diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-07 15:35:18 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-07 15:35:18 +0000 |
commit | b750101eb236130cf056c675997decbac904cc49 (patch) | |
tree | a5df1a06754bdd014cb975c051c83b01c9a97532 /src/libsystemd/sd-bus/bus-socket.c | |
parent | Initial commit. (diff) | |
download | systemd-b750101eb236130cf056c675997decbac904cc49.tar.xz systemd-b750101eb236130cf056c675997decbac904cc49.zip |
Adding upstream version 252.22.upstream/252.22upstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'src/libsystemd/sd-bus/bus-socket.c')
-rw-r--r-- | src/libsystemd/sd-bus/bus-socket.c | 1374 |
1 files changed, 1374 insertions, 0 deletions
diff --git a/src/libsystemd/sd-bus/bus-socket.c b/src/libsystemd/sd-bus/bus-socket.c new file mode 100644 index 0000000..c94befe --- /dev/null +++ b/src/libsystemd/sd-bus/bus-socket.c @@ -0,0 +1,1374 @@ +/* SPDX-License-Identifier: LGPL-2.1-or-later */ + +#include <endian.h> +#include <poll.h> +#include <stdlib.h> +#include <unistd.h> + +#include "sd-bus.h" +#include "sd-daemon.h" + +#include "alloc-util.h" +#include "bus-internal.h" +#include "bus-message.h" +#include "bus-socket.h" +#include "escape.h" +#include "fd-util.h" +#include "format-util.h" +#include "fs-util.h" +#include "hexdecoct.h" +#include "io-util.h" +#include "macro.h" +#include "memory-util.h" +#include "path-util.h" +#include "process-util.h" +#include "rlimit-util.h" +#include "signal-util.h" +#include "stdio-util.h" +#include "string-util.h" +#include "user-util.h" +#include "utf8.h" + +#define SNDBUF_SIZE (8*1024*1024) + +static void iovec_advance(struct iovec iov[], unsigned *idx, size_t size) { + + while (size > 0) { + struct iovec *i = iov + *idx; + + if (i->iov_len > size) { + i->iov_base = (uint8_t*) i->iov_base + size; + i->iov_len -= size; + return; + } + + size -= i->iov_len; + + *i = IOVEC_MAKE(NULL, 0); + + (*idx)++; + } +} + +static int append_iovec(sd_bus_message *m, const void *p, size_t sz) { + assert(m); + assert(p); + assert(sz > 0); + + m->iovec[m->n_iovec++] = IOVEC_MAKE((void*) p, sz); + + return 0; +} + +static int bus_message_setup_iovec(sd_bus_message *m) { + struct bus_body_part *part; + unsigned n, i; + int r; + + assert(m); + assert(m->sealed); + + if (m->n_iovec > 0) + return 0; + + assert(!m->iovec); + + n = 1 + m->n_body_parts; + if (n < ELEMENTSOF(m->iovec_fixed)) + m->iovec = m->iovec_fixed; + else { + m->iovec = new(struct iovec, n); + if (!m->iovec) { + r = -ENOMEM; + goto fail; + } + } + + r = append_iovec(m, m->header, BUS_MESSAGE_BODY_BEGIN(m)); + if (r < 0) + goto fail; + + MESSAGE_FOREACH_PART(part, i, m) { + r = bus_body_part_map(part); + if (r < 0) + goto fail; + + r = append_iovec(m, part->data, part->size); + if (r < 0) + goto fail; + } + + assert(n == m->n_iovec); + + return 0; + +fail: + m->poisoned = true; + return r; +} + +bool bus_socket_auth_needs_write(sd_bus *b) { + + unsigned i; + + if (b->auth_index >= ELEMENTSOF(b->auth_iovec)) + return false; + + for (i = b->auth_index; i < ELEMENTSOF(b->auth_iovec); i++) { + struct iovec *j = b->auth_iovec + i; + + if (j->iov_len > 0) + return true; + } + + return false; +} + +static int bus_socket_auth_verify_client(sd_bus *b) { + char *l, *lines[4] = {}; + sd_id128_t peer; + size_t i, n; + int r; + + assert(b); + + /* + * We expect up to three response lines: + * "DATA\r\n" (optional) + * "OK <server-id>\r\n" + * "AGREE_UNIX_FD\r\n" (optional) + */ + + n = 0; + lines[n] = b->rbuffer; + for (i = 0; i < 3; ++i) { + l = memmem_safe(lines[n], b->rbuffer_size - (lines[n] - (char*) b->rbuffer), "\r\n", 2); + if (l) + lines[++n] = l + 2; + else + break; + } + + /* + * If we sent a non-empty initial response, then we just expect an OK + * reply. We currently do this if, and only if, we picked ANONYMOUS. + * If we did not send an initial response, then we expect a DATA + * challenge, reply with our own DATA, and expect an OK reply. We do + * this for EXTERNAL. + * If FD negotiation was requested, we additionally expect + * an AGREE_UNIX_FD response in all cases. + */ + if (n < (b->anonymous_auth ? 1U : 2U) + !!b->accept_fd) + return 0; /* wait for more data */ + + i = 0; + + /* In case of EXTERNAL, verify the first response was DATA. */ + if (!b->anonymous_auth) { + l = lines[i++]; + if (lines[i] - l == 4 + 2) { + if (memcmp(l, "DATA", 4)) + return -EPERM; + } else if (lines[i] - l == 3 + 32 + 2) { + /* + * Old versions of the server-side implementation of + * `sd-bus` replied with "OK <id>" to "AUTH" requests + * from a client, even if the "AUTH" line did not + * contain inlined arguments. Therefore, we also accept + * "OK <id>" here, even though it is technically the + * wrong reply. We ignore the "<id>" parameter, though, + * since it has no real value. + */ + if (memcmp(l, "OK ", 3)) + return -EPERM; + } else + return -EPERM; + } + + /* Now check the OK line. */ + l = lines[i++]; + + if (lines[i] - l != 3 + 32 + 2) + return -EPERM; + if (memcmp(l, "OK ", 3)) + return -EPERM; + + b->auth = b->anonymous_auth ? BUS_AUTH_ANONYMOUS : BUS_AUTH_EXTERNAL; + + for (unsigned j = 0; j < 32; j += 2) { + int x, y; + + x = unhexchar(l[3 + j]); + y = unhexchar(l[3 + j + 1]); + + if (x < 0 || y < 0) + return -EINVAL; + + peer.bytes[j/2] = ((uint8_t) x << 4 | (uint8_t) y); + } + + if (!sd_id128_is_null(b->server_id) && + !sd_id128_equal(b->server_id, peer)) + return -EPERM; + + b->server_id = peer; + + /* And possibly check the third line, too */ + if (b->accept_fd) { + l = lines[i++]; + b->can_fds = !!memory_startswith(l, lines[i] - l, "AGREE_UNIX_FD"); + } + + assert(i == n); + + b->rbuffer_size -= (lines[i] - (char*) b->rbuffer); + memmove(b->rbuffer, lines[i], b->rbuffer_size); + + r = bus_start_running(b); + if (r < 0) + return r; + + return 1; +} + +static bool line_equals(const char *s, size_t m, const char *line) { + size_t l; + + l = strlen(line); + if (l != m) + return false; + + return memcmp(s, line, l) == 0; +} + +static bool line_begins(const char *s, size_t m, const char *word) { + const char *p; + + p = memory_startswith(s, m, word); + return p && (p == (s + m) || *p == ' '); +} + +static int verify_anonymous_token(sd_bus *b, const char *p, size_t l) { + _cleanup_free_ char *token = NULL; + size_t len; + int r; + + if (!b->anonymous_auth) + return 0; + + if (l <= 0) + return 1; + + assert(p[0] == ' '); + p++; l--; + + if (l % 2 != 0) + return 0; + + r = unhexmem(p, l, (void **) &token, &len); + if (r < 0) + return 0; + + if (memchr(token, 0, len)) + return 0; + + return !!utf8_is_valid(token); +} + +static int verify_external_token(sd_bus *b, const char *p, size_t l) { + _cleanup_free_ char *token = NULL; + size_t len; + uid_t u; + int r; + + /* We don't do any real authentication here. Instead, if + * the owner of this bus wanted authentication they should have + * checked SO_PEERCRED before even creating the bus object. */ + + if (!b->anonymous_auth && !b->ucred_valid) + return 0; + + if (l <= 0) + return 1; + + assert(p[0] == ' '); + p++; l--; + + if (l % 2 != 0) + return 0; + + r = unhexmem(p, l, (void**) &token, &len); + if (r < 0) + return 0; + + if (memchr(token, 0, len)) + return 0; + + r = parse_uid(token, &u); + if (r < 0) + return 0; + + /* We ignore the passed value if anonymous authentication is + * on anyway. */ + if (!b->anonymous_auth && u != b->ucred.uid) + return 0; + + return 1; +} + +static int bus_socket_auth_write(sd_bus *b, const char *t) { + char *p; + size_t l; + + assert(b); + assert(t); + + /* We only make use of the first iovec */ + assert(IN_SET(b->auth_index, 0, 1)); + + l = strlen(t); + p = malloc(b->auth_iovec[0].iov_len + l); + if (!p) + return -ENOMEM; + + memcpy_safe(p, b->auth_iovec[0].iov_base, b->auth_iovec[0].iov_len); + memcpy(p + b->auth_iovec[0].iov_len, t, l); + + b->auth_iovec[0].iov_base = p; + b->auth_iovec[0].iov_len += l; + + free(b->auth_buffer); + b->auth_buffer = p; + b->auth_index = 0; + return 0; +} + +static int bus_socket_auth_write_ok(sd_bus *b) { + char t[3 + 32 + 2 + 1]; + + assert(b); + + xsprintf(t, "OK " SD_ID128_FORMAT_STR "\r\n", SD_ID128_FORMAT_VAL(b->server_id)); + + return bus_socket_auth_write(b, t); +} + +static int bus_socket_auth_verify_server(sd_bus *b) { + char *e; + const char *line; + size_t l; + bool processed = false; + int r; + + assert(b); + + if (b->rbuffer_size < 1) + return 0; + + /* First char must be a NUL byte */ + if (*(char*) b->rbuffer != 0) + return -EIO; + + if (b->rbuffer_size < 3) + return 0; + + /* Begin with the first line */ + if (b->auth_rbegin <= 0) + b->auth_rbegin = 1; + + for (;;) { + /* Check if line is complete */ + line = (char*) b->rbuffer + b->auth_rbegin; + e = memmem_safe(line, b->rbuffer_size - b->auth_rbegin, "\r\n", 2); + if (!e) + return processed; + + l = e - line; + + if (line_begins(line, l, "AUTH ANONYMOUS")) { + + r = verify_anonymous_token(b, + line + strlen("AUTH ANONYMOUS"), + l - strlen("AUTH ANONYMOUS")); + if (r < 0) + return r; + if (r == 0) + r = bus_socket_auth_write(b, "REJECTED\r\n"); + else { + b->auth = BUS_AUTH_ANONYMOUS; + if (l <= strlen("AUTH ANONYMOUS")) + r = bus_socket_auth_write(b, "DATA\r\n"); + else + r = bus_socket_auth_write_ok(b); + } + + } else if (line_begins(line, l, "AUTH EXTERNAL")) { + + r = verify_external_token(b, + line + strlen("AUTH EXTERNAL"), + l - strlen("AUTH EXTERNAL")); + if (r < 0) + return r; + if (r == 0) + r = bus_socket_auth_write(b, "REJECTED\r\n"); + else { + b->auth = BUS_AUTH_EXTERNAL; + if (l <= strlen("AUTH EXTERNAL")) + r = bus_socket_auth_write(b, "DATA\r\n"); + else + r = bus_socket_auth_write_ok(b); + } + + } else if (line_begins(line, l, "AUTH")) + r = bus_socket_auth_write(b, "REJECTED EXTERNAL ANONYMOUS\r\n"); + else if (line_equals(line, l, "CANCEL") || + line_begins(line, l, "ERROR")) { + + b->auth = _BUS_AUTH_INVALID; + r = bus_socket_auth_write(b, "REJECTED\r\n"); + + } else if (line_equals(line, l, "BEGIN")) { + + if (b->auth == _BUS_AUTH_INVALID) + r = bus_socket_auth_write(b, "ERROR\r\n"); + else { + /* We can't leave from the auth phase + * before we haven't written + * everything queued, so let's check + * that */ + + if (bus_socket_auth_needs_write(b)) + return 1; + + b->rbuffer_size -= (e + 2 - (char*) b->rbuffer); + memmove(b->rbuffer, e + 2, b->rbuffer_size); + return bus_start_running(b); + } + + } else if (line_begins(line, l, "DATA")) { + + if (b->auth == _BUS_AUTH_INVALID) + r = bus_socket_auth_write(b, "ERROR\r\n"); + else { + if (b->auth == BUS_AUTH_ANONYMOUS) + r = verify_anonymous_token(b, line + 4, l - 4); + else + r = verify_external_token(b, line + 4, l - 4); + + if (r < 0) + return r; + if (r == 0) { + b->auth = _BUS_AUTH_INVALID; + r = bus_socket_auth_write(b, "REJECTED\r\n"); + } else + r = bus_socket_auth_write_ok(b); + } + } else if (line_equals(line, l, "NEGOTIATE_UNIX_FD")) { + if (b->auth == _BUS_AUTH_INVALID || !b->accept_fd) + r = bus_socket_auth_write(b, "ERROR\r\n"); + else { + b->can_fds = true; + r = bus_socket_auth_write(b, "AGREE_UNIX_FD\r\n"); + } + } else + r = bus_socket_auth_write(b, "ERROR\r\n"); + + if (r < 0) + return r; + + b->auth_rbegin = e + 2 - (char*) b->rbuffer; + + processed = true; + } +} + +static int bus_socket_auth_verify(sd_bus *b) { + assert(b); + + if (b->is_server) + return bus_socket_auth_verify_server(b); + else + return bus_socket_auth_verify_client(b); +} + +static int bus_socket_write_auth(sd_bus *b) { + ssize_t k; + + assert(b); + assert(b->state == BUS_AUTHENTICATING); + + if (!bus_socket_auth_needs_write(b)) + return 0; + + if (b->prefer_writev) + k = writev(b->output_fd, b->auth_iovec + b->auth_index, ELEMENTSOF(b->auth_iovec) - b->auth_index); + else { + struct msghdr mh = { + .msg_iov = b->auth_iovec + b->auth_index, + .msg_iovlen = ELEMENTSOF(b->auth_iovec) - b->auth_index, + }; + + k = sendmsg(b->output_fd, &mh, MSG_DONTWAIT|MSG_NOSIGNAL); + if (k < 0 && errno == ENOTSOCK) { + b->prefer_writev = true; + k = writev(b->output_fd, b->auth_iovec + b->auth_index, ELEMENTSOF(b->auth_iovec) - b->auth_index); + } + } + + if (k < 0) + return ERRNO_IS_TRANSIENT(errno) ? 0 : -errno; + + iovec_advance(b->auth_iovec, &b->auth_index, (size_t) k); + + /* Now crank the state machine since we might be able to make progress after writing. For example, + * the server only processes "BEGIN" when the write buffer is empty. + */ + return bus_socket_auth_verify(b); +} + +static int bus_socket_read_auth(sd_bus *b) { + struct msghdr mh; + struct iovec iov = {}; + size_t n; + ssize_t k; + int r; + void *p; + CMSG_BUFFER_TYPE(CMSG_SPACE(sizeof(int) * BUS_FDS_MAX)) control; + bool handle_cmsg = false; + + assert(b); + assert(b->state == BUS_AUTHENTICATING); + + r = bus_socket_auth_verify(b); + if (r != 0) + return r; + + n = MAX(256u, b->rbuffer_size * 2); + + if (n > BUS_AUTH_SIZE_MAX) + n = BUS_AUTH_SIZE_MAX; + + if (b->rbuffer_size >= n) + return -ENOBUFS; + + p = realloc(b->rbuffer, n); + if (!p) + return -ENOMEM; + + b->rbuffer = p; + + iov = IOVEC_MAKE((uint8_t *)b->rbuffer + b->rbuffer_size, n - b->rbuffer_size); + + if (b->prefer_readv) { + k = readv(b->input_fd, &iov, 1); + if (k < 0) + k = -errno; + } else { + mh = (struct msghdr) { + .msg_iov = &iov, + .msg_iovlen = 1, + .msg_control = &control, + .msg_controllen = sizeof(control), + }; + + k = recvmsg_safe(b->input_fd, &mh, MSG_DONTWAIT|MSG_CMSG_CLOEXEC); + if (k == -ENOTSOCK) { + b->prefer_readv = true; + k = readv(b->input_fd, &iov, 1); + if (k < 0) + k = -errno; + } else + handle_cmsg = true; + } + if (k < 0) { + if (ERRNO_IS_TRANSIENT(k)) + return 0; + return (int) k; + } + if (k == 0) { + if (handle_cmsg) + cmsg_close_all(&mh); /* paranoia, we shouldn't have gotten any fds on EOF */ + return -ECONNRESET; + } + + b->rbuffer_size += k; + + if (handle_cmsg) { + struct cmsghdr *cmsg; + + CMSG_FOREACH(cmsg, &mh) + if (cmsg->cmsg_level == SOL_SOCKET && + cmsg->cmsg_type == SCM_RIGHTS) { + int j; + + /* Whut? We received fds during the auth + * protocol? Somebody is playing games with + * us. Close them all, and fail */ + j = (cmsg->cmsg_len - CMSG_LEN(0)) / sizeof(int); + close_many((int*) CMSG_DATA(cmsg), j); + return -EIO; + } else + log_debug("Got unexpected auxiliary data with level=%d and type=%d", + cmsg->cmsg_level, cmsg->cmsg_type); + } + + r = bus_socket_auth_verify(b); + if (r != 0) + return r; + + return 1; +} + +void bus_socket_setup(sd_bus *b) { + assert(b); + + /* Increase the buffers to 8 MB */ + (void) fd_increase_rxbuf(b->input_fd, SNDBUF_SIZE); + (void) fd_inc_sndbuf(b->output_fd, SNDBUF_SIZE); + + b->message_version = 1; + b->message_endian = 0; +} + +static void bus_get_peercred(sd_bus *b) { + int r; + + assert(b); + assert(!b->ucred_valid); + assert(!b->label); + assert(b->n_groups == SIZE_MAX); + + /* Get the peer for socketpair() sockets */ + b->ucred_valid = getpeercred(b->input_fd, &b->ucred) >= 0; + + /* Get the SELinux context of the peer */ + r = getpeersec(b->input_fd, &b->label); + if (r < 0 && !IN_SET(r, -EOPNOTSUPP, -ENOPROTOOPT)) + log_debug_errno(r, "Failed to determine peer security context: %m"); + + /* Get the list of auxiliary groups of the peer */ + r = getpeergroups(b->input_fd, &b->groups); + if (r >= 0) + b->n_groups = (size_t) r; + else if (!IN_SET(r, -EOPNOTSUPP, -ENOPROTOOPT)) + log_debug_errno(r, "Failed to determine peer's group list: %m"); +} + +static int bus_socket_start_auth_client(sd_bus *b) { + static const char sasl_auth_anonymous[] = { + /* + * We use an arbitrary trace-string for the ANONYMOUS authentication. It can be used by the + * message broker to aid debugging of clients. We fully anonymize the connection and use a + * static default. + */ + /* HEX a n o n y m o u s */ + "\0AUTH ANONYMOUS 616e6f6e796d6f7573\r\n" + }; + static const char sasl_auth_external[] = { + "\0AUTH EXTERNAL\r\n" + "DATA\r\n" + }; + static const char sasl_negotiate_unix_fd[] = { + "NEGOTIATE_UNIX_FD\r\n" + }; + static const char sasl_begin[] = { + "BEGIN\r\n" + }; + size_t i = 0; + + assert(b); + + if (b->anonymous_auth) + b->auth_iovec[i++] = IOVEC_MAKE((char*) sasl_auth_anonymous, sizeof(sasl_auth_anonymous) - 1); + else + b->auth_iovec[i++] = IOVEC_MAKE((char*) sasl_auth_external, sizeof(sasl_auth_external) - 1); + + if (b->accept_fd) + b->auth_iovec[i++] = IOVEC_MAKE_STRING(sasl_negotiate_unix_fd); + + b->auth_iovec[i++] = IOVEC_MAKE_STRING(sasl_begin); + + return bus_socket_write_auth(b); +} + +int bus_socket_start_auth(sd_bus *b) { + assert(b); + + bus_get_peercred(b); + + bus_set_state(b, BUS_AUTHENTICATING); + b->auth_timeout = now(CLOCK_MONOTONIC) + BUS_AUTH_TIMEOUT; + + if (sd_is_socket(b->input_fd, AF_UNIX, 0, 0) <= 0) + b->accept_fd = false; + + if (b->output_fd != b->input_fd) + if (sd_is_socket(b->output_fd, AF_UNIX, 0, 0) <= 0) + b->accept_fd = false; + + if (b->is_server) + return bus_socket_read_auth(b); + else + return bus_socket_start_auth_client(b); +} + +static int bus_socket_inotify_setup(sd_bus *b) { + _cleanup_free_ int *new_watches = NULL; + _cleanup_free_ char *absolute = NULL; + size_t n = 0, done = 0, i; + unsigned max_follow = 32; + const char *p; + int wd, r; + + assert(b); + assert(b->watch_bind); + assert(b->sockaddr.sa.sa_family == AF_UNIX); + assert(b->sockaddr.un.sun_path[0] != 0); + + /* Sets up an inotify fd in case watch_bind