inet_pton(AF_INET6, "fc00::01", &geneve_addr6) <= 0) fail("geneve1 inet_pton failed"); netlink_add_geneve(&nlmsg, sock, "geneve0", 0, &geneve_addr4, 0); netlink_add_geneve(&nlmsg, sock, "geneve1", 1, 0, &geneve_addr6); netdevsim_add((int)procid, 4); // Number of port is in sync with value in sys/linux/socket_netlink_generic_devlink.txt netlink_wireguard_setup(); #if SYZ_EXECUTOR || SYZ_NIC_VF netlink_nicvf_setup(); #endif for (i = 0; i < sizeof(devices) / (sizeof(devices[0])); i++) { // Assign some unique address to devices. Some devices won't up without this. // Shift addresses by 10 because 0 subnet address can mean special things. char addr[32]; sprintf(addr, DEV_IPV4, i + 10); netlink_add_addr4(&nlmsg, sock, devices[i].name, addr); if (!devices[i].noipv6) { sprintf(addr, DEV_IPV6, i + 10); netlink_add_addr6(&nlmsg, sock, devices[i].name, addr); } uint64 macaddr = DEV_MAC + ((i + 10ull) << 40); netlink_device_change(&nlmsg, sock, devices[i].name, true, 0, &macaddr, devices[i].macsize, NULL); } close(sock); } // Same as initialize_netdevices, but called in init net namespace. static void initialize_netdevices_init(void) { #if SYZ_EXECUTOR if (!flag_net_devices) return; #endif int sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_ROUTE); if (sock == -1) fail("socket(AF_NETLINK) failed"); struct { const char* type; int macsize; bool noipv6; bool noup; } devtypes[] = { // NETROM device, see net/netrom/{af_netrom,nr_dev}.c {"nr", 7, true}, // ROSE device, see net/rose/{af_rose,rose_dev}.c // We don't up it yet because it crashes kernel right away: // https://groups.google.com/d/msg/syzkaller/v-4B3zoBC-4/02SCKEzJBwAJ {"rose", 5, true, true}, }; unsigned i; for (i = 0; i < sizeof(devtypes) / sizeof(devtypes[0]); i++) { char dev[32], addr[32]; sprintf(dev, "%s%d", devtypes[i].type, (int)procid); // Note: syscall descriptions know these addresses. sprintf(addr, "172.30.%d.%d", i, (int)procid + 1); netlink_add_addr4(&nlmsg, sock, dev, addr); if (!devtypes[i].noipv6) { sprintf(addr, "fe88::%02x:%02x", i, (int)procid + 1); netlink_add_addr6(&nlmsg, sock, dev, addr); } int macsize = devtypes[i].macsize; uint64 macaddr = 0xbbbbbb + ((unsigned long long)i << (8 * (macsize - 2))) + (procid << (8 * (macsize - 1))); netlink_device_change(&nlmsg, sock, dev, !devtypes[i].noup, 0, &macaddr, macsize, NULL); } close(sock); #if SYZ_EXECUTOR || SYZ_NIC_VF find_vf_interface(); #endif } #endif #if SYZ_EXECUTOR || SYZ_NET_INJECTION && (__NR_syz_extract_tcp_res || SYZ_REPEAT) #include static int read_tun(char* data, int size) { if (tunfd < 0) return -1; int rv = read(tunfd, data, size); if (rv < 0) { // EBADF can be returned if the test closes tunfd with close_range syscall. // Tun sometimes returns EBADFD, unclear if it's a kernel bug or not. if (errno == EAGAIN || errno == EBADF || errno == EBADFD) return -1; fail("tun read failed"); } return rv; } #endif #if SYZ_EXECUTOR || __NR_syz_emit_ethernet && SYZ_NET_INJECTION #include #include #if ENABLE_NAPI_FRAGS #define MAX_FRAGS 4 struct vnet_fragmentation { uint32 full; uint32 count; uint32 frags[MAX_FRAGS]; }; #endif static long syz_emit_ethernet(volatile long a0, volatile long a1, volatile long a2) { // syz_emit_ethernet(len len[packet], packet ptr[in, eth_packet], frags ptr[in, vnet_fragmentation, opt]) // vnet_fragmentation { // full int32[0:1] // count int32[1:4] // frags array[int32[0:4096], 4] // } if (tunfd < 0) return (uintptr_t)-1; uint32 length = a0; char* data = (char*)a1; debug_dump_data(data, length); #if ENABLE_NAPI_FRAGS struct vnet_fragmentation* frags = (struct vnet_fragmentation*)a2; struct iovec vecs[MAX_FRAGS + 1]; uint32 nfrags = 0; if (!tun_frags_enabled || frags == NULL) { vecs[nfrags].iov_base = data; vecs[nfrags].iov_len = length; nfrags++; } else { bool full = frags->full; uint32 count = frags->count; if (count > MAX_FRAGS) count = MAX_FRAGS; uint32 i; for (i = 0; i < count && length != 0; i++) { uint32 size = frags->frags[i]; if (size > length) size = length; vecs[nfrags].iov_base = data; vecs[nfrags].iov_len = size; nfrags++; data += size; length -= size; } if (length != 0 && (full || nfrags == 0)) { vecs[nfrags].iov_base = data; vecs[nfrags].iov_len = length; nfrags++; } } return writev(tunfd, vecs, nfrags); #else return write(tunfd, data, length); #endif } #endif #if SYZ_EXECUTOR || __NR_syz_io_uring_submit || __NR_syz_io_uring_complete || __NR_syz_io_uring_setup #define SIZEOF_IO_URING_SQE 64 #define SIZEOF_IO_URING_CQE 16 // Once a io_uring is set up by calling io_uring_setup, the offsets to the member fields // to be used on the mmap'ed area are set in structs io_sqring_offsets and io_cqring_offsets. // Except io_sqring_offsets.array, the offsets are static while all depend on how struct io_rings // is organized in code. The offsets can be marked as resources in syzkaller descriptions but // this makes it difficult to generate correct programs by the fuzzer. Thus, the offsets are // hard-coded here (and in the descriptions), and array offset is later computed once the number // of entries is available. Another way to obtain the offsets is to setup another io_uring here // and use what it returns. It is slower but might be more maintainable. #define SQ_HEAD_OFFSET 0 #define SQ_TAIL_OFFSET 64 #define SQ_RING_MASK_OFFSET 256 #define SQ_RING_ENTRIES_OFFSET 264 #define SQ_FLAGS_OFFSET 276 #define SQ_DROPPED_OFFSET 272 #define CQ_HEAD_OFFSET 128 #define CQ_TAIL_OFFSET 192 #define CQ_RING_MASK_OFFSET 260 #define CQ_RING_ENTRIES_OFFSET 268 #define CQ_RING_OVERFLOW_OFFSET 284 #define CQ_FLAGS_OFFSET 280 #define CQ_CQES_OFFSET 320 #if SYZ_EXECUTOR || __NR_syz_io_uring_complete // From linux/io_uring.h struct io_uring_cqe { uint64 user_data; uint32 res; uint32 flags; }; static long syz_io_uring_complete(volatile long a0) { // syzlang: syz_io_uring_complete(ring_ptr ring_ptr) // C: syz_io_uring_complete(char* ring_ptr) // It is not checked if the ring is empty // Cast to original char* ring_ptr = (char*)a0; // Compute the head index and the next head value uint32 cq_ring_mask = *(uint32*)(ring_ptr + CQ_RING_MASK_OFFSET); uint32* cq_head_ptr = (uint32*)(ring_ptr + CQ_HEAD_OFFSET); uint32 cq_head = *cq_head_ptr & cq_ring_mask; uint32 cq_head_next = *cq_head_ptr + 1; // Compute the ptr to the src cq entry on the ring char* cqe_src = ring_ptr + CQ_CQES_OFFSET + cq_head * SIZEOF_IO_URING_CQE; // Get the cq entry from the ring struct io_uring_cqe cqe; memcpy(&cqe, cqe_src, sizeof(cqe)); // Advance the head. Head is a free-flowing integer and relies on natural wrapping. // Ensure that the kernel will never see a head update without the preceeding CQE // stores being done. __atomic_store_n(cq_head_ptr, cq_head_next, __ATOMIC_RELEASE); // In the descriptions (sys/linux/io_uring.txt), openat and openat2 are passed // with a unique range of sqe.user_data (0x12345 and 0x23456) to identify the operations // which produces an fd instance. Check cqe.user_data, which should be the same // as sqe.user_data for that operation. If it falls in that unique range, return // cqe.res as fd. Otherwise, just return an invalid fd. return (cqe.user_data == 0x12345 || cqe.user_data == 0x23456) ? (long)cqe.res : (long)-1; } #endif #if SYZ_EXECUTOR || __NR_syz_io_uring_setup struct io_sqring_offsets { uint32 head; uint32 tail; uint32 ring_mask; uint32 ring_entries; uint32 flags; uint32 dropped; uint32 array; uint32 resv1; uint64 resv2; }; struct io_cqring_offsets { uint32 head; uint32 tail; uint32 ring_mask; uint32 ring_entries; uint32 overflow; uint32 cqes; uint64 resv[2]; }; struct io_uring_params { uint32 sq_entries; uint32 cq_entries; uint32 flags; uint32 sq_thread_cpu; uint32 sq_thread_idle; uint32 features; uint32 resv[4]; struct io_sqring_offsets sq_off; struct io_cqring_offsets cq_off; }; #define IORING_OFF_SQ_RING 0 #define IORING_OFF_SQES 0x10000000ULL #define IORING_SETUP_SQE128 (1U << 10) #define IORING_SETUP_CQE32 (1U << 11) #include #include // Wrapper for io_uring_setup and the subsequent mmap calls that map the ring and the sqes static long syz_io_uring_setup(volatile long a0, volatile long a1, volatile long a2, volatile long a3) { // syzlang: syz_io_uring_setup(entries int32[1:IORING_MAX_ENTRIES], params ptr[inout, io_uring_params], ring_ptr ptr[out, ring_ptr], sqes_ptr ptr[out, sqes_ptr]) fd_io_uring // C: syz_io_uring_setup(uint32 entries, struct io_uring_params* params, void** ring_ptr_out, void** sqes_ptr_out) // returns uint32 fd_io_uring // Cast to original uint32 entries = (uint32)a0; struct io_uring_params* setup_params = (struct io_uring_params*)a1; void** ring_ptr_out = (void**)a2; void** sqes_ptr_out = (void**)a3; // Temporarily disable IORING_SETUP_CQE32 and IORING_SETUP_SQE128 that may change SIZEOF_IO_URING_CQE and SIZEOF_IO_URING_SQE. // Tracking bug: https://github.com/google/syzkaller/issues/4531. setup_params->flags &= ~(IORING_SETUP_CQE32 | IORING_SETUP_SQE128); uint32 fd_io_uring = syscall(__NR_io_uring_setup, entries, setup_params); // Compute the ring sizes uint32 sq_ring_sz = setup_params->sq_off.array + setup_params->sq_entries * sizeof(uint32); uint32 cq_ring_sz = setup_params->cq_off.cqes + setup_params->cq_entries * SIZEOF_IO_URING_CQE; // Asssumed IORING_FEAT_SINGLE_MMAP, which is always the case with the current implementation // The implication is that the sq_ring_ptr and the cq_ring_ptr are the same but the // difference is in the offsets to access the fields of these rings. uint32 ring_sz = sq_ring_sz > cq_ring_sz ? sq_ring_sz : cq_ring_sz; *ring_ptr_out = mmap(0, ring_sz, PROT_READ | PROT_WRITE, MAP_SHARED | MAP_POPULATE, fd_io_uring, IORING_OFF_SQ_RING); uint32 sqes_sz = setup_params->sq_entries * SIZEOF_IO_URING_SQE; *sqes_ptr_out = mmap(0, sqes_sz, PROT_READ | PROT_WRITE, MAP_SHARED | MAP_POPULATE, fd_io_uring, IORING_OFF_SQES); uint32* array = (uint32*)((uintptr_t)*ring_ptr_out + setup_params->sq_off.array); for (uint32 index = 0; index < entries; index++) array[index] = index; return fd_io_uring; } #endif #if SYZ_EXECUTOR || __NR_syz_io_uring_submit static long syz_io_uring_submit(volatile long a0, volatile long a1, volatile long a2) { // syzlang: syz_io_uring_submit(ring_ptr ring_ptr, sqes_ptr sqes_ptr, sqe ptr[in, io_uring_sqe]) // C: syz_io_uring_submit(char* ring_ptr, io_uring_sqe* sqes_ptr, io_uring_sqe* sqe) // It is not checked if the ring is full // Cast to original char* ring_ptr = (char*)a0; // This will be exposed to offsets in bytes char* sqes_ptr = (char*)a1; char* sqe = (char*)a2; uint32 sq_ring_mask = *(uint32*)(ring_ptr + SQ_RING_MASK_OFFSET); uint32* sq_tail_ptr = (uint32*)(ring_ptr + SQ_TAIL_OFFSET); uint32 sq_tail = *sq_tail_ptr & sq_ring_mask; // Get the ptr to the destination for the sqe char* sqe_dest = sqes_ptr + sq_tail * SIZEOF_IO_URING_SQE; // Write the sqe entry to its destination in sqes memcpy(sqe_dest, sqe, SIZEOF_IO_URING_SQE); // Write the index to the sqe array uint32 sq_tail_next = *sq_tail_ptr + 1; // Advance the tail. Tail is a free-flowing integer and relies on natural wrapping. // Ensure that the kernel will never see a tail update without the preceeding SQE // stores being done. __atomic_store_n(sq_tail_ptr, sq_tail_next, __ATOMIC_RELEASE); // Now the application is free to call io_uring_enter() to submit the sqe return 0; } #endif #endif #if SYZ_EXECUTOR || __NR_syz_usbip_server_init #include #include #include #include #include #include #include #include #include // This should be coherent with CONFIG_USBIP_VHCI_HC_PORTS. #define VHCI_HC_PORTS 8 #define VHCI_PORTS (VHCI_HC_PORTS * 2) static long syz_usbip_server_init(volatile long a0) { // port_alloc[0] corresponds to ports which can be used by usb2 and // port_alloc[1] corresponds to ports which can be used by usb3. static int port_alloc[2]; int speed = (int)a0; bool usb3 = (speed == USB_SPEED_SUPER); int socket_pair[2]; if (socketpair(AF_UNIX, SOCK_STREAM, 0, socket_pair)) { // This can happen if the test calls prlimit(RLIMIT_AS). debug("syz_usbip_server_init: socketpair failed (%d)\n", errno); return -1; } int client_fd = socket_pair[0]; int server_fd = socket_pair[1]; int available_port_num = __atomic_fetch_add(&port_alloc[usb3], 1, __ATOMIC_RELAXED); if (available_port_num > VHCI_HC_PORTS) { debug("syz_usbip_server_init : no more available port for : %d\n", available_port_num); return -1; } // Each port number corresponds to a particular vhci_hcd (USB/IP Virtual Host Controller) and it is used by either // an usb2 device or usb3 device. There are 16 ports available in each vhci_hcd. // (VHCI_PORTS = 16 in our case.) When they are occupied, the following vhci_hcd's ports are used. // First 16 ports correspond to vhci_hcd0, next 16 ports correspond to // vhci_hcd1 etc. In a vhci_hcd, first 8 ports are used by usb2 devices and last 8 are used by usb3 devices. int port_num = procid * VHCI_PORTS + usb3 * VHCI_HC_PORTS + available_port_num; // Under normal USB/IP usage, devid represents the device ID on the server. // When fuzzing with syzkaller we don't have an actual server or an actual device, so use 0 for devid. char buffer[100]; sprintf(buffer, "%d %d %s %d", port_num, client_fd, "0", speed); write_file("/sys/devices/platform/vhci_hcd.0/attach", buffer); return server_fd; } #endif #if SYZ_EXECUTOR || __NR_syz_btf_id_by_name #include #include #include #include #include #include #include #include #include // Some items in linux/btf.h are relatively new, so we copy them here for // backward compatibility. #define BTF_MAGIC 0xeB9F struct btf_header { __u16 magic; __u8 version; __u8 flags; __u32 hdr_len; __u32 type_off; __u32 type_len; __u32 str_off; __u32 str_len; }; #define BTF_INFO_KIND(info) (((info) >> 24) & 0x0f) #define BTF_INFO_VLEN(info) ((info) & 0xffff) #define BTF_KIND_INT 1 #define BTF_KIND_ARRAY 3 #define BTF_KIND_STRUCT 4 #define BTF_KIND_UNION 5 #define BTF_KIND_ENUM 6 #define BTF_KIND_FUNC_PROTO 13 #define BTF_KIND_VAR 14 #define BTF_KIND_DATASEC 15 struct btf_type { __u32 name_off; __u32 info; union { __u32 size; __u32 type; }; }; struct btf_enum { __u32 name_off; __s32 val; }; struct btf_array { __u32 type; __u32 index_type; __u32 nelems; }; struct btf_member { __u32 name_off; __u32 type; __u32 offset; }; struct btf_param { __u32 name_off; __u32 type; }; struct btf_var { __u32 linkage; }; struct btf_var_secinfo { __u32 type; __u32 offset; __u32 size; }; // Set the limit on the maximum size of btf/vmlinux to be 10 MiB. #define VMLINUX_MAX_SUPPORT_SIZE (10 * 1024 * 1024) // Read out all the content of /sys/kernel/btf/vmlinux to the fixed address // buffer and return it. Return NULL if failed. static char* read_btf_vmlinux() { static bool is_read = false; static char buf[VMLINUX_MAX_SUPPORT_SIZE]; // There could be a race condition here, but it should not be harmful. if (is_read) return buf; int fd = open("/sys/kernel/btf/vmlinux", O_RDONLY); if (fd < 0) return NULL; unsigned long bytes_read = 0; for (;;) { ssize_t ret = read(fd, buf + bytes_read, VMLINUX_MAX_SUPPORT_SIZE - bytes_read); if (ret < 0 || bytes_read + ret == VMLINUX_MAX_SUPPORT_SIZE) return NULL; if (ret == 0) break; bytes_read += ret; } is_read = true; return buf; } // Given a pointer to a C-string as the only argument a0, return the // corresponding btf ID for this name. Return -1 if there is an error when // opening the vmlinux file or the name is not found in vmlinux. static long syz_btf_id_by_name(volatile long a0) { // syzlang: syz_btf_id_by_name(name ptr[in, string]) btf_id // C: syz_btf_id_by_name(char* name) char* target = (char*)a0; char* vmlinux = read_btf_vmlinux(); if (vmlinux == NULL) return -1; struct btf_header* btf_header = (struct btf_header*)vmlinux; if (btf_header->magic != BTF_MAGIC) return -1; // These offsets are bytes relative to the end of the header. char* btf_type_sec = vmlinux + btf_header->hdr_len + btf_header->type_off; char* btf_str_sec = vmlinux + btf_header->hdr_len + btf_header->str_off; // Scan through the btf type section, and find a type description that // matches the provided name. unsigned int bytes_parsed = 0; // BTF index starts at 1. long idx = 1; while (bytes_parsed < btf_header->type_len) { struct btf_type* btf_type = (struct btf_type*)(btf_type_sec + bytes_parsed); uint32 kind = BTF_INFO_KIND(btf_type->info); uint32 vlen = BTF_INFO_VLEN(btf_type->info); char* name = btf_str_sec + btf_type->name_off; if (strcmp(name, target) == 0) return idx; // From /include/uapi/linux/btf.h, some kinds of types are // followed by extra data. size_t skip; switch (kind) { case BTF_KIND_INT: skip = sizeof(uint32); break; case BTF_KIND_ENUM: skip = sizeof(struct btf_enum) * vlen; break; case BTF_KIND_ARRAY: skip = sizeof(struct btf_array); break; case BTF_KIND_STRUCT: case BTF_KIND_UNION: skip = sizeof(struct btf_member) * vlen; break; case BTF_KIND_FUNC_PROTO: skip = sizeof(struct btf_param) * vlen; break; case BTF_KIND_VAR: skip = sizeof(struct btf_var); break; case BTF_KIND_DATASEC: skip = sizeof(struct btf_var_secinfo) * vlen; break; default: skip = 0; } bytes_parsed += sizeof(struct btf_type) + skip; idx++; } return -1; } #endif // SYZ_EXECUTOR || __NR_syz_btf_id_by_name // Same as memcpy except that it accepts offset to dest and src. #if SYZ_EXECUTOR || __NR_syz_memcpy_off static long syz_memcpy_off(volatile long a0, volatile long a1, volatile long a2, volatile long a3, volatile long a4) { // C: syz_memcpy_off(void* dest, uint32 dest_off, void* src, uint32 src_off, size_t n) // Cast to original char* dest = (char*)a0; uint32 dest_off = (uint32)a1; char* src = (char*)a2; uint32 src_off = (uint32)a3; size_t n = (size_t)a4; return (long)memcpy(dest + dest_off, src + src_off, n); } #endif #if SYZ_EXECUTOR || __NR_syz_create_resource // syz_create_resource(val intptr) intptr // Variants of this pseudo-syscall are used to create resources from arbitrary values. // For example: // syz_create_resource$foo(x int32) resource_foo // allows the fuzzer to use the same random int32 value in multiple syscalls, // and should increase probability of generation of syscalls related to foo. static long syz_create_resource(volatile long val) { return val; } #endif #if (SYZ_EXECUTOR || SYZ_REPEAT && SYZ_NET_INJECTION) && SYZ_EXECUTOR_USES_FORK_SERVER static void flush_tun() { #if SYZ_EXECUTOR if (!flag_net_injection) return; #endif char data[1000]; while (read_tun(&data[0], sizeof(data)) != -1) { } } #endif #if SYZ_EXECUTOR || __NR_syz_extract_tcp_res && SYZ_NET_INJECTION #ifndef __ANDROID__ // Can't include , since it causes // conflicts due to some structs redefinition. struct ipv6hdr { __u8 priority : 4, version : 4; __u8 flow_lbl[3]; __be16 payload_len; __u8 nexthdr; __u8 hop_limit; struct in6_addr saddr; struct in6_addr daddr; }; #endif struct tcp_resources { uint32 seq; uint32 ack; }; static long syz_extract_tcp_res(volatile long a0, volatile long a1, volatile long a2) { // syz_extract_tcp_res(res ptr[out, tcp_resources], seq_inc int32, ack_inc int32) if (tunfd < 0) return (uintptr_t)-1; // We just need this to be large enough to hold headers that we parse (ethernet/ip/tcp). // Rest of the packet (if any) will be silently truncated which is fine. char data[1000]; int rv = read_tun(&data[0], sizeof(data)); if (rv == -1) return (uintptr_t)-1; size_t length = rv; debug_dump_data(data, length); if (length < sizeof(struct ethhdr)) return (uintptr_t)-1; struct ethhdr* ethhdr = (struct ethhdr*)&data[0]; struct tcphdr* tcphdr = 0; if (ethhdr->h_proto == htons(ETH_P_IP)) { if (length < sizeof(struct ethhdr) + sizeof(struct iphdr)) return (uintptr_t)-1; struct iphdr* iphdr = (struct iphdr*)&data[sizeof(struct ethhdr)]; if (iphdr->protocol != IPPROTO_TCP) return (uintptr_t)-1; if (length < sizeof(struct ethhdr) + iphdr->ihl * 4 + sizeof(struct tcphdr)) return (uintptr_t)-1; tcphdr = (struct tcphdr*)&data[sizeof(struct ethhdr) + iphdr->ihl * 4]; } else { if (length < sizeof(struct ethhdr) + sizeof(struct ipv6hdr)) return (uintptr_t)-1; struct ipv6hdr* ipv6hdr = (struct ipv6hdr*)&data[sizeof(struct ethhdr)]; // TODO: parse and skip extension headers. if (ipv6hdr->nexthdr != IPPROTO_TCP) return (uintptr_t)-1; if (length < sizeof(struct ethhdr) + sizeof(struct ipv6hdr) + sizeof(struct tcphdr)) return (uintptr_t)-1; tcphdr = (struct tcphdr*)&data[sizeof(struct ethhdr) + sizeof(struct ipv6hdr)]; } struct tcp_resources* res = (struct tcp_resources*)a0; res->seq = htonl((ntohl(tcphdr->seq) + (uint32)a1)); res->ack = htonl((ntohl(tcphdr->ack_seq) + (uint32)a2)); debug("extracted seq: %08x\n", res->seq); debug("extracted ack: %08x\n", res->ack); return 0; } #endif #if SYZ_EXECUTOR || SYZ_CLOSE_FDS || __NR_syz_usb_connect || __NR_syz_usb_connect_ath9k #define MAX_FDS 30 #endif #if SYZ_EXECUTOR || __NR_syz_usb_connect || __NR_syz_usb_connect_ath9k || \ __NR_syz_usb_ep_write || __NR_syz_usb_ep_read || __NR_syz_usb_control_io || \ __NR_syz_usb_disconnect #include #include #include #include #include #include #include #include #include #include #include "common_usb_linux.h" #endif #if SYZ_EXECUTOR || __NR_syz_open_dev #include #include #include #include static long syz_open_dev(volatile long a0, volatile long a1, volatile long a2) { if (a0 == 0xc || a0 == 0xb) { // syz_open_dev$char(dev const[0xc], major intptr, minor intptr) fd // syz_open_dev$block(dev const[0xb], major intptr, minor intptr) fd char buf[128]; sprintf(buf, "/dev/%s/%d:%d", a0 == 0xc ? "char" : "block", (uint8)a1, (uint8)a2); return open(buf, O_RDWR, 0); } else { // syz_open_dev(dev ptr[in, string["/dev/foo#"]], id intptr, flags flags[open_flags]) fd unsigned long nb = a1; char buf[1024]; char* hash; strncpy(buf, (char*)a0, sizeof(buf) - 1); buf[sizeof(buf) - 1] = 0; while ((hash = strchr(buf, '#'))) { *hash = '0' + (char)(nb % 10); // 10 devices should be enough for everyone. nb /= 10; } return open(buf, a2 & ~O_CREAT, 0); } } #endif #if SYZ_EXECUTOR || __NR_syz_open_procfs #include #include #include #include static long syz_open_procfs(volatile long a0, volatile long a1) { // syz_open_procfs(pid pid, file ptr[in, string[procfs_file]]) fd char buf[128]; memset(buf, 0, sizeof(buf)); if (a0 == 0) { snprintf(buf, sizeof(buf), "/proc/self/%s", (char*)a1); } else if (a0 == -1) { snprintf(buf, sizeof(buf), "/proc/thread-self/%s", (char*)a1); } else { snprintf(buf, sizeof(buf), "/proc/self/task/%d/%s", (int)a0, (char*)a1); } int fd = open(buf, O_RDWR); if (fd == -1) fd = open(buf, O_RDONLY); return fd; } #endif #if SYZ_EXECUTOR || __NR_syz_open_pts #include #include #include #include static long syz_open_pts(volatile long a0, volatile long a1) { // syz_openpts(fd fd[tty], flags flags[open_flags]) fd[tty] int ptyno = 0; if (ioctl(a0, TIOCGPTN, &ptyno)) return -1; char buf[128]; sprintf(buf, "/dev/pts/%d", ptyno); return open(buf, a1, 0); } #endif #if SYZ_EXECUTOR || __NR_syz_init_net_socket #if SYZ_EXECUTOR || SYZ_SANDBOX_NONE || SYZ_SANDBOX_SETUID || SYZ_SANDBOX_NAMESPACE || SYZ_SANDBOX_ANDROID #include #include #include #include #include // syz_init_net_socket opens a socket in init net namespace. // Used for families that can only be created in init net namespace. static long syz_init_net_socket(volatile long domain, volatile long type, volatile long proto) { int netns = open("/proc/self/ns/net", O_RDONLY); if (netns == -1) return netns; if (setns(kInitNetNsFd, 0)) return -1; int sock = syscall(__NR_socket, domain, type, proto); int err = errno; if (setns(netns, 0)) { // The operation may fail if the fd is closed by // a syscall from another thread. exitf("setns(netns) failed"); } close(netns); errno = err; return sock; } #else static long syz_init_net_socket(volatile long domain, volatile long type, volatile long proto) { return syscall(__NR_socket, domain, type, proto); } #endif #endif #if SYZ_EXECUTOR || __NR_syz_socket_connect_nvme_tcp #if SYZ_EXECUTOR || SYZ_SANDBOX_NONE || SYZ_SANDBOX_SETUID || SYZ_SANDBOX_NAMESPACE #include #include #include #include #include #include #include #include static long syz_socket_connect_nvme_tcp() { struct sockaddr_in nvme_local_address; int netns = open("/proc/self/ns/net", O_RDONLY); if (netns == -1) return netns; if (setns(kInitNetNsFd, 0)) return -1; int sock = syscall(__NR_socket, AF_INET, SOCK_STREAM, 0x0); int err = errno; if (setns(netns, 0)) { // The operation may fail if the fd is closed by // a syscall from another thread. exitf("setns(netns) failed"); } close(netns); errno = err; // We only connect to an NVMe-oF/TCP server on 127.0.0.1:4420 nvme_local_address.sin_family = AF_INET; nvme_local_address.sin_port = htobe16(4420); nvme_local_address.sin_addr.s_addr = htobe32(0x7f000001); err = syscall(__NR_connect, sock, &nvme_local_address, sizeof(nvme_local_address)); if (err != 0) { close(sock); return -1; } return sock; } #else static long syz_socket_connect_nvme_tcp() { return syscall(__NR_socket, -1, 0, 0); } #endif #endif #if SYZ_EXECUTOR || SYZ_VHCI_INJECTION #include #include #include #include #include #include #include #include #define BTPROTO_HCI 1 #define ACL_LINK 1 #define SCAN_PAGE 2 typedef struct { uint8 b[6]; } __attribute__((packed)) bdaddr_t; #define HCI_COMMAND_PKT 1 #define HCI_EVENT_PKT 4 #define HCI_VENDOR_PKT 0xff struct hci_command_hdr { uint16 opcode; uint8 plen; } __attribute__((packed)); struct hci_event_hdr { uint8 evt; uint8 plen; } __attribute__((packed)); #define HCI_EV_CONN_COMPLETE 0x03 struct hci_ev_conn_complete { uint8 status; uint16 handle; bdaddr_t bdaddr; uint8 link_type; uint8 encr_mode; } __attribute__((packed)); #define HCI_EV_CONN_REQUEST 0x04 struct hci_ev_conn_request { bdaddr_t bdaddr; uint8 dev_class[3]; uint8 link_type; } __attribute__((packed)); #define HCI_EV_REMOTE_FEATURES 0x0b struct hci_ev_remote_features { uint8 status; uint16 handle; uint8 features[8]; } __attribute__((packed)); #define HCI_EV_CMD_COMPLETE 0x0e struct hci_ev_cmd_complete { uint8 ncmd; uint16 opcode; } __attribute__((packed)); #define HCI_OP_WRITE_SCAN_ENABLE 0x0c1a #define HCI_OP_READ_BUFFER_SIZE 0x1005 struct hci_rp_read_buffer_size { uint8 status; uint16 acl_mtu; uint8 sco_mtu; uint16 acl_max_pkt; uint16 sco_max_pkt; } __attribute__((packed)); #define HCI_OP_READ_BD_ADDR 0x1009 struct hci_rp_read_bd_addr { uint8 status; bdaddr_t bdaddr; } __attribute__((packed)); #define HCI_EV_LE_META 0x3e struct hci_ev_le_meta { uint8 subevent; } __attribute__((packed)); #define HCI_EV_LE_CONN_COMPLETE 0x01 struct hci_ev_le_conn_complete { uint8 status; uint16 handle; uint8 role; uint8 bdaddr_type; bdaddr_t bdaddr; uint16 interval; uint16 latency; uint16 supervision_timeout; uint8 clk_accurancy; } __attribute__((packed)); struct hci_dev_req { uint16 dev_id; uint32 dev_opt; }; struct vhci_vendor_pkt_request { uint8 type; uint8 opcode; } __attribute__((packed)); struct vhci_pkt { uint8 type; union { struct { uint8 opcode; uint16 id; } __attribute__((packed)) vendor_pkt; struct hci_command_hdr command_hdr; }; } __attribute__((packed)); #define HCIDEVUP _IOW('H', 201, int) #define HCISETSCAN _IOW('H', 221, int) static int vhci_fd = -1; static void rfkill_unblock_all() { int fd = open("/dev/rfkill", O_WRONLY); if (fd < 0) fail("open /dev/rfkill failed"); struct rfkill_event event = {0}; event.idx = 0; event.type = RFKILL_TYPE_ALL; event.op = RFKILL_OP_CHANGE_ALL; event.soft = 0; event.hard = 0; if (write(fd, &event, sizeof(event)) < 0) fail("write rfkill event failed"); close(fd); } static void hci_send_event_packet(int fd, uint8 evt, void* data, size_t data_len) { struct iovec iv[3]; struct hci_event_hdr hdr; hdr.evt = evt; hdr.plen = data_len; uint8 type = HCI_EVENT_PKT; iv[0].iov_base = &type; iv[0].iov_len = sizeof(type); iv[1].iov_base = &hdr; iv[1].iov_len = sizeof(hdr); iv[2].iov_base = data; iv[2].iov_len = data_len; if (writev(fd, iv, sizeof(iv) / sizeof(struct iovec)) < 0) fail("writev failed"); } static void hci_send_event_cmd_complete(int fd, uint16 opcode, void* data, size_t data_len) { struct iovec iv[4]; struct hci_event_hdr hdr; hdr.evt = HCI_EV_CMD_COMPLETE; hdr.plen = sizeof(struct hci_ev_cmd_complete) + data_len; struct hci_ev_cmd_complete evt_hdr; evt_hdr.ncmd = 1; evt_hdr.opcode = opcode; uint8 type = HCI_EVENT_PKT; iv[0].iov_base = &type; iv[0].iov_len = sizeof(type); iv[1].iov_base = &hdr; iv[1].iov_len = sizeof(hdr); iv[2].iov_base = &evt_hdr; iv[2].iov_len = sizeof(evt_hdr); iv[3].iov_base = data; iv[3].iov_len = data_len; if (writev(fd, iv, sizeof(iv) / sizeof(struct iovec)) < 0) fail("writev failed"); } static bool process_command_pkt(int fd, char* buf, ssize_t buf_size) { struct hci_command_hdr* hdr = (struct hci_command_hdr*)buf; if (buf_size < (ssize_t)sizeof(struct hci_command_hdr) || hdr->plen != buf_size - sizeof(struct hci_command_hdr)) failmsg("process_command_pkt: invalid size", "suze=%zx", buf_size); switch (hdr->opcode) { case HCI_OP_WRITE_SCAN_ENABLE: { uint8 status = 0; hci_send_event_cmd_complete(fd, hdr->opcode, &status, sizeof(status)); return true; } case HCI_OP_READ_BD_ADDR: { struct hci_rp_read_bd_addr rp = {0}; rp.status = 0; memset(&rp.bdaddr, 0xaa, 6); hci_send_event_cmd_complete(fd, hdr->opcode, &rp, sizeof(rp)); return false; } case HCI_OP_READ_BUFFER_SIZE: { struct hci_rp_read_buffer_size rp = {0}; rp.status = 0; rp.acl_mtu = 1021; rp.sco_mtu = 96; rp.acl_max_pkt = 4; rp.sco_max_pkt = 6; hci_send_event_cmd_complete(fd, hdr->opcode, &rp, sizeof(rp)); return false; } } char dummy[0xf9] = {0}; hci_send_event_cmd_complete(fd, hdr->opcode, dummy, sizeof(dummy)); return false; } static void* event_thread(void* arg) { while (1) { char buf[1024] = {0}; ssize_t buf_size = read(vhci_fd, buf, sizeof(buf)); if (buf_size < 0) fail("read failed"); debug_dump_data(buf, buf_size); if (buf_size > 0 && buf[0] == HCI_COMMAND_PKT) { if (process_command_pkt(vhci_fd, buf + 1, buf_size - 1)) break; } } return NULL; } // Matches hci_handles in sys/linux/dev_vhci.txt. #define HCI_HANDLE_1 200 #define HCI_HANDLE_2 201 #define HCI_PRIMARY 0 #define HCI_OP_RESET 0x0c03 static void initialize_vhci() { #if SYZ_EXECUTOR if (!flag_vhci_injection) return; #endif int hci_sock = socket(AF_BLUETOOTH, SOCK_RAW, BTPROTO_HCI); if (hci_sock < 0) fail("socket(AF_BLUETOOTH, SOCK_RAW, BTPROTO_HCI) failed"); vhci_fd = open("/dev/vhci", O_RDWR); if (vhci_fd == -1) fail("open /dev/vhci failed"); // Remap vhci onto higher fd number to hide it from fuzzer and to