is enabled: wait until the configured AF_UNIX file system socket + * appears before connecting to it. The implemented is pretty simplistic: we just subscribe to relevant changes + * to all prefix components of the path, and every time we get an event for that we try to reconnect again, + * without actually caring what precisely the event we got told us. If we still can't connect we re-subscribe + * to all relevant changes of anything in the path, so that our watches include any possibly newly created path + * components. */ + + if (b->inotify_fd < 0) { + b->inotify_fd = inotify_init1(IN_NONBLOCK|IN_CLOEXEC); + if (b->inotify_fd < 0) + return -errno; + + b->inotify_fd = fd_move_above_stdio(b->inotify_fd); + } + + /* Make sure the path is NUL terminated */ + p = strndupa_safe(b->sockaddr.un.sun_path, + sizeof(b->sockaddr.un.sun_path)); + + /* Make sure the path is absolute */ + r = path_make_absolute_cwd(p, &absolute); + if (r < 0) + goto fail; + + /* Watch all parent directories, and don't mind any prefix that doesn't exist yet. For the innermost directory + * that exists we want to know when files are created or moved into it. For all parents of it we just care if + * they are removed or renamed. */ + + if (!GREEDY_REALLOC(new_watches, n + 1)) { + r = -ENOMEM; + goto fail; + } + + /* Start with the top-level directory, which is a bit simpler than the rest, since it can't be a symlink, and + * always exists */ + wd = inotify_add_watch(b->inotify_fd, "/", IN_CREATE|IN_MOVED_TO); + if (wd < 0) { + r = log_debug_errno(errno, "Failed to add inotify watch on /: %m"); + goto fail; + } else + new_watches[n++] = wd; + + for (;;) { + _cleanup_free_ char *component = NULL, *prefix = NULL, *destination = NULL; + size_t n_slashes, n_component; + char *c = NULL; + + n_slashes = strspn(absolute + done, "/"); + n_component = n_slashes + strcspn(absolute + done + n_slashes, "/"); + + if (n_component == 0) /* The end */ + break; + + component = strndup(absolute + done, n_component); + if (!component) { + r = -ENOMEM; + goto fail; + } + + /* A trailing slash? That's a directory, and not a socket then */ + if (path_equal(component, "/")) { + r = -EISDIR; + goto fail; + } + + /* A single dot? Let's eat this up */ + if (path_equal(component, "/.")) { + done += n_component; + continue; + } + + prefix = strndup(absolute, done + n_component); + if (!prefix) { + r = -ENOMEM; + goto fail; + } + + if (!GREEDY_REALLOC(new_watches, n + 1)) { + r = -ENOMEM; + goto fail; + } + + wd = inotify_add_watch(b->inotify_fd, prefix, IN_DELETE_SELF|IN_MOVE_SELF|IN_ATTRIB|IN_CREATE|IN_MOVED_TO|IN_DONT_FOLLOW); + log_debug("Added inotify watch for %s on bus %s: %i", prefix, strna(b->description), wd); + + if (wd < 0) { + if (IN_SET(errno, ENOENT, ELOOP)) + break; /* This component doesn't exist yet, or the path contains a cyclic symlink right now */ + + r = log_debug_errno(errno, "Failed to add inotify watch on %s: %m", empty_to_root(prefix)); + goto fail; + } else + new_watches[n++] = wd; + + /* Check if this is possibly a symlink. If so, let's follow it and watch it too. */ + r = readlink_malloc(prefix, &destination); + if (r == -EINVAL) { /* not a symlink */ + done += n_component; + continue; + } + if (r < 0) + goto fail; + + if (isempty(destination)) { /* Empty symlink target? Yuck! */ + r = -EINVAL; + goto fail; + } + + if (max_follow <= 0) { /* Let's make sure we don't follow symlinks forever */ + r = -ELOOP; + goto