keep // fd numbers stable regardless of whether vhci is opened or not (also see kMaxFd). const int kVhciFd = 202; if (dup2(vhci_fd, kVhciFd) < 0) fail("dup2(vhci_fd, kVhciFd) failed"); close(vhci_fd); vhci_fd = kVhciFd; struct vhci_vendor_pkt_request vendor_pkt_req = {HCI_VENDOR_PKT, HCI_PRIMARY}; if (write(vhci_fd, &vendor_pkt_req, sizeof(vendor_pkt_req)) != sizeof(vendor_pkt_req)) fail("vendor_pkt_req write failed"); struct vhci_pkt vhci_pkt; if (read(vhci_fd, &vhci_pkt, sizeof(vhci_pkt)) != sizeof(vhci_pkt)) fail("vhci_pkt read failed"); if (vhci_pkt.type == HCI_COMMAND_PKT && vhci_pkt.command_hdr.opcode == HCI_OP_RESET) { char response[1] = {0}; hci_send_event_cmd_complete(vhci_fd, HCI_OP_RESET, response, sizeof(response)); if (read(vhci_fd, &vhci_pkt, sizeof(vhci_pkt)) != sizeof(vhci_pkt)) fail("vhci_pkt read failed"); } if (vhci_pkt.type != HCI_VENDOR_PKT) fail("wrong response packet"); int dev_id = vhci_pkt.vendor_pkt.id; debug("hci dev id: %x\n", dev_id); pthread_t th; if (pthread_create(&th, NULL, event_thread, NULL)) fail("pthread_create failed"); // Bring hci device up int ret = ioctl(hci_sock, HCIDEVUP, dev_id); if (ret) { if (errno == ERFKILL) { rfkill_unblock_all(); ret = ioctl(hci_sock, HCIDEVUP, dev_id); } if (ret && errno != EALREADY) fail("ioctl(HCIDEVUP) failed"); } // Activate page scanning mode which is required to fake a connection. struct hci_dev_req dr = {0}; dr.dev_id = dev_id; dr.dev_opt = SCAN_PAGE; if (ioctl(hci_sock, HCISETSCAN, &dr)) fail("ioctl(HCISETSCAN) failed"); // Fake a connection with bd address 10:aa:aa:aa:aa:aa. // This is a fixed address used in sys/linux/socket_bluetooth.txt. struct hci_ev_conn_request request; memset(&request, 0, sizeof(request)); memset(&request.bdaddr, 0xaa, 6); *(uint8*)&request.bdaddr.b[5] = 0x10; request.link_type = ACL_LINK; hci_send_event_packet(vhci_fd, HCI_EV_CONN_REQUEST, &request, sizeof(request)); struct hci_ev_conn_complete complete; memset(&complete, 0, sizeof(complete)); complete.status = 0; complete.handle = HCI_HANDLE_1; memset(&complete.bdaddr, 0xaa, 6); *(uint8*)&complete.bdaddr.b[5] = 0x10; complete.link_type = ACL_LINK; complete.encr_mode = 0; hci_send_event_packet(vhci_fd, HCI_EV_CONN_COMPLETE, &complete, sizeof(complete)); struct hci_ev_remote_features features; memset(&features, 0, sizeof(features)); features.status = 0; features.handle = HCI_HANDLE_1; hci_send_event_packet(vhci_fd, HCI_EV_REMOTE_FEATURES, &features, sizeof(features)); // Fake a low-energy connection with bd address 11:aa:aa:aa:aa:aa. // This is a fixed address used in sys/linux/socket_bluetooth.txt. struct { struct hci_ev_le_meta le_meta; struct hci_ev_le_conn_complete le_conn; } le_conn; memset(&le_conn, 0, sizeof(le_conn)); le_conn.le_meta.subevent = HCI_EV_LE_CONN_COMPLETE; memset(&le_conn.le_conn.bdaddr, 0xaa, 6); *(uint8*)&le_conn.le_conn.bdaddr.b[5] = 0x11; le_conn.le_conn.role = 1; le_conn.le_conn.handle = HCI_HANDLE_2; hci_send_event_packet(vhci_fd, HCI_EV_LE_META, &le_conn, sizeof(le_conn)); pthread_join(th, NULL); close(hci_sock); } #endif #if SYZ_EXECUTOR || __NR_syz_emit_vhci && SYZ_VHCI_INJECTION static long syz_emit_vhci(volatile long a0, volatile long a1) { if (vhci_fd < 0) return (uintptr_t)-1; char* data = (char*)a0; uint32 length = a1; return write(vhci_fd, data, length); } #endif #if SYZ_EXECUTOR || __NR_syz_genetlink_get_family_id #include #include static long syz_genetlink_get_family_id(volatile long name, volatile long sock_arg) { debug("syz_genetlink_get_family_id(%s, %d)\n", (char*)name, (int)sock_arg); int fd = sock_arg; if (fd < 0) { fd = socket(AF_NETLINK, SOCK_RAW, NETLINK_GENERIC); if (fd == -1) { debug("syz_genetlink_get_family_id: socket failed: %d\n", errno); return -1; } } struct nlmsg nlmsg_tmp; int ret = netlink_query_family_id(&nlmsg_tmp, fd, (char*)name, false); if ((int)sock_arg < 0) close(fd); if (ret < 0) { debug("syz_genetlink_get_family_id: netlink_query_family_id failed: %d\n", ret); return -1; } return ret; } #endif #if SYZ_EXECUTOR || __NR_syz_mount_image || __NR_syz_read_part_table #include "common_zlib.h" #include #include #include #include #include #include #include // Setup the loop device needed for mounting a filesystem image. Takes care of // creating and initializing the underlying file backing the loop device and // returns the fds to the file and device. // Returns 0 on success, -1 otherwise. static int setup_loop_device(unsigned char* data, unsigned long size, const char* loopname, int* loopfd_p) { int err = 0, loopfd = -1; int memfd = syscall(__NR_memfd_create, "syzkaller", 0); if (memfd == -1) { err = errno; goto error; } if (puff_zlib_to_file(data, size, memfd)) { err = errno; debug("setup_loop_device: could not decompress data: %d\n", errno); goto error_close_memfd; } loopfd = open(loopname, O_RDWR); if (loopfd == -1) { err = errno; debug("setup_loop_device: open failed: %d\n", errno); goto error_close_memfd; } if (ioctl(loopfd, LOOP_SET_FD, memfd)) { if (errno != EBUSY) { err = errno; goto error_close_loop; } ioctl(loopfd, LOOP_CLR_FD, 0); usleep(1000); if (ioctl(loopfd, LOOP_SET_FD, memfd)) { err = errno; goto error_close_loop; } } close(memfd); *loopfd_p = loopfd; return 0; error_close_loop: close(loopfd); error_close_memfd: close(memfd); error: errno = err; return -1; } #if SYZ_EXECUTOR || __NR_syz_mount_image static void reset_loop_device(const char* loopname) { int loopfd = open(loopname, O_RDWR); if (loopfd == -1) { debug("reset_loop_device: open failed: %d\n", errno); return; } if (ioctl(loopfd, LOOP_CLR_FD, 0)) { debug("reset_loop_device: LOOP_CLR_FD failed: %d\n", errno); } close(loopfd); } #endif #endif #if SYZ_EXECUTOR || __NR_syz_read_part_table // syz_read_part_table(size len[img], img ptr[in, compressed_image]) static long syz_read_part_table(volatile unsigned long size, volatile long image) { unsigned char* data = (unsigned char*)image; int err = 0, res = -1, loopfd = -1; char loopname[64]; snprintf(loopname, sizeof(loopname), "/dev/loop%llu", procid); if (setup_loop_device(data, size, loopname, &loopfd) == -1) return -1; struct loop_info64 info; if (ioctl(loopfd, LOOP_GET_STATUS64, &info)) { err = errno; goto error_clear_loop; } #if SYZ_EXECUTOR cover_reset(0); #endif info.lo_flags |= LO_FLAGS_PARTSCAN; if (ioctl(loopfd, LOOP_SET_STATUS64, &info)) { err = errno; goto error_clear_loop; } res = 0; // If we managed to parse some partitions, symlink them into our work dir. for (unsigned long i = 1, j = 0; i < 8; i++) { snprintf(loopname, sizeof(loopname), "/dev/loop%llup%d", procid, (int)i); struct stat statbuf; if (stat(loopname, &statbuf) == 0) { char linkname[64]; snprintf(linkname, sizeof(linkname), "./file%d", (int)j++); if (symlink(loopname, linkname)) { debug("syz_read_part_table: symlink(%s, %s) failed: %d\n", loopname, linkname, errno); } } } error_clear_loop: if (res) ioctl(loopfd, LOOP_CLR_FD, 0); close(loopfd); errno = err; return res; } #endif #if SYZ_EXECUTOR || __NR_syz_mount_image #include #include #include // syz_mount_image( // fs ptr[in, string[fs]], // dir ptr[in, filename], // flags flags[mount_flags], // opts ptr[in, fs_options], // chdir bool8, // size len[img], // img ptr[in, compressed_image] // ) fd_dir static long syz_mount_image( volatile long fsarg, volatile long dir, volatile long flags, volatile long optsarg, volatile long change_dir, volatile unsigned long size, volatile long image) { unsigned char* data = (unsigned char*)image; int res = -1, err = 0, need_loop_device = !!size; char* mount_opts = (char*)optsarg; char* target = (char*)dir; char* fs = (char*)fsarg; char* source = NULL; char loopname[64]; if (need_loop_device) { int loopfd; // Some filesystems (e.g. FUSE) do not need a backing device or // filesystem image. memset(loopname, 0, sizeof(loopname)); snprintf(loopname, sizeof(loopname), "/dev/loop%llu", procid); if (setup_loop_device(data, size, loopname, &loopfd) == -1) return -1; // If BLK_DEV_WRITE_MOUNTED is set, we won't be able to mount() // while holding the loop device fd. close(loopfd); source = loopname; } mkdir(target, 0777); char opts[256]; memset(opts, 0, sizeof(opts)); // Leave some space for the additional options we append below. if (strlen(mount_opts) > (sizeof(opts) - 32)) { debug("ERROR: syz_mount_image parameter optsarg bigger than internal opts\n"); } strncpy(opts, mount_opts, sizeof(opts) - 32); if (strcmp(fs, "iso9660") == 0) { flags |= MS_RDONLY; } else if (strncmp(fs, "ext", 3) == 0) { // For ext2/3/4 we have to have errors=continue because the image // can contain errors=panic flag and can legally crash kernel. bool has_remount_ro = false; char* remount_ro_start = strstr(opts, "errors=remount-ro"); if (remount_ro_start != NULL) { // syzkaller can sometimes break the options format, so we have to make sure this option can really be parsed. char after = *(remount_ro_start + strlen("errors=remount-ro")); char before = remount_ro_start == opts ? '\0' : *(remount_ro_start - 1); has_remount_ro = ((before == '\0' || before == ',') && (after == '\0' || after == ',')); } if (strstr(opts, "errors=panic") || !has_remount_ro) strcat(opts, ",errors=continue"); } else if (strcmp(fs, "xfs") == 0) { // For xfs we need nouuid because xfs has a global uuids table // and if two parallel executors mounts fs with the same uuid, second mount fails. strcat(opts, ",nouuid"); } else if (strncmp(fs, "gfs2", 4) == 0 && (strstr(opts, "errors=panic") || strstr(opts, "debug"))) { // Otherwise ordinary withdrawals turn into kernel panics, see #6189. strcat(opts, ",errors=withdraw"); } debug("syz_mount_image: size=%llu loop='%s' dir='%s' fs='%s' flags=%llu opts='%s'\n", (uint64)size, loopname, target, fs, (uint64)flags, opts); #if SYZ_EXECUTOR cover_reset(0); #endif res = mount(source, target, fs, flags, opts); if (res == -1) { debug("syz_mount_image > mount error: %d\n", errno); err = errno; goto error_clear_loop; } res = open(target, O_RDONLY | O_DIRECTORY); if (res == -1) { debug("syz_mount_image > open error: %d\n", errno); err = errno; goto error_clear_loop; } if (change_dir) { res = chdir(target); if (res == -1) { debug("syz_mount_image > chdir error: %d\n", errno); err = errno; } } error_clear_loop: if (need_loop_device) reset_loop_device(loopname); errno = err; return res; } #endif #if SYZ_EXECUTOR || __NR_syz_kvm_setup_cpu || __NR_syz_kvm_vgic_v3_setup || __NR_syz_kvm_setup_syzos_vm || __NR_syz_kvm_add_vcpu || __NR_syz_kvm_assert_syzos_uexit || __NR_syz_kvm_assert_reg || __NR_syz_kvm_assert_syzos_kvm_exit #if !GOARCH_arm #include #include #include #include #include #include #include #if GOARCH_amd64 #include "common_kvm_amd64.h" #elif GOARCH_386 #include "common_kvm_386.h" #elif GOARCH_arm64 #include "common_kvm_arm64.h" #elif GOARCH_ppc64 || GOARCH_ppc64le #include "common_kvm_ppc64.h" #elif GOARCH_riscv64 #include "common_kvm_riscv64.h" #elif SYZ_EXECUTOR || __NR_syz_kvm_setup_cpu static volatile long syz_kvm_setup_cpu(volatile long a0, volatile long a1, volatile long a2, volatile long a3, volatile long a4, volatile long a5, volatile long a6, volatile long a7) { return 0; } #endif #endif #endif #if (SYZ_EXECUTOR || SYZ_NET_RESET) && SYZ_EXECUTOR_USES_FORK_SERVER #include #include #include #include #include #include // checkpoint/reset_net_namespace partially resets net namespace to initial state // after each test. Currently it resets only ipv4 netfilter state. // Ideally, we just create a new net namespace for each test, // however it's too slow (1-1.5 seconds per namespace, not parallelizable). // Linux headers do not compile for C++, so we have to define the structs manualy. #define XT_TABLE_SIZE 1536 #define XT_MAX_ENTRIES 10 struct xt_counters { uint64 pcnt, bcnt; }; struct ipt_getinfo { char name[32]; unsigned int valid_hooks; unsigned int hook_entry[5]; unsigned int underflow[5]; unsigned int num_entries; unsigned int size; }; struct ipt_get_entries { char name[32]; unsigned int size; uint64 entrytable[XT_TABLE_SIZE / sizeof(uint64)]; }; struct ipt_replace { char name[32]; unsigned int valid_hooks; unsigned int num_entries; unsigned int size; unsigned int hook_entry[5]; unsigned int underflow[5]; unsigned int num_counters; struct xt_counters* counters; uint64 entrytable[XT_TABLE_SIZE / sizeof(uint64)]; }; struct ipt_table_desc { const char* name; struct ipt_getinfo info; struct ipt_replace replace; }; static struct ipt_table_desc ipv4_tables[] = { {.name = "filter"}, {.name = "nat"}, {.name = "mangle"}, {.name = "raw"}, {.name = "security"}, }; static struct ipt_table_desc ipv6_tables[] = { {.name = "filter"}, {.name = "nat"}, {.name = "mangle"}, {.name = "raw"}, {.name = "security"}, }; #define IPT_BASE_CTL 64 #define IPT_SO_SET_REPLACE (IPT_BASE_CTL) #define IPT_SO_GET_INFO (IPT_BASE_CTL) #define IPT_SO_GET_ENTRIES (IPT_BASE_CTL + 1) struct arpt_getinfo { char name[32]; unsigned int valid_hooks; unsigned int hook_entry[3]; unsigned int underflow[3]; unsigned int num_entries; unsigned int size; }; struct arpt_get_entries { char name[32]; unsigned int size; uint64 entrytable[XT_TABLE_SIZE / sizeof(uint64)]; }; struct arpt_replace { char name[32]; unsigned int valid_hooks; unsigned int num_entries; unsigned int size; unsigned int hook_entry[3]; unsigned int underflow[3]; unsigned int num_counters; struct xt_counters* counters; uint64 entrytable[XT_TABLE_SIZE / sizeof(uint64)]; }; struct arpt_table_desc { const char* name; struct arpt_getinfo info; struct arpt_replace replace; }; static struct arpt_table_desc arpt_tables[] = { {.name = "filter"}, }; #define ARPT_BASE_CTL 96 #define ARPT_SO_SET_REPLACE (ARPT_BASE_CTL) #define ARPT_SO_GET_INFO (ARPT_BASE_CTL) #define ARPT_SO_GET_ENTRIES (ARPT_BASE_CTL + 1) static void checkpoint_iptables(struct ipt_table_desc* tables, int num_tables, int family, int level) { int fd = socket(family, SOCK_STREAM, IPPROTO_TCP); if (fd == -1) { switch (errno) { case EAFNOSUPPORT: case ENOPROTOOPT: // ENOENT can be returned if smack lsm is used. Smack tried to aplly netlbl to created sockets, // but the fuzzer can manage to remove netlbl entry for SOCK_STREAM/IPPROTO_TCP using // NLBL_MGMT_C_REMOVE, which is unfortunately global (not part of net namespace). In this state // creation of such sockets will fail all the time in all processes (so in some sense the machine // is indeed broken), but ignoring the error is still probably the best option given we allow // the fuzzer to invoke NLBL_MGMT_C_REMOVE in the first place. case ENOENT: return; } failmsg("iptable checkpoint: socket(SOCK_STREAM, IPPROTO_TCP) failed", "family=%d", family); } for (int i = 0; i < num_tables; i++) { struct ipt_table_desc* table = &tables[i]; strcpy(table->info.name, table->name); strcpy(table->replace.name, table->name); socklen_t optlen = sizeof(table->info); if (getsockopt(fd, level, IPT_SO_GET_INFO, &table->info, &optlen)) { switch (errno) { case EPERM: case ENOENT: case ENOPROTOOPT: continue; } failmsg("iptable checkpoint: getsockopt(IPT_SO_GET_INFO) failed", "table=%s, family=%d", table->name, family); } debug("iptable checkpoint %s/%d: checkpoint entries=%d hooks=%x size=%d\n", table->name, family, table->info.num_entries, table->info.valid_hooks, table->info.size); if (table->info.size > sizeof(table->replace.entrytable)) failmsg("iptable checkpoint: table size is too large", "table=%s, family=%d, size=%u", table->name, family, table->info.size); if (table->info.num_entries > XT_MAX_ENTRIES) failmsg("iptable checkpoint: too many counters", "table=%s, family=%d, counters=%d", table->name, family, table->info.num_entries); struct ipt_get_entries entries; memset(&entries, 0, sizeof(entries)); strcpy(entries.name, table->name); entries.size = table->info.size; optlen = sizeof(entries) - sizeof(entries.entrytable) + table->info.size; if (getsockopt(fd, level, IPT_SO_GET_ENTRIES, &entries, &optlen)) failmsg("iptable checkpoint: getsockopt(IPT_SO_GET_ENTRIES) failed", "table=%s, family=%d", table->name, family); table->replace.valid_hooks = table->info.valid_hooks; table->replace.num_entries = table->info.num_entries; table->replace.size = table->info.size; memcpy(table->replace.hook_entry, table->info.hook_entry, sizeof(table->replace.hook_entry)); memcpy(table->replace.underflow, table->info.underflow, sizeof(table->replace.underflow)); memcpy(table->replace.entrytable, entries.entrytable, table->info.size); } close(fd); } static void reset_iptables(struct ipt_table_desc* tables, int num_tables, int family, int level) { int fd = socket(family, SOCK_STREAM, IPPROTO_TCP); if (fd == -1) { switch (errno) { case EAFNOSUPPORT: case ENOPROTOOPT: case ENOENT: return; } failmsg("iptable: socket(SOCK_STREAM, IPPROTO_TCP) failed", "family=%d", family); } for (int i = 0; i < num_tables; i++) { struct ipt_table_desc* table = &tables[i]; if (table->info.valid_hooks == 0) continue; struct ipt_getinfo info; memset(&info, 0, sizeof(info)); strcpy(info.name, table->name); socklen_t optlen = sizeof(info); if (getsockopt(fd, level, IPT_SO_GET_INFO, &info, &optlen)) failmsg("iptable: getsockopt(IPT_SO_GET_INFO) failed", "table=%s, family=%d", table->name, family); if (memcmp(&table->info, &info, sizeof(table->info)) == 0) { struct ipt_get_entries entries; memset(&entries, 0, sizeof(entries)); strcpy(entries.name, table->name); entries.size = table->info.size; optlen = sizeof(entries) - sizeof(entries.entrytable) + entries.size; if (getsockopt(fd, level, IPT_SO_GET_ENTRIES, &entries, &optlen)) failmsg("iptable: getsockopt(IPT_SO_GET_ENTRIES) failed", "table=%s, family=%d", table->name, family); if (memcmp(table->replace.entrytable, entries.entrytable, table->info.size) == 0) continue; } debug("iptable %s/%d: resetting\n", table->name, family); struct xt_counters counters[XT_MAX_ENTRIES]; table->replace.num_counters = info.num_entries; table->replace.counters = counters; optlen = sizeof(table->replace) - sizeof(table->replace.entrytable) + table->replace.size; if (setsockopt(fd, level, IPT_SO_SET_REPLACE, &table->replace, optlen)) failmsg("iptable: setsockopt(IPT_SO_SET_REPLACE) failed", "table=%s, family=%d", table->name, family); } close(fd); } static void checkpoint_arptables(void) { int fd = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP); if (fd == -1) { switch (errno) { case EAFNOSUPPORT: case ENOPROTOOPT: case ENOENT: return; } fail("arptable checkpoint: socket(AF_INET, SOCK_STREAM, IPPROTO_TCP) failed"); } for (unsigned i = 0; i < sizeof(arpt_tables) / sizeof(arpt_tables[0]); i++) { struct arpt_table_desc* table = &arpt_tables[i]; strcpy(table->info.name, table->name); strcpy(table->replace.name, table->name); socklen_t optlen = sizeof(table->info); if (getsockopt(fd, SOL_IP, ARPT_SO_GET_INFO, &table->info, &optlen)) { switch (errno) { case EPERM: case ENOENT: case ENOPROTOOPT: continue; } failmsg("arptable checkpoint: getsockopt(ARPT_SO_GET_INFO) failed", "table=%s", table->name); } debug("arptable checkpoint %s: entries=%d hooks=%x size=%d\n", table->name, table->info.num_entries, table->info.valid_hooks, table->info.size); if (table->info.size > sizeof(table->replace.entrytable)) failmsg("arptable checkpoint: table size is too large", "table=%s, size=%u", table->name, table->info.size); if (table->info.num_entries > XT_MAX_ENTRIES) failmsg("arptable checkpoint: too many counters", "table=%s, counters=%u", table->name, table->info.num_entries); struct arpt_get_entries entries; memset(&entries, 0, sizeof(entries)); strcpy(entries.name, table->name); entries.size = table->info.size; optlen = sizeof(entries) - sizeof(entries.entrytable) + table->info.size; if (getsockopt(fd, SOL_IP, ARPT_SO_GET_ENTRIES, &entries, &optlen)) failmsg("arptable checkpoint: getsockopt(ARPT_SO_GET_ENTRIES) failed", "table=%s", table->name); table->replace.valid_hooks = table->info.valid_hooks; table->replace.num_entries = table->info.num_entries; table->replace.size = table->info.size; memcpy(table->replace.hook_entry, table->info.hook_entry, sizeof(table->replace.hook_entry)); memcpy(table->replace.underflow, table->info.underflow, sizeof(table->replace.underflow)); memcpy(table->replace.entrytable, entries.entrytable, table->info.size); } close(fd); } static void reset_arptables() { int fd = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP); if (fd == -1) { switch (errno) { case EAFNOSUPPORT: case ENOPROTOOPT: case ENOENT: return; } fail("arptable: socket(AF_INET, SOCK_STREAM, IPPROTO_TCP)"); } for (unsigned i = 0; i < sizeof(arpt_tables) / sizeof(arpt_tables[0]); i++) { struct arpt_table_desc* table = &arpt_tables[i]; if (table->info.valid_hooks == 0) continue; struct arpt_getinfo info; memset(&info, 0, sizeof(info)); strcpy(info.name, table->name); socklen_t optlen = sizeof(info); if (getsockopt(fd, SOL_IP, ARPT_SO_GET_INFO, &info, &optlen)) failmsg("arptable: getsockopt(ARPT_SO_GET_INFO) failed", "table=%s", table->name); if (memcmp(&table->info, &info, sizeof(table->info)) == 0) { struct arpt_get_entries entries; memset(&entries, 0, sizeof(entries)); strcpy(entries.name, table->name); entries.size = table->info.size; optlen = sizeof(entries) - sizeof(entries.entrytable) + entries.size; if (getsockopt(fd, SOL_IP, ARPT_SO_GET_ENTRIES, &entries, &optlen)) failmsg("arptable: getsockopt(ARPT_SO_GET_ENTRIES) failed", "table=%s", table->name); if (memcmp(table->replace.entrytable, entries.entrytable, table->info.size) == 0) continue; debug("arptable %s: data changed\n", table->name); } else { debug("arptable %s: header changed\n", table->name); } debug("arptable %s: resetting\n", table->name); struct xt_counters counters[XT_MAX_ENTRIES]; table->replace.num_counters = info.num_entries; table->replace.counters = counters; optlen = sizeof(table->replace) - sizeof(table->replace.entrytable) + table->replace.size; if (setsockopt(fd, SOL_IP, ARPT_SO_SET_REPLACE, &table->replace, optlen)) failmsg("arptable: setsockopt(ARPT_SO_SET_REPLACE) failed", "table=%s", table->name); } close(fd); } // ebtables.h is broken too: // ebtables.h: In function ‘ebt_entry_target* ebt_get_target(ebt_entry*)’: // ebtables.h:197:19: error: invalid conversion from ‘void*’ to ‘ebt_entry_target*’ #define NF_BR_NUMHOOKS 6 #define EBT_TABLE_MAXNAMELEN 32 #define EBT_CHAIN_MAXNAMELEN 32 #define EBT_BASE_CTL 128 #define EBT_SO_SET_ENTRIES (EBT_BASE_CTL) #define EBT_SO_GET_INFO (EBT_BASE_CTL) #define EBT_SO_GET_ENTRIES (EBT_SO_GET_INFO + 1) #define EBT_SO_GET_INIT_INFO (EBT_SO_GET_ENTRIES + 1) #define EBT_SO_GET_INIT_ENTRIES (EBT_SO_GET_INIT_INFO + 1) struct ebt_replace { char name[EBT_TABLE_MAXNAMELEN]; unsigned int valid_hooks; unsigned int nentries; unsigned int entries_size; struct ebt_entries* hook_entry[NF_BR_NUMHOOKS]; unsigned int num_counters; struct ebt_counter* counters; char* entries; }; struct ebt_entries { unsigned int distinguisher; char name[EBT_CHAIN_MAXNAMELEN]; unsigned int counter_offset; int policy; unsigned int nentries; char data[0] __attribute__((aligned(__alignof__(struct ebt_replace)))); }; struct ebt_table_desc { const char* name; struct ebt_replace replace; char entrytable[XT_TABLE_SIZE]; }; static struct ebt_table_desc ebt_tables[] = { {.name = "filter"}, {.name = "nat"}, {.name = "broute"}, }; static void checkpoint_ebtables(void) { int fd = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP); if (fd == -1) { switch (errno) { case EAFNOSUPPORT: case ENOPROTOOPT: case ENOENT: return; } fail("ebtable checkpoint: socket(AF_INET, SOCK_STREAM, IPPROTO_TCP)"); } for (size_t i = 0; i < sizeof(ebt_tables) / sizeof(ebt_tables[0]); i++) { struct ebt_table_desc* table = &ebt_tables[i]; strcpy(table->replace.name, table->name); socklen_t optlen = sizeof(table->replace); if (getsockopt(fd, SOL_IP, EBT_SO_GET_INIT_INFO, &table->replace, &optlen)) { switch (errno) { case EPERM: case ENOENT: case ENOPROTOOPT: continue; } failmsg("ebtable checkpoint: getsockopt(EBT_SO_GET_INIT_INFO) failed", "table=%s", table->name); } debug("ebtable checkpoint %s: entries=%d hooks=%x size=%d\n", table->name, table->replace.nentries, table->replace.valid_hooks, table->replace.entries_size); if (table->replace.entries_size > sizeof(table->entrytable)) failmsg("ebtable checkpoint: table size is too large", "table=%s, size=%u", table->name, table->replace.entries_size); table->replace.num_counters = 0; table->replace.entries = table->entrytable; optlen = sizeof(table->replace) + table->replace.entries_size; if (getsockopt(fd, SOL_IP, EBT_SO_GET_INIT_ENTRIES, &table->replace, &optlen)) failmsg("ebtable checkpoint: getsockopt(EBT_SO_GET_INIT_ENTRIES) failed", "table=%s", table->name); } close(fd); } static void reset_ebtables() { int fd = socket(AF_INET, SOCK_STREAM, IPPROTO_TCP); if (fd == -1) { switch (errno) { case EAFNOSUPPORT: case ENOPROTOOPT: case ENOENT: return; } fail("ebtable: socket(AF_INET, SOCK_STREAM, IPPROTO_TCP)"); } for (unsigned i = 0; i < sizeof(ebt_tables) / sizeof(ebt_tables[0]); i++) { struct ebt_table_desc* table = &ebt_tables[i]; if (table->replace.valid_hooks == 0) continue; struct ebt_replace replace; memset(&replace, 0, sizeof(replace)); strcpy(replace.name, table->name); socklen_t optlen = sizeof(replace); if (getsockopt(fd, SOL_IP, EBT_SO_GET_INFO, &replace, &optlen)) failmsg("ebtable: getsockopt(EBT_SO_GET_INFO)", "table=%s", table->name); replace.num_counters = 0; table->replace.entries = 0; for (unsigned h = 0; h < NF_BR_NUMHOOKS; h++) table->replace.hook_entry[h] = 0; if (memcmp(&table->replace, &replace, sizeof(table->replace)) == 0) { char entrytable[XT_TABLE_SIZE]; memset(&entrytable, 0, sizeof(entrytable)); replace.entries = entrytable; optlen = sizeof(replace) + replace.entries_size; if (getsockopt(fd, SOL_IP, EBT_SO_GET_ENTRIES, &replace, &optlen)) failmsg("ebtable: getsockopt(EBT_SO_GET_ENTRIES) failed", "table=%s", table->name); if (memcmp(table->entrytable, entrytable, replace.entries_size) == 0) continue; } debug("ebtable %s: resetting\n", table->name); // Kernel does not seem to return actual entry points (wat?). for (unsigned j = 0, h = 0; h < NF_BR_NUMHOOKS; h++) { if (table->replace.valid_hooks & (1 << h)) { table->replace.hook_entry[h] = (struct ebt_entries*)table->entrytable + j; j++; } } table->replace.entries = table->entrytable; optlen = sizeof(table->replace) + table->replace.entries_size; if (setsockopt(fd, SOL_IP, EBT_SO_SET_ENTRIES, &table->replace, optlen)) failmsg("ebtable: setsockopt(EBT_SO_SET_ENTRIES) failed", "table=%s", table->name); } close(fd); } static void checkpoint_net_namespace(void) { #if SYZ_EXECUTOR if (!flag_net_reset || flag_sandbox_setuid) return; #endif checkpoint_ebtables(); checkpoint_arptables(); checkpoint_iptables(ipv4_tables, sizeof(ipv4_tables) / sizeof(ipv4_tables[0]), AF_INET, SOL_IP); checkpoint_iptables(ipv6_tables, sizeof(ipv6_tables) / sizeof(ipv6_tables[0]), AF_INET6, SOL_IPV6); } static void reset_net_namespace(void) { #if SYZ_EXECUTOR if (!flag_net_reset || flag_sandbox_setuid) return; #endif reset_ebtables(); reset_arptables(); reset_iptables(ipv4_tables, sizeof(ipv4_tables) / sizeof(ipv4_tables[0]), AF_INET, SOL_IP); reset_iptables(ipv6_tables, sizeof(ipv6_tables) / sizeof(ipv6_tables[0]), AF_INET6, SOL_IPV6); } #endif #if SYZ_EXECUTOR || (SYZ_CGROUPS && (SYZ_SANDBOX_NONE || SYZ_SANDBOX_SETUID || SYZ_SANDBOX_NAMESPACE || SYZ_SANDBOX_ANDROID)) #include #include #include #include #include static void mount_cgroups(const char* dir, const char** controllers, int count) { if (mkdir(dir, 0777)) { debug("mkdir(%s) failed: %d\n", dir, errno); return; } // First, probe one-by-one to understand what controllers are present. char enabled[128] = {0}; int i = 0; for (; i < count; i++) { if (mount("none", dir, "cgroup", 0, controllers[i])) { debug("mount(%s, %s) failed: %d\n", dir, controllers[i], errno); continue; } umount(dir); strcat(enabled, ","); strcat(enabled, controllers[i]); } if (enabled[0] == 0) { if (rmdir(dir) && errno != EBUSY) failmsg("rmdir failed", "dir=%s", dir); return; } // Now mount all at once. if (mount("none", dir, "cgroup", 0, enabled + 1)) { // In systemd/stretch images this is failing with EBUSY // (systemd starts messing with these mounts?), // so we don't fail, but just log the error. debug("mount(%s, %s) failed: %d\n", dir, enabled + 1, errno); if (rmdir(dir) && errno != EBUSY) failmsg("rmdir failed", "dir=%s enabled=%s", dir, enabled); } if (chmod(dir, 0777)) { debug("chmod(%s) failed: %d\n", dir, errno); } } static void mount_cgroups2(const char** controllers, int count) { if (mkdir("/syzcgroup/unified", 0777)) { debug("mkdir(/syzcgroup/unified) failed: %d\n", errno); return; } if (mount("none", "/syzcgroup/unified", "cgroup2", 0, NULL)) { debug("mount(cgroup2) failed: %d\n", errno); // For all cases when we don't end up mounting cgroup/cgroup2 // in /syzcgroup/{unified,net,cpu}, we need to remove the dir. // Otherwise these will end up as normal dirs and the fuzzer may // create huge files there. These files won't be cleaned up // after tests and may easily consume all disk space. // EBUSY usually means that cgroup is already mounted there // by a previous run of e.g. syz-execprog. if (rmdir("/syzcgroup/unified") && errno != EBUSY) fail("rmdir(/syzcgroup/unified) failed"); return; } if (chmod("/syzcgroup/unified", 0777)) { debug("chmod(/syzcgroup/unified) failed: %d\n", errno); } int control = open("/syzcgroup/unified/cgroup.subtree_control", O_WRONLY); if (control == -1) return; int i; for (i = 0; i < count; i++) if (write(control, controllers[i], strlen(controllers[i])) < 0) { debug("write(cgroup.subtree_control, %s) failed: %d\n", controllers[i], errno); } close(control); } static void setup_cgroups() { // We want to cover both cgroup and cgroup2. // Each resource controller can be bound to only one of them, // so to cover both we divide all controllers into 3 arbitrary groups. // One group is then bound to cgroup2/unified, and 2 other groups // are bound to 2 cgroup hierarchies. // Note: we need to enable controllers one-by-one for both cgroup and cgroup2. // If we enable all at the same time and one of them fails (b/c of older kernel // or not enabled configs), then all will fail. const char* unified_controllers[] = {"+cpu", "+io", "+pids"}; const char* net_controllers[] = {"net", "net_prio", "devices", "blkio", "freezer"}; const char* cpu_controllers[] = {"cpuset", "cpuacct", "hugetlb", "rlimit", "memory"}; if (mkdir("/syzcgroup", 0777)) { // Can happen due to e.g. read-only file system (EROFS). debug("mkdir(/syzcgroup) failed: %d\n", errno); return; } mount_cgroups2(unified_controllers, sizeof(unified_controllers) / sizeof(unified_controllers[0])); mount_cgroups("/syzcgroup/net", net_controllers, sizeof(net_controllers) / sizeof(net_controllers[0])); mount_cgroups("/syzcgroup/cpu", cpu_controllers, sizeof(cpu_controllers) / sizeof(cpu_controllers[0])); write_file("/syzcgroup/cpu/cgroup.clone_children", "1"); write_file("/syzcgroup/cpu/cpuset.memory_pressure_enabled", "1"); } #if (SYZ_EXECUTOR || SYZ_REPEAT) && SYZ_EXECUTOR_USES_FORK_SERVER static void setup_cgroups_loop() { #if SYZ_EXECUTOR if (!flag_cgroups) return; #endif int pid = getpid(); char file[128]; char cgroupdir[64]; snprintf(cgroupdir, sizeof(cgroupdir), "/syzcgroup/unified/syz%llu", procid); if (mkdir(cgroupdir, 0777)) { debug("mkdir(%s) failed: %d\n", cgroupdir, errno); } // Restrict number of pids per test process to prevent fork bombs. // We have up to 16 threads + main process + loop. // 32 pids should be enough for everyone. snprintf(file, sizeof(file), "%s/pids.max", cgroupdir); write_file(file, "32"); // Setup some v1 groups to make things more interesting. snprintf(file, sizeof(file), "%s/cgroup.procs", cgroupdir); write_file(file, "%d", pid); snprintf(cgroupdir, sizeof(cgroupdir), "/syzcgroup/cpu/syz%llu", procid); if (mkdir(cgroupdir, 0777)) { debug("mkdir(%s) failed: %d\n", cgroupdir, errno); } snprintf(file, sizeof(file), "%s/cgroup.procs", cgroupdir); write_file(file, "%d", pid); // Restrict memory consumption. // We have some syscalls that inherently consume lots of memory, // e.g. mounting some filesystem images requires at least 128MB // image in memory. We restrict RLIMIT_AS to 200MB. Here we gradually // increase memory limits to make things more interesting. // Also this takes into account KASAN quarantine size. // If the limit is lower than KASAN quarantine size, then it can happen // so that we kill the process, but all of its memory is in quarantine // and is still accounted against memcg. As the result memcg won't // allow to allocate any memory in the parent and in the new test process. // The current limit of 300MB supports up to 9.6GB RAM (quarantine is 1/32). snprintf(file, sizeof(file), "%s/memory.soft_limit_in_bytes", cgroupdir); write_file(file, "%d", 299 << 20); snprintf(file, sizeof(file), "%s/memory.limit_in_bytes", cgroupdir); write_file(file, "%d", 300 << 20); snprintf(cgroupdir, sizeof(cgroupdir), "/syzcgroup/net/syz%llu", procid); if (mkdir(cgroupdir, 0777)) { debug("mkdir(%s) failed: %d\n", cgroupdir, errno); } snprintf(file, sizeof(file), "%s/cgroup.procs", cgroupdir); write_file(file, "%d", pid); } static void setup_cgroups_test() { #if SYZ_EXECUTOR if (!flag_cgroups) return; #endif char cgroupdir[64]; snprintf(cgroupdir, sizeof(cgroupdir), "/syzcgroup/unified/syz%llu", procid); if (symlink(cgroupdir, "./cgroup")) { debug("symlink(%s, ./cgroup) failed: %d\n", cgroupdir, errno); } snprintf(cgroupdir, sizeof(cgroupdir), "/syzcgroup/cpu/syz%llu", procid); if (symlink(cgroupdir, "./cgroup.cpu")) { debug("symlink(%s, ./cgroup.cpu) failed: %d\n", cgroupdir, errno); } snprintf(cgroupdir, sizeof(cgroupdir), "/syzcgroup/net/syz%llu", procid); if (symlink(cgroupdir, "./cgroup.net")) { debug("symlink(%s, ./cgroup.net) failed: %d\n", cgroupdir, errno); } } #endif #if SYZ_EXECUTOR || SYZ_SANDBOX_NONE || SYZ_SANDBOX_NAMESPACE static void initialize_cgroups() { #if SYZ_EXECUTOR if (!flag_cgroups) return; #endif if (mkdir("./syz-tmp/newroot/syzcgroup", 0700)) fail("mkdir failed"); if (mkdir("./syz-tmp/newroot/syzcgroup/unified", 0700)) fail("mkdir failed"); if (mkdir("./syz-tmp/newroot/syzcgroup/cpu", 0700)) fail("mkdir failed"); if (mkdir("./syz-tmp/newroot/syzcgroup/net", 0700)) fail("mkdir failed"); unsigned bind_mount_flags = MS_BIND | MS_REC | MS_PRIVATE; if (mount("/syzcgroup/unified", "./syz-tmp/newroot/syzcgroup/unified", NULL, bind_mount_flags, NULL)) { debug("mount(cgroup2, MS_BIND) failed: %d\n", errno); } if (mount("/syzcgroup/cpu", "./syz-tmp/newroot/syzcgroup/cpu", NULL, bind_mount_flags, NULL)) { debug("mount(cgroup/cpu, MS_BIND) failed: %d\n", errno); } if (mount("/syzcgroup/net", "./syz-tmp/newroot/syzcgroup/net", NULL, bind_mount_flags, NULL)) { debug("mount(cgroup/net, MS_BIND) failed: %d\n", errno); } } #endif #endif #if SYZ_EXECUTOR || SYZ_SANDBOX_NONE || SYZ_SANDBOX_NAMESPACE static void setup_gadgetfs(); static void setup_binderfs(); static void setup_fusectl(); // Mount tmpfs and chroot into it in sandbox=none and sandbox=namespace. // This is to prevent persistent changes to the root file system (e.g. setting attributes) that may // hinder fuzzing. // See https://github.com/google/syzkaller/issues/4939 for more details. static void sandbox_common_mount_tmpfs(void) { // Android systems set fs.mount-max to a very low value, causing ENOSPC when doing the mounts below // (see https://github.com/google/syzkaller/issues/4972). 100K mounts should be enough for everyone. write_file("/proc/sys/fs/mount-max", "100000"); if (mkdir("./syz-tmp", 0777)) fail("mkdir(syz-tmp) failed"); if (mount("", "./syz-tmp", "tmpfs", 0, NULL)) fail("mount(tmpfs) failed"); if (mkdir("./syz-tmp/newroot", 0777)) fail("mkdir failed"); if (mkdir("./syz-tmp/newroot/dev", 0700)) fail("mkdir failed"); unsigned bind_mount_flags = MS_BIND | MS_REC | MS_PRIVATE; if (mount("/dev", "./syz-tmp/newroot/dev", NULL, bind_mount_flags, NULL)) fail("mount(dev) failed"); if (mkdir("./syz-tmp/newroot/proc", 0700)) fail("mkdir failed"); if (mount("syz-proc", "./syz-tmp/newroot/proc", "proc", 0, NULL)) fail("mount(proc) failed"); if (mkdir("./syz-tmp/newroot/selinux", 0700)) fail("mkdir failed"); // selinux mount used to be at /selinux, but then moved to /sys/fs/selinux. const char* selinux_path = "./syz-tmp/newroot/selinux"; if (mount("/selinux", selinux_path, NULL, bind_mount_flags, NULL)) { if (errno != ENOENT) fail("mount(/selinux) failed"); if (mount("/sys/fs/selinux", selinux_path, NULL, bind_mount_flags, NULL) && errno != ENOENT) fail("mount(/sys/fs/selinux) failed"); } if (mkdir("./syz-tmp/newroot/sys", 0700)) fail("mkdir(/sys) failed"); if (mount("/sys", "./syz-tmp/newroot/sys", 0, bind_mount_flags, NULL)) fail("mount(sysfs) failed"); if (mount("/sys/kernel/debug", "./syz-tmp/newroot/sys/kernel/debug", NULL, bind_mount_flags, NULL) && errno != ENOENT) fail("mount(debug) failed"); if (mount("/sys/fs/smackfs", "./syz-tmp/newroot/sys/fs/smackfs", NULL, bind_mount_flags, NULL) && errno != ENOENT) fail("mount(smackfs) failed"); if (mount("/proc/sys/fs/binfmt_misc", "./syz-tmp/newroot/proc/sys/fs/binfmt_misc", NULL, bind_mount_flags, NULL) && errno != ENOENT) fail("mount(binfmt_misc) failed"); // If user wants to supply custom inputs, those can be placed to /syz-inputs // That folder will be mounted to fuzzer sandbox // https://groups.google.com/g/syzkaller/c/U-DISFjKLzg if (mkdir("./syz-tmp/newroot/syz-inputs", 0700)) fail("mkdir(/syz-inputs) failed"); if (mount("/syz-inputs", "./syz-tmp/newroot/syz-inputs", NULL, bind_mount_flags | MS_RDONLY, NULL) && errno != ENOENT) fail("mount(syz-inputs) failed"); #if SYZ_EXECUTOR || SYZ_CGROUPS initialize_cgroups(); #endif if (mkdir("./syz-tmp/pivot", 0777)) fail("mkdir failed"); if (syscall(SYS_pivot_root, "./syz-tmp", "./syz-tmp/pivot")) { debug("pivot_root failed\n"); if (chdir("./syz-tmp")) fail("chdir failed"); } else { debug("pivot_root OK\n"); if (chdir("/")) fail("chdir failed"); if (umount2("./pivot", MNT_DETACH)) fail("umount failed"); } if (chroot("./newroot")) fail("chroot failed"); if (chdir("/")) fail("chdir failed"); setup_gadgetfs(); setup_binderfs(); setup_fusectl(); } #endif #if SYZ_EXECUTOR || SYZ_SANDBOX_NONE || SYZ_SANDBOX_NAMESPACE #include #include static void setup_gadgetfs() { if (mkdir("/dev/gadgetfs", 0777)) { debug("mkdir(/dev/gadgetfs) failed: %d\n", errno); } if (mount("gadgetfs", "/dev/gadgetfs", "gadgetfs", 0, NULL)) { debug("mount of gadgetfs at /dev/gadgetfs failed: %d\n", errno); } } #endif #if SYZ_EXECUTOR || SYZ_SANDBOX_NONE || SYZ_SANDBOX_SETUID || SYZ_SANDBOX_NAMESPACE || SYZ_SANDBOX_ANDROID #include #include #include #include static void setup_fusectl() { if (mount(0, "/sys/fs/fuse/connections", "fusectl", 0, 0)) { debug("mount(fusectl) failed: %d\n", errno); } } static void setup_binderfs() { // NOTE: this function must be called after chroot. // Bind an instance of binderfs specific just to this executor - it will // only be visible in its mount namespace and will help isolate binder // devices during fuzzing. // These commands will just silently fail if binderfs is not supported. // Ideally it should have been added as a separate feature (with lots of // minor changes throughout the code base), but it seems to be an overkill // for just 2 simple lines of code. if (mkdir("/dev/binderfs", 0777)) { debug("mkdir(/dev/binderfs) failed: %d\n", errno); } if (mount("binder", "/dev/binderfs", "binder", 0, NULL)) { debug("mount of binder at /dev/binderfs failed: %d\n", errno); } #if !SYZ_EXECUTOR && !SYZ_USE_TMP_DIR // Do a local symlink right away. if (symlink("/dev/binderfs", "./binderfs")) { debug("symlink(/dev/binderfs, ./binderfs) failed: %d\n", errno); } #endif } #include #include #include #include #include static void loop(); static void sandbox_common() { prctl(PR_SET_PDEATHSIG, SIGKILL, 0, 0, 0); if (getppid() == 1) exitf("the sandbox parent process was killed"); #if SYZ_EXECUTOR || __NR_syz_init_net_socket || SYZ_DEVLINK_PCI || __NR_syz_socket_connect_nvme_tcp int netns = open("/proc/self/ns/net", O_RDONLY); if (netns == -1) fail("open(/proc/self/ns/net) failed"); if (dup2(netns, kInitNetNsFd) < 0) fail("dup2(netns, kInitNetNsFd) failed"); close(netns); #endif struct rlimit rlim; #if SYZ_EXECUTOR rlim.rlim_cur = rlim.rlim_max = (200 << 20) + (kMaxThreads * kCoverSize + kExtraCoverSize) * sizeof(void*); #else rlim.rlim_cur = rlim.rlim_max = (200 << 20); #endif setrlimit(RLIMIT_AS, &rlim); rlim.rlim_cur = rlim.rlim_max = 32 << 20; setrlimit(RLIMIT_MEMLOCK, &rlim); rlim.rlim_cur = rlim.rlim_max = 136 << 20; setrlimit(RLIMIT_FSIZE, &rlim); rlim.rlim_cur = rlim.rlim_max = 1 << 20; setrlimit(RLIMIT_STACK, &rlim); // Note: core size is also restricted by RLIMIT_FSIZE. rlim.rlim_cur = rlim.rlim_max = 128 << 20; setrlimit(RLIMIT_CORE, &rlim); rlim.rlim_cur = rlim.rlim_max = 256; // see kMaxFd setrlimit(RLIMIT_NOFILE, &rlim); // CLONE_NEWNS/NEWCGROUP cause EINVAL on some systems, // so we do them separately of clone in do_sandbox_namespace. if (unshare(CLONE_NEWNS)) { debug("unshare(CLONE_NEWNS): %d\n", errno); } if (mount(NULL, "/", NULL, MS_REC | MS_PRIVATE, NULL)) { debug("mount(\"/\", MS_REC | MS_PRIVATE): %d\n", errno); } if (unshare(CLONE_NEWIPC)) { debug("unshare(CLONE_NEWIPC): %d\n", errno); } if (unshare(0x02000000)) { debug("unshare(CLONE_NEWCGROUP): %d\n", errno); } if (unshare(CLONE_NEWUTS)) { debug("unshare(CLONE_NEWUTS): %d\n", errno); } if (unshare(CLONE_SYSVSEM)) { debug("unshare(CLONE_SYSVSEM): %d\n", errno); } // These sysctl's restrict ipc resource usage (by default it's possible // to eat all system memory by creating e.g. lots of large sem sets). // These sysctl's are per-namespace, so we need to set them inside // of the test ipc namespace (after CLONE_NEWIPC). typedef struct { const char* name; const char* value; } sysctl_t; static const sysctl_t sysctls[] = { {"/proc/sys/kernel/shmmax", "16777216"}, {"/proc/sys/kernel/shmall", "536870912"}, {"/proc/sys/kernel/shmmni", "1024"}, {"/proc/sys/kernel/msgmax", "8192"}, {"/proc/sys/kernel/msgmni", "1024"}, {"/proc/sys/kernel/msgmnb", "1024"}, {"/proc/sys/kernel/sem", "1024 1048576 500 1024"}, }; unsigned i; for (i = 0; i < sizeof(sysctls) / sizeof(sysctls[0]); i++) write_file(sysctls[i].name, sysctls[i].value); } #endif #if SYZ_EXECUTOR || SYZ_SANDBOX_NONE || SYZ_SANDBOX_SETUID || SYZ_SANDBOX_NAMESPACE static int wait_for_loop(int pid) { if (pid < 0) fail("sandbox fork failed"); debug("spawned loop pid %d\n", pid); int status = 0; while (waitpid(-1, &status, __WALL) != pid) { } return WEXITSTATUS(status); } #endif #if SYZ_EXECUTOR || SYZ_SANDBOX_NONE || SYZ_SANDBOX_NAMESPACE || SYZ_SANDBOX_ANDROID #include static void drop_caps(void) { struct __user_cap_header_struct cap_hdr = {}; struct __user_cap_data_struct cap_data[2] = {}; cap_hdr.version = _LINUX_CAPABILITY_VERSION_3; cap_hdr.pid = getpid(); if (syscall(SYS_capget, &cap_hdr, &cap_data)) fail("capget failed"); // Drop CAP_SYS_PTRACE so that test processes can't attach to parent processes. // Previously it lead to hangs because the loop process stopped due to SIGSTOP. // Note that a process can always ptrace its direct children, which is enough for testing purposes. // // A process with CAP_SYS_NICE can bring kernel down by asking for too high SCHED_DEADLINE priority, // as the result rcu and other system services that use kernel threads will stop functioning. // Some parameters for SCHED_DEADLINE should be OK, but we don't have means to enforce // values of indirect syscall arguments. Peter Zijlstra proposed sysctl_deadline_period_{min,max} // which could be used to enfore safe limits without droppping CAP_SYS_NICE, but we don't have it yet. // See the following bug for details: // https://groups.google.com/forum/#!topic/syzkaller-bugs/G6Wl_PKPIWI const int drop = (1 << CAP_SYS_PTRACE) | (1 << CAP_SYS_NICE); cap_data[0].effective &= ~drop; cap_data[0].permitted &= ~drop; cap_data[0].inheritable &= ~drop; if (syscall(SYS_capset, &cap_hdr, &cap_data)) fail("capset failed"); } #endif #if SYZ_EXECUTOR || SYZ_SANDBOX_NONE #include #include static int do_sandbox_none(void) { // CLONE_NEWPID takes effect for the first child of the current process, // so we do it before fork to make the loop "init" process of the namespace. // We ought to do fail here, but sandbox=none is used in pkg/ipc tests // and they are usually run under non-root. // Also since debug is stripped by pkg/csource, we need to do {} // even though we generally don't do {} around single statements. if (unshare(CLONE_NEWPID)) { debug("unshare(CLONE_NEWPID): %d\n", errno); } int pid = fork(); if (pid != 0) return wait_for_loop(pid); #if SYZ_EXECUTOR || SYZ_VHCI_INJECTION initialize_vhci(); #endif sandbox_common(); drop_caps(); #if SYZ_EXECUTOR || SYZ_NET_DEVICES initialize_netdevices_init(); #endif if (unshare(CLONE_NEWNET)) { debug("unshare(CLONE_NEWNET): %d\n", errno); } // Enable access to IPPROTO_ICMP sockets, must be done after CLONE_NEWNET. write_file("/proc/sys/net/ipv4/ping_group_range", "0 65535"); #if SYZ_EXECUTOR || SYZ_DEVLINK_PCI initialize_devlink_pci(); #endif #if SYZ_EXECUTOR || SYZ_NET_INJECTION initialize_tun(); #endif #if SYZ_EXECUTOR || SYZ_NET_DEVICES initialize_netdevices(); #endif #if SYZ_EXECUTOR || SYZ_WIFI initialize_wifi_devices(); #endif sandbox_common_mount_tmpfs(); loop(); doexit(1); } #endif #if SYZ_EXECUTOR || SYZ_SANDBOX_SETUID #include #include #include #define SYZ_HAVE_SANDBOX_SETUID 1 static int do_sandbox_setuid(void) { if (unshare(CLONE_NEWPID)) { debug("unshare(CLONE_NEWPID): %d\n", errno); } int pid = fork(); if (pid != 0) return wait_for_loop(pid); #if SYZ_EXECUTOR || SYZ_VHCI_INJECTION initialize_vhci(); #endif sandbox_common(); #if SYZ_EXECUTOR || SYZ_NET_DEVICES initialize_netdevices_init(); #endif if (unshare(CLONE_NEWNET)) { debug("unshare(CLONE_NEWNET): %d\n", errno); } #if SYZ_EXECUTOR || SYZ_DEVLINK_PCI initialize_devlink_pci(); #endif #if SYZ_EXECUTOR || SYZ_NET_INJECTION initialize_tun(); #endif #if SYZ_EXECUTOR || SYZ_NET_DEVICES initialize_netdevices(); #endif #if SYZ_EXECUTOR || SYZ_WIFI initialize_wifi_devices(); #endif setup_binderfs(); setup_fusectl(); const int nobody = 65534; if (setgroups(0, NULL)) fail("failed to setgroups"); if (syscall(SYS_setresgid, nobody, nobody, nobody)) fail("failed to setresgid"); if (syscall(SYS_setresuid, nobody, nobody, nobody)) fail("failed to setresuid"); // setresuid and setresgid clear the parent-death signal. prctl(PR_SET_PDEATHSIG, SIGKILL, 0, 0, 0); // This is required to open /proc/self/ files. // Otherwise they are owned by root and we can't open them after setuid. // See task_dump_owner function in kernel. prctl(PR_SET_DUMPABLE, 1, 0, 0, 0); loop(); doexit(1); } #endif #if SYZ_EXECUTOR || SYZ_SANDBOX_NAMESPACE #include #include #include static int real_uid; static int real_gid; __attribute__((aligned(64 << 10))) static char sandbox_stack[1 << 20]; static int namespace_sandbox_proc(void* arg) { sandbox_common(); // /proc/self/setgroups is not present on some systems, ignore error. write_file("/proc/self/setgroups", "deny"); if (!write_file("/proc/self/uid_map", "0 %d 1\n", real_uid)) fail("write of /proc/self/uid_map failed"); if (!write_file("/proc/self/gid_map", "0 %d 1\n", real_gid)) fail("write of /proc/self/gid_map failed"); #if SYZ_EXECUTOR || SYZ_NET_DEVICES initialize_netdevices_init(); #endif // CLONE_NEWNET must always happen before tun setup, // because we want the tun device in the test namespace. if (unshare(CLONE_NEWNET)) fail("unshare(CLONE_NEWNET)"); // Enable access to IPPROTO_ICMP sockets, must be done after CLONE_NEWNET. write_file("/proc/sys/net/ipv4/ping_group_range", "0 65535"); #if SYZ_EXECUTOR || SYZ_DEVLINK_PCI initialize_devlink_pci(); #endif #if SYZ_EXECUTOR || SYZ_NET_INJECTION // We setup tun here as it needs to be in the test net namespace, // which in turn needs to be in the test user namespace. // However, IFF_NAPI_FRAGS will fail as we are not root already. // TODO: we should create tun in the init net namespace and use setns // to move it to the target namespace. initialize_tun(); #endif #if SYZ_EXECUTOR || SYZ_NET_DEVICES initialize_netdevices(); #endif #if SYZ_EXECUTOR || SYZ_WIFI initialize_wifi_devices(); #endif sandbox_common_mount_tmpfs(); drop_caps(); loop(); doexit(1); } #define SYZ_HAVE_SANDBOX_NAMESPACE 1 static int do_sandbox_namespace(void) { #if SYZ_EXECUTOR || SYZ_VHCI_INJECTION // HCIDEVUP requires CAP_ADMIN, so this needs to happen early. initialize_vhci(); #endif real_uid = getuid(); real_gid = getgid(); mprotect(sandbox_stack, 4096, PROT_NONE); // to catch stack underflows int pid = clone(namespace_sandbox_proc, &sandbox_stack[sizeof(sandbox_stack) - 64], CLONE_NEWUSER | CLONE_NEWPID, 0); return wait_for_loop(pid); } #endif #if SYZ_EXECUTOR || SYZ_SANDBOX_ANDROID // seccomp only supported for Arm, Arm64, X86, and X86_64 archs #if GOARCH_arm || GOARCH_arm64 || GOARCH_386 || GOARCH_amd64 #include #include #include #include #include #include #include #include #include #include "android/android_seccomp.h" #if GOARCH_amd64 || GOARCH_386 // Syz-executor is linked against glibc when fuzzing runs on Cuttlefish x86-x64. // However Android blocks calls into mkdir, rmdir, symlink which causes // syz-executor to crash. When fuzzing runs on Android device this issue // is not observed, because syz-executor is linked against Bionic. Under // the hood Bionic invokes mkdirat, inlinkat and symlinkat, which are // allowed by seccomp-bpf. // This issue may exist not only in Android, but also in Linux in general // where seccomp filtering is enforced. // // This trick makes linker believe it matched the correct version of mkdir, // rmdir, symlink. So now behavior is the same across ARM and non-ARM builds. inline int mkdir(const char* path, mode_t mode) { return mkdirat(AT_FDCWD, path, mode); } inline int rmdir(const char* path) { return unlinkat(AT_FDCWD, path, AT_REMOVEDIR); } inline int symlink(const char* old_path, const char* new_path) { return symlinkat(old_path, AT_FDCWD, new_path); } #endif #endif #include // open(2) #include // setgroups #include // setxattr, getxattr #define AID_NET_BT_ADMIN 3001 #define AID_NET_BT 3002 #define AID_INET 3003 #define AID_EVERYBODY 9997 #define AID_APP 10000 #define UNTRUSTED_APP_UID (AID_APP + 999) #define UNTRUSTED_APP_GID (AID_APP + 999) #define SYSTEM_UID 1000 #define SYSTEM_GID 1000 const char* const SELINUX_CONTEXT_UNTRUSTED_APP = "u:r:untrusted_app:s0:c512,c768"; const char* const SELINUX_LABEL_APP_DATA_FILE = "u:object_r:app_data_file:s0:c512,c768"; const char* const SELINUX_CONTEXT_FILE = "/proc/thread-self/attr/current"; const char* const SELINUX_XATTR_NAME = "security.selinux"; const gid_t UNTRUSTED_APP_GROUPS[] = {UNTRUSTED_APP_GID, AID_NET_BT_ADMIN, AID_NET_BT, AID_INET, AID_EVERYBODY}; const size_t UNTRUSTED_APP_NUM_GROUPS = sizeof(UNTRUSTED_APP_GROUPS) / sizeof(UNTRUSTED_APP_GROUPS[0]); const gid_t SYSTEM_GROUPS[] = {SYSTEM_GID, AID_NET_BT_ADMIN, AID_NET_BT, AID_INET, AID_EVERYBODY}; const size_t SYSTEM_NUM_GROUPS = sizeof(SYSTEM_GROUPS) / sizeof(SYSTEM_GROUPS[0]); // Similar to libselinux getcon(3), but: // - No library dependency // - No dynamic memory allocation // - Uses fail() instead of returning an error code static void getcon(char* context, size_t context_size) { int fd = open(SELINUX_CONTEXT_FILE, O_RDONLY); if (fd < 0) fail("getcon: couldn't open context file"); ssize_t nread = read(fd, context, context_size); close(fd); if (nread <= 0) fail("getcon: failed to read context file"); // The contents of the context file MAY end with a newline // and MAY not have a null terminator. Handle this here. if (context[nread - 1] == '\n') context[nread - 1] = '\0'; } // Similar to libselinux setcon(3), but: // - No library dependency // - No dynamic memory allocation // - Uses fail() instead of returning an error code static void setcon(const char* context) { char new_context[512]; // Attempt to write the new context int fd = open(SELINUX_CONTEXT_FILE, O_WRONLY); if (fd < 0) fail("setcon: could not open context file"); ssize_t bytes_written = write(fd, context, strlen(context)); // N.B.: We cannot reuse this file descriptor, since the target SELinux context // may not be able to read from it. close(fd); if (bytes_written != (ssize_t)strlen(context)) failmsg("setcon: could not write entire context", "wrote=%zi, expected=%zu", bytes_written, strlen(context)); // Validate the transition by checking the context getcon(new_context, sizeof(new_context)); if (strcmp(context, new_context) != 0) failmsg("setcon: failed to change", "want=%s, context=%s", context, new_context); } // Similar to libselinux setfilecon(3), but: // - No library dependency // - No dynamic memory allocation // - Uses fail() instead of returning an error code static void setfilecon(const char* path, const char* context) { char new_context[512]; if (setxattr(path, SELINUX_XATTR_NAME, context, strlen(context) + 1, 0) != 0) fail("setfilecon: setxattr failed"); if (getxattr(path, SELINUX_XATTR_NAME, new_context, sizeof(new_context)) < 0) fail("setfilecon: getxattr failed"); if (strcmp(context, new_context) != 0) failmsg("setfilecon: could not set context", "want=%s, got=%s", context, new_context); } #define SYZ_HAVE_SANDBOX_ANDROID 1 static int do_sandbox_android(uint64 sandbox_arg) { setup_fusectl(); #if SYZ_EXECUTOR || SYZ_VHCI_INJECTION initialize_vhci(); #endif sandbox_common(); drop_caps(); #if SYZ_EXECUTOR || SYZ_NET_DEVICES initialize_netdevices_init(); #endif // CLONE_NEWNET must always happen before tun setup, because we want the tun // device in the test namespace. If we don't do this, executor will crash with // SYZFATAL: executor NUM failed NUM times: executor NUM: EOF if (unshare(CLONE_NEWNET)) { debug("unshare(CLONE_NEWNET): %d\n", errno); } // Enable access to IPPROTO_ICMP sockets, must be done after CLONE_NEWNET. write_file("/proc/sys/net/ipv4/ping_group_range", "0 65535"); #if SYZ_EXECUTOR || SYZ_DEVLINK_PCI initialize_devlink_pci(); #endif #if SYZ_EXECUTOR || SYZ_NET_INJECTION initialize_tun(); #endif #if SYZ_EXECUTOR || SYZ_NET_DEVICES initialize_netdevices(); #endif uid_t uid = UNTRUSTED_APP_UID; size_t num_groups = UNTRUSTED_APP_NUM_GROUPS; const gid_t* groups = UNTRUSTED_APP_GROUPS; gid_t gid = UNTRUSTED_APP_GID; debug("executor received sandbox_arg=%llu\n", sandbox_arg); if (sandbox_arg == 1) { uid = SYSTEM_UID; num_groups = SYSTEM_NUM_GROUPS; groups = SYSTEM_GROUPS; gid = SYSTEM_GID; debug("fuzzing under SYSTEM account\n"); } if (chown(".", uid, uid) != 0) failmsg("do_sandbox_android: chmod failed", "sandbox_arg=%llu", sandbox_arg); if (setgroups(num_groups, groups) != 0) failmsg("do_sandbox_android: setgroups failed", "sandbox_arg=%llu", sandbox_arg); if (setresgid(gid, gid, gid) != 0) failmsg("do_sandbox_android: setresgid failed", "sandbox_arg=%llu", sandbox_arg); setup_binderfs(); #if GOARCH_arm || GOARCH_arm64 || GOARCH_386 || GOARCH_amd64 // Will fail() if anything fails. // Must be called when the new process still has CAP_SYS_ADMIN, in this case, // before changing uid from 0, which clears capabilities. int account = SCFS_RestrictedApp; if (sandbox_arg == 1) account = SCFS_SystemAccount; set_app_seccomp_filter(account); #endif if (setresuid(uid, uid, uid) != 0) failmsg("do_sandbox_android: setresuid failed", "sandbox_arg=%llu", sandbox_arg); // setresuid and setresgid clear the parent-death signal. prctl(PR_SET_PDEATHSIG, SIGKILL, 0, 0, 0); setfilecon(".", SELINUX_LABEL_APP_DATA_FILE); if (uid == UNTRUSTED_APP_UID) setcon(SELINUX_CONTEXT_UNTRUSTED_APP); loop(); doexit(1); } #endif #if SYZ_EXECUTOR || SYZ_REPEAT && SYZ_USE_TMP_DIR #include #include #include #include #include #define FS_IOC_SETFLAGS _IOW('f', 2, long) // One does not simply remove a directory. // There can be mounts, so we need to try to umount. // Moreover, a mount can be mounted several times, so we need to try to umount in a loop. // Moreover, after umount a dir can become non-empty again, so we need another loop. // Moreover, a mount can be re-mounted as read-only and then we will fail to make a dir empty. static void remove_dir(const char* dir) { int iter = 0; DIR* dp = 0; #if SYZ_EXECUTOR || !SYZ_SANDBOX_ANDROID // Starting from v6.9, it does no longer make sense to use MNT_DETACH, because // a loop device may only be reused in RW mode if no mounted filesystem keeps a // reference to it. So we have to umount them synchronously. // MNT_FORCE should hopefully prevent hangs for filesystems that may require a complex cleanup. // // This declaration should not be moved under retry label, since label followed by a declaration // is not supported by old compilers. const int umount_flags = MNT_FORCE | UMOUNT_NOFOLLOW; #endif retry: #if SYZ_EXECUTOR || !SYZ_SANDBOX_ANDROID #if SYZ_EXECUTOR if (!flag_sandbox_android) #endif while (umount2(dir, umount_flags) == 0) { debug("umount(%s)\n", dir); } #endif dp = opendir(dir); if (dp == NULL) { if (errno == EMFILE) { // This happens when the test process casts prlimit(NOFILE) on us. // Ideally we somehow prevent test processes from messing with parent processes. // But full sandboxing is expensive, so let's ignore this error for now. exitf("opendir(%s) failed due to NOFILE, exiting", dir); } exitf("opendir(%s) failed", dir); } struct dirent* ep = 0; while ((ep = readdir(dp))) { if (strcmp(ep->d_name, ".") == 0 || strcmp(ep->d_name, "..") == 0) continue; char filename[FILENAME_MAX]; snprintf(filename, sizeof(filename), "%s/%s", dir, ep->d_name); // If it's 9p mount with broken transport, lstat will fail. // So try to umount first. #if SYZ_EXECUTOR || !SYZ_SANDBOX_ANDROID #if SYZ_EXECUTOR if (!flag_sandbox_android) #endif while (umount2(filename, umount_flags) == 0) { debug("umount(%s)\n", filename); } #endif struct stat st; if (lstat(filename, &st)) exitf("lstat(%s) failed", filename); if (S_ISDIR(st.st_mode)) { remove_dir(filename); continue; } int i; for (i = 0;; i++) { if (unlink(filename) == 0) break; if (errno == EPERM) { // Try to reset FS_XFLAG_IMMUTABLE. int fd = open(filename, O_RDONLY); if (fd != -1) { long flags = 0; if (ioctl(fd, FS_IOC_SETFLAGS, &flags) == 0) { debug("reset FS_XFLAG_IMMUTABLE\n"); } close(fd); continue; } } if (errno == EROFS) { debug("ignoring EROFS\n"); break; } if (errno != EBUSY || i > 100) exitf("unlink(%s) failed", filename); #if SYZ_EXECUTOR || !SYZ_SANDBOX_ANDROID #if SYZ_EXECUTOR if (!flag_sandbox_android) { #endif debug("umount(%s)\n", filename); if (umount2(filename, umount_flags)) exitf("umount(%s) failed", filename); #if SYZ_EXECUTOR } #endif #endif } } closedir(dp); for (int i = 0;; i++) { if (rmdir(dir) == 0) break; if (i < 100) { if (errno == EPERM) { // Try to reset FS_XFLAG_IMMUTABLE. int fd = open(dir, O_RDONLY); if (fd != -1) { long flags = 0; if (ioctl(fd, FS_IOC_SETFLAGS, &flags) == 0) { debug("reset FS_XFLAG_IMMUTABLE\n"); } close(fd); continue; } } if (errno == EROFS) { debug("ignoring EROFS\n"); break; } if (errno == EBUSY) { #if SYZ_EXECUTOR || !SYZ_SANDBOX_ANDROID #if SYZ_EXECUTOR if (!flag_sandbox_android) { #endif debug("umount(%s)\n", dir); if (umount2(dir, umount_flags)) exitf("umount(%s) failed", dir); #if SYZ_EXECUTOR } #endif #endif continue; } if (errno == ENOTEMPTY) { if (iter < 100) { iter++; goto retry; } } } exitf("rmdir(%s) failed", dir); } } #endif #if SYZ_EXECUTOR || SYZ_FAULT #include #include #include #include static int inject_fault(int nth) { int fd; fd = open("/proc/thread-self/fail-nth", O_RDWR); // We treat errors here as temporal/non-critical because we see // occasional ENOENT/EACCES errors returned. It seems that fuzzer // somehow gets its hands to it. if (fd == -1) exitf("failed to open /proc/thread-self/fail-nth"); char buf[16]; sprintf(buf, "%d", nth); if (write(fd, buf, strlen(buf)) != (ssize_t)strlen(buf)) exitf("failed to write /proc/thread-self/fail-nth"); return fd; } #endif #if SYZ_EXECUTOR static int fault_injected(int fail_fd) { char buf[16]; int n = read(fail_fd, buf, sizeof(buf) - 1); if (n <= 0) exitf("failed to read /proc/thread-self/fail-nth"); int res = n == 2 && buf[0] == '0' && buf[1] == '\n'; buf[0] = '0'; if (write(fail_fd, buf, 1) != 1) exitf("failed to write /proc/thread-self/fail-nth"); close(fail_fd); return res; } #endif #if (SYZ_EXECUTOR || SYZ_REPEAT) && SYZ_EXECUTOR_USES_FORK_SERVER #include #include #include #include #include #include #include #include static void kill_and_wait(int pid, int* status) { kill(-pid, SIGKILL); kill(pid, SIGKILL); // First, give it up to 100 ms to surrender. for (int i = 0; i < 100; i++) { if (waitpid(-1, status, WNOHANG | __WALL) == pid) return; usleep(1000); } // Now, try