fail; + } + + if (path_is_absolute(destination)) { + /* For absolute symlinks we build the new path and start anew */ + c = strjoin(destination, absolute + done + n_component); + done = 0; + } else { + _cleanup_free_ char *t = NULL; + + /* For relative symlinks we replace the last component, and try again */ + t = strndup(absolute, done); + if (!t) + return -ENOMEM; + + c = strjoin(t, "/", destination, absolute + done + n_component); + } + if (!c) { + r = -ENOMEM; + goto fail; + } + + free(absolute); + absolute = c; + + max_follow--; + } + + /* And now, let's remove all watches from the previous iteration we don't need anymore */ + for (i = 0; i < b->n_inotify_watches; i++) { + bool found = false; + size_t j; + + for (j = 0; j < n; j++) + if (new_watches[j] == b->inotify_watches[i]) { + found = true; + break; + } + + if (found) + continue; + + (void) inotify_rm_watch(b->inotify_fd, b->inotify_watches[i]); + } + + free_and_replace(b->inotify_watches, new_watches); + b->n_inotify_watches = n; + + return 0; + +fail: + bus_close_inotify_fd(b); + return r; +} + +int bus_socket_connect(sd_bus *b) { + bool inotify_done = false; + int r; + + assert(b); + + for (;;) { + assert(b->input_fd < 0); + assert(b->output_fd < 0); + assert(b->sockaddr.sa.sa_family != AF_UNSPEC); + + if (DEBUG_LOGGING) { + _cleanup_free_ char *pretty = NULL; + (void) sockaddr_pretty(&b->sockaddr.sa, b->sockaddr_size, false, true, &pretty); + log_debug("sd-bus: starting bus%s%s by connecting to %s...", + b->description ? " " : "", strempty(b->description), strnull(pretty)); + } + + b->input_fd = socket(b->sockaddr.sa.sa_family, SOCK_STREAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0); + if (b->input_fd < 0) + return -errno; + + b->input_fd = fd_move_above_stdio(b->input_fd); + + b->output_fd = b->input_fd; + bus_socket_setup(b); + + if (connect(b->input_fd, &b->sockaddr.sa, b->sockaddr_size) < 0) { + if (errno == EINPROGRESS) { + + /* If we have any inotify watches open, close them now, we don't need them anymore, as + * we have successfully initiated a connection */ + bus_close_inotify_fd(b); + + /* Note that very likely we are already in BUS_OPENING state here, as we enter it when + * we start parsing the address string. The only reason we set the state explicitly + * here, is to undo BUS_WATCH_BIND, in case we did the inotify magic. */ + bus_set_state(b, BUS_OPENING); + return 1; + } + + if (IN_SET(errno, ENOENT, ECONNREFUSED) && /* ENOENT → unix socket doesn't exist at all; ECONNREFUSED → unix socket stale */ + b->watch_bind && + b->sockaddr.sa.sa_family == AF_UNIX && + b->sockaddr.un.sun_path[0] != 0) { + + /* This connection attempt failed, let's release the socket for now, and start with a + * fresh one when reconnecting. */ + bus_close_io_fds(b); + + if (inotify_done) { + /* inotify set up already, don't do it again, just return now, and remember + * that we are waiting for inotify events now. */ + bus_set_state(b, BUS_WATCH_BIND); + return 1; + } + + /* This is a file system socket, and the inotify logic is enabled. Let's create the necessary inotify fd. */ + r = bus_socket_inotify_setup(b); + if (r < 0) + return r; + + /* Let's now try to connect a second time, because in theory there's otherwise a race + * here: the socket might have been created in the time between our first connect() and + * the time we set up the inotify logic. But let's remember that we set up inotify now, + * so that we don't do the