to abort fuse connections as they cause deadlocks, // see Documentation/filesystems/fuse.txt for details. // There is no good way to figure out the right connections // provided that the process could use unshare(CLONE_NEWNS), // so we abort all. debug("kill is not working\n"); DIR* dir = opendir("/sys/fs/fuse/connections"); if (dir) { for (;;) { struct dirent* ent = readdir(dir); if (!ent) break; if (strcmp(ent->d_name, ".") == 0 || strcmp(ent->d_name, "..") == 0) continue; char abort[300]; snprintf(abort, sizeof(abort), "/sys/fs/fuse/connections/%s/abort", ent->d_name); int fd = open(abort, O_WRONLY); if (fd == -1) { debug("failed to open %s: %d\n", abort, errno); continue; } debug("aborting fuse conn %s\n", ent->d_name); if (write(fd, abort, 1) < 0) { debug("failed to abort: %d\n", errno); } close(fd); } closedir(dir); } else { debug("failed to open /sys/fs/fuse/connections: %d\n", errno); } // Now, just wait, no other options. while (waitpid(-1, status, __WALL) != pid) { } } #endif #if (SYZ_EXECUTOR || SYZ_REPEAT && (SYZ_CGROUPS || SYZ_NET_RESET)) && SYZ_EXECUTOR_USES_FORK_SERVER #include #include #include #include #include #define SYZ_HAVE_SETUP_LOOP 1 static void setup_loop() { #if SYZ_EXECUTOR || SYZ_CGROUPS setup_cgroups_loop(); #endif #if SYZ_EXECUTOR || SYZ_NET_RESET checkpoint_net_namespace(); #endif } #endif #if (SYZ_EXECUTOR || SYZ_REPEAT && (SYZ_NET_RESET || __NR_syz_mount_image || __NR_syz_read_part_table)) && SYZ_EXECUTOR_USES_FORK_SERVER #define SYZ_HAVE_RESET_LOOP 1 static void reset_loop() { #if SYZ_EXECUTOR || __NR_syz_mount_image || __NR_syz_read_part_table char buf[64]; snprintf(buf, sizeof(buf), "/dev/loop%llu", procid); int loopfd = open(buf, O_RDWR); if (loopfd != -1) { ioctl(loopfd, LOOP_CLR_FD, 0); close(loopfd); } #endif #if SYZ_EXECUTOR || SYZ_NET_RESET reset_net_namespace(); #endif } #endif #if (SYZ_EXECUTOR || SYZ_REPEAT) && SYZ_EXECUTOR_USES_FORK_SERVER #include #include #define SYZ_HAVE_SETUP_TEST 1 static void setup_test() { prctl(PR_SET_PDEATHSIG, SIGKILL, 0, 0, 0); // We don't check for getppid() == 1 here b/c of unshare(CLONE_NEWPID), // our parent is normally pid 1. setpgrp(); #if SYZ_EXECUTOR || SYZ_CGROUPS setup_cgroups_test(); #endif // It's the leaf test process we want to be always killed first. write_file("/proc/self/oom_score_adj", "1000"); #if SYZ_EXECUTOR || SYZ_NET_INJECTION // Read all remaining packets from tun to better // isolate consequently executing programs. flush_tun(); #endif #if SYZ_EXECUTOR || SYZ_USE_TMP_DIR // Add a binderfs symlink to the tmp folder. if (symlink("/dev/binderfs", "./binderfs")) { debug("symlink(/dev/binderfs, ./binderfs) failed: %d", errno); } #endif } #endif #if SYZ_EXECUTOR || SYZ_CLOSE_FDS #include #define SYZ_HAVE_CLOSE_FDS 1 static void close_fds() { #if SYZ_EXECUTOR if (!flag_close_fds) return; #endif #ifdef SYS_close_range if (!syscall(SYS_close_range, 3, MAX_FDS, 0)) return; #endif // Keeping a 9p transport pipe open will hang the proccess dead, // so close all opened file descriptors. // Also close all USB emulation descriptors to trigger exit from USB // event loop to collect coverage. for (int fd = 3; fd < MAX_FDS; fd++) close(fd); } #endif #if SYZ_EXECUTOR || SYZ_FAULT #include static const char* setup_fault() { int fd = open("/proc/self/make-it-fail", O_WRONLY); if (fd == -1) return "CONFIG_FAULT_INJECTION is not enabled"; close(fd); fd = open("/proc/thread-self/fail-nth", O_WRONLY); if (fd == -1) return "kernel does not have systematic fault injection support"; close(fd); static struct { const char* file; const char* val; bool fatal; } files[] = { {"/sys/kernel/debug/failslab/ignore-gfp-wait", "N", true}, // These are enabled by separate configs (e.g. CONFIG_FAIL_FUTEX) // and we did not check all of them in host.checkFaultInjection, so we ignore errors. {"/sys/kernel/debug/fail_futex/ignore-private", "N", false}, {"/sys/kernel/debug/fail_page_alloc/ignore-gfp-highmem", "N", false}, {"/sys/kernel/debug/fail_page_alloc/ignore-gfp-wait", "N", false}, {"/sys/kernel/debug/fail_page_alloc/min-order", "0", false}, }; unsigned i; for (i = 0; i < sizeof(files) / sizeof(files[0]); i++) { if (!write_file(files[i].file, files[i].val)) { debug("failed to write %s: %d\n", files[i].file, errno); if (files[i].fatal) return "failed to write fault injection file"; } } return NULL; } #endif #if SYZ_EXECUTOR || SYZ_LEAK #include #include #include #include #include #define KMEMLEAK_FILE "/sys/kernel/debug/kmemleak" static const char* setup_leak() { if (!write_file(KMEMLEAK_FILE, "scan=off")) { if (errno == EBUSY) return "KMEMLEAK disabled: increase CONFIG_DEBUG_KMEMLEAK_EARLY_LOG_SIZE" " or unset CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF"; return "failed to write(kmemleak, \"scan=off\")"; } // Flush boot leaks. if (!write_file(KMEMLEAK_FILE, "scan")) return "failed to write(kmemleak, \"scan\")"; sleep(5); // account for MSECS_MIN_AGE if (!write_file(KMEMLEAK_FILE, "scan")) return "failed to write(kmemleak, \"scan\")"; if (!write_file(KMEMLEAK_FILE, "clear")) return "failed to write(kmemleak, \"clear\")"; return NULL; } #define SYZ_HAVE_LEAK_CHECK 1 #if SYZ_EXECUTOR static void check_leaks(char** frames, int nframes) #else static void check_leaks(void) #endif { int fd = open(KMEMLEAK_FILE, O_RDWR); if (fd == -1) fail("failed to open(kmemleak)"); // KMEMLEAK has false positives. To mitigate most of them, it checksums // potentially leaked objects, and reports them only on the next scan // iff the checksum does not change. Because of that we do the following // intricate dance: // Scan, sleep, scan again. At this point we can get some leaks. // If there are leaks, we sleep and scan again, this can remove // false leaks. Then, read kmemleak again. If we get leaks now, then // hopefully these are true positives during the previous testing cycle. uint64 start = current_time_ms(); if (write(fd, "scan", 4) != 4) fail("failed to write(kmemleak, \"scan\")"); sleep(1); // Account for MSECS_MIN_AGE // (1 second less because scanning will take at least a second). while (current_time_ms() - start < 4 * 1000) sleep(1); if (write(fd, "scan", 4) != 4) fail("failed to write(kmemleak, \"scan\")"); static char buf[128 << 10]; ssize_t n = read(fd, buf, sizeof(buf) - 1); if (n < 0) fail("failed to read(kmemleak)"); int nleaks = 0; if (n != 0) { sleep(1); if (write(fd, "scan", 4) != 4) fail("failed to write(kmemleak, \"scan\")"); if (lseek(fd, 0, SEEK_SET) < 0) fail("failed to lseek(kmemleak)"); n = read(fd, buf, sizeof(buf) - 1); if (n < 0) fail("failed to read(kmemleak)"); buf[n] = 0; char* pos = buf; char* end = buf + n; while (pos < end) { char* next = strstr(pos + 1, "unreferenced object"); if (!next) next = end; char prev = *next; *next = 0; #if SYZ_EXECUTOR int f; for (f = 0; f < nframes; f++) { if (strstr(pos, frames[f])) break; } if (f != nframes) { *next = prev; pos = next; continue; } #endif // BUG in output should be recognized by manager. fprintf(stderr, "BUG: memory leak\n%s\n", pos); *next = prev; pos = next; nleaks++; } } if (write(fd, "clear", 5) != 5) fail("failed to write(kmemleak, \"clear\")"); close(fd); if (nleaks) doexit(1); } #endif #if SYZ_EXECUTOR || SYZ_BINFMT_MISC #include #include #include #include static const char* setup_binfmt_misc() { // EBUSY means it's already mounted here. if (mount(0, "/proc/sys/fs/binfmt_misc", "binfmt_misc", 0, 0) && errno != EBUSY) { debug("mount(binfmt_misc) failed: %d\n", errno); return NULL; } if (!write_file("/proc/sys/fs/binfmt_misc/register", ":syz0:M:0:\x01::./file0:") || !write_file("/proc/sys/fs/binfmt_misc/register", ":syz1:M:1:\x02::./file0:POC")) return "write(/proc/sys/fs/binfmt_misc/register) failed"; return NULL; } #endif #if SYZ_EXECUTOR || SYZ_KCSAN static const char* setup_kcsan() { if (!write_file("/sys/kernel/debug/kcsan", "on")) return "write(/sys/kernel/debug/kcsan, on) failed"; return NULL; } #endif #if SYZ_EXECUTOR || SYZ_USB static const char* setup_usb() { if (chmod("/dev/raw-gadget", 0666)) return "failed to chmod /dev/raw-gadget"; return NULL; } #endif #if SYZ_EXECUTOR || SYZ_SYSCTL #include #include #include #include #include static void setup_sysctl() { // See ctrl-alt-del comment below. int cad_pid = fork(); if (cad_pid < 0) fail("fork failed"); if (cad_pid == 0) { for (;;) sleep(100); } char tmppid[32]; snprintf(tmppid, sizeof(tmppid), "%d", cad_pid); // TODO: consider moving all sysctl's into CMDLINE config later. // Kernel has support for setting sysctl's via command line since 3db978d480e28 (v5.8). struct { const char* name; const char* data; } files[] = { #if GOARCH_amd64 || GOARCH_386 // nmi_check_duration() prints "INFO: NMI handler took too long" on slow debug kernels. // It happens a lot in qemu, and the messages are frequently corrupted // (intermixed with other kernel output as they are printed from NMI) // and are not matched against the suppression in pkg/report. // This write prevents these messages from being printed. {"/sys/kernel/debug/x86/nmi_longest_ns", "10000000000"}, #endif {"/proc/sys/kernel/hung_task_check_interval_secs", "20"}, // bpf_jit_kallsyms and disabling bpf_jit_harden are required // for unwinding through bpf functions. {"/proc/sys/net/core/bpf_jit_kallsyms", "1"}, {"/proc/sys/net/core/bpf_jit_harden", "0"}, // This is to provide more useful info in crash reports. {"/proc/sys/kernel/kptr_restrict", "0"}, {"/proc/sys/kernel/softlockup_all_cpu_backtrace", "1"}, // This is to restrict effects of recursive exponential mounts, for details see // "mnt: Add a per mount namespace limit on the number of mounts" commit. {"/proc/sys/fs/mount-max", "100"}, // Dumping all tasks to console can take too long. {"/proc/sys/vm/oom_dump_tasks", "0"}, // Executor hits lots of SIGSEGVs, no point in logging them. {"/proc/sys/debug/exception-trace", "0"}, {"/proc/sys/kernel/printk", "7 4 1 3"}, // Faster gc (1 second) is intended to make tests more repeatable. {"/proc/sys/kernel/keys/gc_delay", "1"}, // We always want to prefer killing the allocating test process rather than somebody else // (sshd or another random test process). {"/proc/sys/vm/oom_kill_allocating_task", "1"}, // This blocks some of the ways the fuzzer can trigger a reboot. // ctrl-alt-del=0 tells kernel to signal cad_pid instead of rebooting. // We set cad_pid to a transient process pid ctrl-alt-del a no-op. // Note: we need to write a live process pid. // For context see: https://groups.google.com/g/syzkaller-bugs/c/WqOY4TiRnFg/m/6P9u8lWZAQAJ {"/proc/sys/kernel/ctrl-alt-del", "0"}, {"/proc/sys/kernel/cad_pid", tmppid}, }; for (size_t i = 0; i < sizeof(files) / sizeof(files[0]); i++) { if (!write_file(files[i].name, files[i].data)) { debug("write to %s failed: %s\n", files[i].name, strerror(errno)); } } kill(cad_pid, SIGKILL); while (waitpid(cad_pid, NULL, 0) != cad_pid) ; } #endif #if SYZ_EXECUTOR || SYZ_802154 #include #include #include #include #define NL802154_CMD_SET_SHORT_ADDR 11 #define NL802154_ATTR_IFINDEX 3 #define NL802154_ATTR_SHORT_ADDR 10 static const char* setup_802154() { const char* error = NULL; int sock_generic = -1; int sock_route = socket(AF_NETLINK, SOCK_RAW, NETLINK_ROUTE); if (sock_route == -1) { error = "socket(AF_NETLINK, SOCK_RAW, NETLINK_ROUTE) failed"; goto fail; } sock_generic = socket(AF_NETLINK, SOCK_RAW, NETLINK_GENERIC); if (sock_generic == -1) { error = "socket(AF_NETLINK, SOCK_RAW, NETLINK_GENERIC) failed"; goto fail; } { int nl802154_family_id = netlink_query_family_id(&nlmsg, sock_generic, "nl802154", true); if (nl802154_family_id < 0) { error = "netlink_query_family_id failed"; goto fail; } for (int i = 0; i < 2; i++) { // wpan0/1 are created by CONFIG_IEEE802154_HWSIM. // sys/linux/socket_ieee802154.txt knowns about these names and consts. char devname[] = "wpan0"; devname[strlen(devname) - 1] += i; uint64 hwaddr = 0xaaaaaaaaaaaa0002 + (i << 8); uint16 shortaddr = 0xaaa0 + i; int ifindex = if_nametoindex(devname); struct genlmsghdr genlhdr; memset(&genlhdr, 0, sizeof(genlhdr)); genlhdr.cmd = NL802154_CMD_SET_SHORT_ADDR; netlink_init(&nlmsg, nl802154_family_id, 0, &genlhdr, sizeof(genlhdr)); netlink_attr(&nlmsg, NL802154_ATTR_IFINDEX, &ifindex, sizeof(ifindex)); netlink_attr(&nlmsg, NL802154_ATTR_SHORT_ADDR, &shortaddr, sizeof(shortaddr)); if (netlink_send(&nlmsg, sock_generic) < 0) { error = "NL802154_CMD_SET_SHORT_ADDR failed"; goto fail; } netlink_device_change(&nlmsg, sock_route, devname, true, 0, &hwaddr, sizeof(hwaddr), 0); if (i == 0) { netlink_add_device_impl(&nlmsg, "lowpan", "lowpan0", false); netlink_done(&nlmsg); netlink_attr(&nlmsg, IFLA_LINK, &ifindex, sizeof(ifindex)); if (netlink_send(&nlmsg, sock_route) < 0) { error = "netlink: adding device lowpan0 type lowpan link wpan0"; goto fail; } } } } fail: close(sock_route); close(sock_generic); return error; } #endif #if GOARCH_s390x #include // Ugly way to work around gcc's "error: function called through a non-compatible type". // Simply casting via (void*) inline does not work b/c gcc sees through a chain of casts. // The macro is used in generated C code. #define CAST(f) ({void* p = (void*)f; p; }) #endif #if SYZ_EXECUTOR || __NR_syz_fuse_handle_req #include #include #include #include #include // From linux/fuse.h #define FUSE_MIN_READ_BUFFER 8192 // From linux/fuse.h enum fuse_opcode { FUSE_LOOKUP = 1, FUSE_FORGET = 2, // no reply FUSE_GETATTR = 3, FUSE_SETATTR = 4, FUSE_READLINK = 5, FUSE_SYMLINK = 6, FUSE_MKNOD = 8, FUSE_MKDIR = 9, FUSE_UNLINK = 10, FUSE_RMDIR = 11, FUSE_RENAME = 12, FUSE_LINK = 13, FUSE_OPEN = 14, FUSE_READ = 15, FUSE_WRITE = 16, FUSE_STATFS = 17, FUSE_RELEASE = 18, FUSE_FSYNC = 20, FUSE_SETXATTR = 21, FUSE_GETXATTR = 22, FUSE_LISTXATTR = 23, FUSE_REMOVEXATTR = 24, FUSE_FLUSH = 25, FUSE_INIT = 26, FUSE_OPENDIR = 27, FUSE_READDIR = 28, FUSE_RELEASEDIR = 29, FUSE_FSYNCDIR = 30, FUSE_GETLK = 31, FUSE_SETLK = 32, FUSE_SETLKW = 33, FUSE_ACCESS = 34, FUSE_CREATE = 35, FUSE_INTERRUPT = 36, FUSE_BMAP = 37, FUSE_DESTROY = 38, FUSE_IOCTL = 39, FUSE_POLL = 40, FUSE_NOTIFY_REPLY = 41, FUSE_BATCH_FORGET = 42, FUSE_FALLOCATE = 43, FUSE_READDIRPLUS = 44, FUSE_RENAME2 = 45, FUSE_LSEEK = 46, FUSE_COPY_FILE_RANGE = 47, FUSE_SETUPMAPPING = 48, FUSE_REMOVEMAPPING = 49, FUSE_SYNCFS = 50, FUSE_TMPFILE = 51, FUSE_STATX = 52, // CUSE specific operations CUSE_INIT = 4096, // Reserved opcodes: helpful to detect structure endian-ness CUSE_INIT_BSWAP_RESERVED = 1048576, // CUSE_INIT << 8 FUSE_INIT_BSWAP_RESERVED = 436207616, // FUSE_INIT << 24 }; // From linux/fuse.h struct fuse_in_header { uint32 len; uint32 opcode; uint64 unique; uint64 nodeid; uint32 uid; uint32 gid; uint32 pid; uint32 padding; }; // From linux/fuse.h struct fuse_out_header { uint32 len; // This is actually a int32_t but *_t variants fail to compile inside // the executor (it appends an additional _t for some reason) and int32 // does not exist. Since we don't touch this field, defining it as // unsigned should not cause any problems. uint32 error; uint64 unique; }; // Struct shared between syz_fuse_handle_req() and the fuzzer. Used to provide // a fuzzed response for each request type. struct syz_fuse_req_out { struct fuse_out_header* init; struct fuse_out_header* lseek; struct fuse_out_header* bmap; struct fuse_out_header* poll; struct fuse_out_header* getxattr; struct fuse_out_header* lk; struct fuse_out_header* statfs; struct fuse_out_header* write; struct fuse_out_header* read; struct fuse_out_header* open; struct fuse_out_header* attr; struct fuse_out_header* entry; struct fuse_out_header* dirent; struct fuse_out_header* direntplus; struct fuse_out_header* create_open; struct fuse_out_header* ioctl; struct fuse_out_header* statx; }; // Link the reponse to the request and send it to /dev/fuse. static int fuse_send_response(int fd, const struct fuse_in_header* in_hdr, struct fuse_out_header* out_hdr) { if (!out_hdr) { debug("fuse_send_response: received a NULL out_hdr\n"); return -1; } out_hdr->unique = in_hdr->unique; if (write(fd, out_hdr, out_hdr->len) == -1) { debug("fuse_send_response > write failed: %d\n", errno); return -1; } return 0; } // This function reads a request from /dev/fuse and tries to pick the correct // response from the input struct syz_fuse_req_out (a3). Responses are still // generated by the fuzzer. static volatile long syz_fuse_handle_req(volatile long a0, // /dev/fuse fd. volatile long a1, // Read buffer. volatile long a2, // Buffer len. volatile long a3) // syz_fuse_req_out. { struct syz_fuse_req_out* req_out = (struct syz_fuse_req_out*)a3; struct fuse_out_header* out_hdr = NULL; char* buf = (char*)a1; int buf_len = (int)a2; int fd = (int)a0; if (!req_out) { debug("syz_fuse_handle_req: received a NULL syz_fuse_req_out\n"); return -1; } if (buf_len < FUSE_MIN_READ_BUFFER) { debug("FUSE requires the read buffer to be at least %u\n", FUSE_MIN_READ_BUFFER); return -1; } int ret = read(fd, buf, buf_len); if (ret == -1) { debug("syz_fuse_handle_req > read failed: %d\n", errno); return -1; } // Safe to do because ret > 0 (!= -1) and < FUSE_MIN_READ_BUFFER (= 8192). if ((size_t)ret < sizeof(struct fuse_in_header)) { debug("syz_fuse_handle_req: received a truncated FUSE header\n"); return -1; } const struct fuse_in_header* in_hdr = (const struct fuse_in_header*)buf; debug("syz_fuse_handle_req: received opcode %d\n", in_hdr->opcode); if (in_hdr->len > (uint32)ret) { debug("syz_fuse_handle_req: received a truncated message\n"); return -1; } switch (in_hdr->opcode) { case FUSE_GETATTR: case FUSE_SETATTR: out_hdr = req_out->attr; break; case FUSE_LOOKUP: case FUSE_SYMLINK: case FUSE_LINK: case FUSE_MKNOD: case FUSE_MKDIR: out_hdr = req_out->entry; break; case FUSE_OPEN: case FUSE_OPENDIR: out_hdr = req_out->open; break; case FUSE_STATFS: out_hdr = req_out->statfs; break; case FUSE_RMDIR: case FUSE_RENAME: case FUSE_RENAME2: case FUSE_FALLOCATE: case FUSE_SETXATTR: case FUSE_REMOVEXATTR: case FUSE_FSYNCDIR: case FUSE_FSYNC: case FUSE_SETLKW: case FUSE_SETLK: case FUSE_ACCESS: case FUSE_FLUSH: case FUSE_RELEASE: case FUSE_RELEASEDIR: case FUSE_UNLINK: case FUSE_DESTROY: // These opcodes do not have any reply data. Hence, we pick // another response and only use the shared header. out_hdr = req_out->init; if (!out_hdr) { debug("syz_fuse_handle_req: received a NULL out_hdr\n"); return -1; } out_hdr->len = sizeof(struct fuse_out_header); break; case FUSE_READ: out_hdr = req_out->read; break; case FUSE_READDIR: out_hdr = req_out->dirent; break; case FUSE_READDIRPLUS: out_hdr = req_out->direntplus; break; case FUSE_INIT: out_hdr = req_out->init; break; case FUSE_LSEEK: out_hdr = req_out->lseek; break; case FUSE_GETLK: out_hdr = req_out->lk; break; case FUSE_BMAP: out_hdr = req_out->bmap; break; case FUSE_POLL: out_hdr = req_out->poll; break; case FUSE_GETXATTR: case FUSE_LISTXATTR: out_hdr = req_out->getxattr; break; case FUSE_WRITE: case FUSE_COPY_FILE_RANGE: out_hdr = req_out->write; break; case FUSE_FORGET: case FUSE_BATCH_FORGET: // FUSE_FORGET and FUSE_BATCH_FORGET expect no reply. return 0; case FUSE_CREATE: out_hdr = req_out->create_open; break; case FUSE_IOCTL: out_hdr = req_out->ioctl; break; case FUSE_STATX: out_hdr = req_out->statx; break; default: debug("syz_fuse_handle_req: unknown FUSE opcode\n"); return -1; } return fuse_send_response(fd, in_hdr, out_hdr); } #endif #if SYZ_EXECUTOR || __NR_syz_80211_inject_frame #include #include #include #include #include #include // This pseudo syscall performs 802.11 frame injection. // // Its current implementation performs the injection by means of mac80211_hwsim. // The procedure consists of the following steps: // 1. Open a netlink socket // 2. Register as an application responsible for wireless medium simulation by executing // HWSIM_CMD_REGISTER. This is a preq-requisite for the following step. After HWSIM_CMD_REGISTER // is executed, mac80211_hwsim stops simulating a perfect medium. // It is also important to note that this command registers a specific socket, not a netlink port. // 3. Inject a frame to the required interface by executing HWSIM_CMD_FRAME. // 4. Close the socket. mac80211_hwsim will detect this and return to perfect medium simulation. // // Note that we cannot (should not) open a socket, register it once and then use it for frame injection // throughout the lifetime of a proc. When some socket is registered, mac80211_hwsim does not broadcast // frames to all interfaces itself. As we do not perform this activity either, a permanently registered // socket will disrupt normal network operation. #define HWSIM_ATTR_RX_RATE 5 #define HWSIM_ATTR_SIGNAL 6 #define HWSIM_ATTR_ADDR_RECEIVER 1 #define HWSIM_ATTR_FRAME 3 #define WIFI_MAX_INJECT_LEN 2048 static int hwsim_register_socket(struct nlmsg* nlmsg, int sock, int hwsim_family) { struct genlmsghdr genlhdr; memset(&genlhdr, 0, sizeof(genlhdr)); genlhdr.cmd = HWSIM_CMD_REGISTER; netlink_init(nlmsg, hwsim_family, 0, &genlhdr, sizeof(genlhdr)); int err = netlink_send_ext(nlmsg, sock, 0, NULL, false); if (err < 0) { debug("hwsim_register_device failed: %s\n", strerror(errno)); } return err; } static int hwsim_inject_frame(struct nlmsg* nlmsg, int sock, int hwsim_family, uint8* mac_addr, uint8* data, int len) { struct genlmsghdr genlhdr; uint32 rx_rate = WIFI_DEFAULT_RX_RATE; uint32 signal = WIFI_DEFAULT_SIGNAL; memset(&genlhdr, 0, sizeof(genlhdr)); genlhdr.cmd = HWSIM_CMD_FRAME; netlink_init(nlmsg, hwsim_family, 0, &genlhdr, sizeof(genlhdr)); netlink_attr(nlmsg, HWSIM_ATTR_RX_RATE, &rx_rate, sizeof(rx_rate)); netlink_attr(nlmsg, HWSIM_ATTR_SIGNAL, &signal, sizeof(signal)); netlink_attr(nlmsg, HWSIM_ATTR_ADDR_RECEIVER, mac_addr, ETH_ALEN); netlink_attr(nlmsg, HWSIM_ATTR_FRAME, data, len); int err = netlink_send_ext(nlmsg, sock, 0, NULL, false); if (err < 0) { debug("hwsim_inject_frame failed: %s\n", strerror(errno)); } return err; } static long syz_80211_inject_frame(volatile long a0, volatile long a1, volatile long a2) { uint8* mac_addr = (uint8*)a0; uint8* buf = (uint8*)a1; int buf_len = (int)a2; struct nlmsg tmp_msg; if (buf_len < 0 || buf_len > WIFI_MAX_INJECT_LEN) { debug("syz_80211_inject_frame: wrong buffer size %d\n", buf_len); return -1; } int sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_GENERIC); if (sock < 0) { debug("syz_80211_inject_frame: socket creation failed, errno %d\n", errno); return -1; } int hwsim_family_id = netlink_query_family_id(&tmp_msg, sock, "MAC80211_HWSIM", false); if (hwsim_family_id < 0) { debug("syz_80211_inject_frame: failed to query family id\n"); close(sock); return -1; } int ret = hwsim_register_socket(&tmp_msg, sock, hwsim_family_id); if (ret < 0) { debug("syz_80211_inject_frame: failed to register socket, ret %d\n", ret); close(sock); return -1; } ret = hwsim_inject_frame(&tmp_msg, sock, hwsim_family_id, mac_addr, buf, buf_len); close(sock); if (ret < 0) { debug("syz_80211_inject_frame: failed to inject message, ret %d\n", ret); return -1; } return 0; } #endif #if SYZ_EXECUTOR || __NR_syz_80211_join_ibss #define WIFI_MAX_SSID_LEN 32 #define WIFI_JOIN_IBSS_NO_SCAN 0 #define WIFI_JOIN_IBSS_BG_SCAN 1 #define WIFI_JOIN_IBSS_BG_NO_SCAN 2 static long syz_80211_join_ibss(volatile long a0, volatile long a1, volatile long a2, volatile long a3) { char* interface = (char*)a0; uint8* ssid = (uint8*)a1; int ssid_len = (int)a2; int mode = (int)a3; // This parameter essentially determines whether it will perform a scan struct nlmsg tmp_msg; uint8 bssid[ETH_ALEN] = WIFI_IBSS_BSSID; if (ssid_len < 0 || ssid_len > WIFI_MAX_SSID_LEN) { debug("syz_80211_join_ibss: invalid ssid len %d\n", ssid_len); return -1; } if (mode < 0 || mode > WIFI_JOIN_IBSS_BG_NO_SCAN) { debug("syz_80211_join_ibss: invalid mode %d\n", mode); return -1; } int sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_GENERIC); if (sock < 0) { debug("syz_80211_join_ibss: socket creation failed, errno %d\n", errno); return -1; } int nl80211_family_id = netlink_query_family_id(&tmp_msg, sock, "nl80211", false); if (nl80211_family_id < 0) { debug("syz_80211_join_ibss: netlink_query_family_id failed\n"); close(sock); return -1; } struct join_ibss_props ibss_props = { .wiphy_freq = WIFI_DEFAULT_FREQUENCY, .wiphy_freq_fixed = (mode == WIFI_JOIN_IBSS_NO_SCAN || mode == WIFI_JOIN_IBSS_BG_NO_SCAN), .mac = bssid, .ssid = ssid, .ssid_len = ssid_len}; int ret = nl80211_setup_ibss_interface(&tmp_msg, sock, nl80211_family_id, interface, &ibss_props, false); close(sock); if (ret < 0) { debug("syz_80211_join_ibss: failed set up IBSS network for %.32s\n", interface); return -1; } if (mode == WIFI_JOIN_IBSS_NO_SCAN) { ret = await_ifla_operstate(&tmp_msg, interface, IF_OPER_UP, false); if (ret < 0) { debug("syz_80211_join_ibss: await_ifla_operstate failed for %.32s, ret %d\n", interface, ret); return -1; } } return 0; } #endif #if SYZ_EXECUTOR || __NR_syz_clone || __NR_syz_clone3 #if SYZ_EXECUTOR // The slowdown multiplier is already taken into account. #define USLEEP_FORKED_CHILD (3 * syscall_timeout_ms * 1000) #else #define USLEEP_FORKED_CHILD (3 * /*{{{BASE_CALL_TIMEOUT_MS}}}*/ *1000) #endif static long handle_clone_ret(long ret) { if (ret != 0) { #if SYZ_EXECUTOR || SYZ_HANDLE_SEGV __atomic_store_n(&clone_ongoing, 0, __ATOMIC_RELAXED); #endif return ret; } // Exit if we're in the child process - not all kernels provide the proper means // to prevent fork-bombs. // But first sleep for some time. This will hopefully foster IPC fuzzing. usleep(USLEEP_FORKED_CHILD); // Note that exit_group is a bad choice here because if we created just a thread, then // the whole process will be killed. A plain exit will work fine in any case. syscall(__NR_exit, 0); while (1) { } } #endif #if SYZ_EXECUTOR || __NR_syz_clone #include // syz_clone is mostly needed on kernels which do not suport clone3. static long syz_clone(volatile long flags, volatile long stack, volatile long stack_len, volatile long ptid, volatile long ctid, volatile long tls) { // ABI requires 16-byte stack alignment. long sp = (stack + stack_len) & ~15; #if SYZ_EXECUTOR || SYZ_HANDLE_SEGV __atomic_store_n(&clone_ongoing, 1, __ATOMIC_RELAXED); #endif // Clear the CLONE_VM flag. Otherwise it'll very likely corrupt syz-executor. long ret = (long)syscall(__NR_clone, flags & ~CLONE_VM, sp, ptid, ctid, tls); return handle_clone_ret(ret); } #endif #if SYZ_EXECUTOR || __NR_syz_clone3 #include #include #define MAX_CLONE_ARGS_BYTES 256 static long syz_clone3(volatile long a0, volatile long a1) { unsigned long copy_size = a1; if (copy_size < sizeof(uint64) || copy_size > MAX_CLONE_ARGS_BYTES) return -1; // The structure may have different sizes on different kernel versions, so copy it as raw bytes. char clone_args[MAX_CLONE_ARGS_BYTES]; memcpy(&clone_args, (void*)a0, copy_size); // As in syz_clone, clear the CLONE_VM flag. Flags are in the first 8-byte integer field. uint64* flags = (uint64*)&clone_args; *flags &= ~CLONE_VM; #if SYZ_EXECUTOR || SYZ_HANDLE_SEGV __atomic_store_n(&clone_ongoing, 1, __ATOMIC_RELAXED); #endif return handle_clone_ret((long)syscall(__NR_clone3, &clone_args, copy_size)); } #endif #if SYZ_EXECUTOR || __NR_syz_pkey_set #include #define RESERVED_PKEY 15 // syz_pkey_set(key pkey, val flags[pkey_flags]) static long syz_pkey_set(volatile long pkey, volatile long val) { #if GOARCH_amd64 || GOARCH_386 if (pkey == RESERVED_PKEY) { errno = EINVAL; return -1; } uint32 eax = 0; uint32 ecx = 0; asm volatile("rdpkru" : "=a"(eax) : "c"(ecx) : "edx"); // PKRU register contains 2 bits per key. // Max number of keys is 16. // Clear old bits for the key: eax &= ~(3 << ((pkey % 16) * 2)); // Set new bits for the key: eax |= (val & 3) << ((pkey % 16) * 2); uint32 edx = 0; asm volatile("wrpkru" ::"a"(eax), "c"(ecx), "d"(edx)); #endif return 0; } #endif #if SYZ_EXECUTOR || SYZ_SWAP #include #include #include #include #include #include #include #define SWAP_FILE "./swap-file" #define SWAP_FILE_SIZE (128 * 1000 * 1000) // 128 MB. static const char* setup_swap() { // The call must be idempotent, so first disable swap and remove the swap file. swapoff(SWAP_FILE); unlink(SWAP_FILE); // Zero-fill the file. int fd = open(SWAP_FILE, O_CREAT | O_WRONLY | O_CLOEXEC, 0600); if (fd == -1) return "swap file open failed"; // We cannot do ftruncate -- swapon complains about this. Do fallocate instead. fallocate(fd, FALLOC_FL_ZERO_RANGE, 0, SWAP_FILE_SIZE); close(fd); // Set up the swap file. char cmdline[64]; sprintf(cmdline, "mkswap %s", SWAP_FILE); if (runcmdline(cmdline)) return "mkswap failed"; if (swapon(SWAP_FILE, SWAP_FLAG_PREFER) == 1) return "swapon failed"; return NULL; } #endif #if SYZ_EXECUTOR || __NR_syz_pidfd_open #include // TODO: long-term we should improve our sandboxing rules since there are also // many other opportunities for a fuzzer process to access what it shouldn't. // Here we only shut down one of the recently discovered ways. static long syz_pidfd_open(volatile long pid, volatile long flags) { if (pid == 1) { // Under a PID namespace, pid=1 is the parent process. // We don't want a forked child to mangle parent syz-executor's fds. pid = 0; } return syscall(__NR_pidfd_open, pid, flags); } #endif #if SYZ_EXECUTOR || __NR_syz_kfuzztest_run #include #include #include #include #include #include #include #include static long syz_kfuzztest_run(volatile long test_name_ptr, volatile long input_data, volatile long input_data_size, volatile long buffer) { const char* test_name = (const char*)test_name_ptr; if (!test_name) { debug("syz_kfuzztest_run: test name was NULL\n"); return -1; } if (!buffer) { debug("syz_kfuzztest_run: buffer was NULL\n"); return -1; } char buf[256]; int ret = snprintf(buf, sizeof(buf), "/sys/kernel/debug/kfuzztest/%s/input", test_name); if (ret < 0 || (unsigned long)ret >= sizeof(buf)) { debug("syz_kfuzztest_run: constructed path is too long or snprintf failed\n"); return -1; } int fd = openat(AT_FDCWD, buf, O_WRONLY, 0); if (fd < 0) { debug("syz_kfuzztest_run: failed to open %s\n", buf); return -1; } ssize_t bytes_written = write(fd, (void*)buffer, (size_t)input_data_size); if (bytes_written != input_data_size) { debug("syz_kfuzztest_run: failed to write to %s, reason: %s\n", buf, strerror(errno)); close(fd); return -1; } if (close(fd) != 0) { debug("syz_kfuzztest_run: failed to close file\n"); return -1; } return 0; } #endif