connect() more than twice. */ + inotify_done = true; + + } else + return -errno; + } else + break; + } + + /* Yay, established, we don't need no inotify anymore! */ + bus_close_inotify_fd(b); + + return bus_socket_start_auth(b); +} + +int bus_socket_exec(sd_bus *b) { + int s[2], r; + + assert(b); + assert(b->input_fd < 0); + assert(b->output_fd < 0); + assert(b->exec_path); + assert(b->busexec_pid == 0); + + if (DEBUG_LOGGING) { + _cleanup_free_ char *line = NULL; + + if (b->exec_argv) + line = quote_command_line(b->exec_argv, SHELL_ESCAPE_EMPTY); + + log_debug("sd-bus: starting bus%s%s with %s%s", + b->description ? " " : "", strempty(b->description), + line ?: b->exec_path, + b->exec_argv && !line ? "…" : ""); + } + + r = socketpair(AF_UNIX, SOCK_STREAM|SOCK_NONBLOCK|SOCK_CLOEXEC, 0, s); + if (r < 0) + return -errno; + + r = safe_fork_full("(sd-busexec)", s+1, 1, FORK_RESET_SIGNALS|FORK_CLOSE_ALL_FDS, &b->busexec_pid); + if (r < 0) { + safe_close_pair(s); + return r; + } + if (r == 0) { + /* Child */ + + r = rearrange_stdio(s[1], s[1], STDERR_FILENO); + TAKE_FD(s[1]); + if (r < 0) + _exit(EXIT_FAILURE); + + (void) rlimit_nofile_safe(); + + if (b->exec_argv) + execvp(b->exec_path, b->exec_argv); + else + execvp(b->exec_path, STRV_MAKE(b->exec_path)); + + _exit(EXIT_FAILURE); + } + + safe_close(s[1]); + b->output_fd = b->input_fd = fd_move_above_stdio(s[0]); + + bus_socket_setup(b); + + return bus_socket_start_auth(b); +} + +int bus_socket_take_fd(sd_bus *b) { + assert(b); + + bus_socket_setup(b); + + return bus_socket_start_auth(b); +} + +int bus_socket_write_message(sd_bus *bus, sd_bus_message *m, size_t *idx) { + struct iovec *iov; + ssize_t k; + size_t n; + unsigned j; + int r; + + assert(bus); + assert(m); + assert(idx); + assert(IN_SET(bus->state, BUS_RUNNING, BUS_HELLO)); + + if (*idx >= BUS_MESSAGE_SIZE(m)) + return 0; + + r = bus_message_setup_iovec(m); + if (r < 0) + return r; + + n = m->n_iovec * sizeof(struct iovec); + iov = newa(struct iovec, n); + memcpy_safe(iov, m->iovec, n); + + j = 0; + iovec_advance(iov, &j, *idx); + + if (bus->prefer_writev) + k = writev(bus->output_fd, iov, m->n_iovec); + else { + struct msghdr mh = { + .msg_iov = iov, + .msg_iovlen = m->n_iovec, + }; + + if (m->n_fds > 0 && *idx == 0) { + struct cmsghdr *control; + + mh.msg_controllen = CMSG_SPACE(sizeof(int) * m->n_fds); + mh.msg_control = alloca0(mh.msg_controllen); + control = CMSG_FIRSTHDR(&mh); + control->cmsg_len = CMSG_LEN(sizeof(int) * m->n_fds); + control->cmsg_level = SOL_SOCKET; + control->cmsg_type = SCM_RIGHTS; + memcpy(CMSG_DATA(control), m->fds, sizeof(int) * m->n_fds); + } + + k = sendmsg(bus->output_fd, &mh, MSG_DONTWAIT|MSG_NOSIGNAL); + if (k < 0 && errno == ENOTSOCK) { + bus->prefer_writev = true; + k = writev(bus->output_fd, iov, m->n_iovec); + } + } + + if (k < 0) + return ERRNO_IS_TRANSIENT(errno) ? 0 : -errno; + + *idx += (size_t) k; + return 1; +} + +static int bus_socket_read_message_need(sd_bus *bus, size_t *need) { + uint32_t a, b; + uint8_t e; + uint64_t sum; + + assert(bus); + assert(need); + assert(IN_SET(bus->state, BUS_RUNNING, BUS_HELLO)); + + if (bus->rbuffer_size < sizeof(struct bus_header)) { + *need = sizeof(struct bus_header) + 8; + + /* Minimum message size: + * + * Header + + * + * Method Call: +2 string headers + * Signal: +3 string headers + * Method Error: +1 string headers + * +1 uint32 headers + * Method Reply: +1 uint32 headers + * + * A string header is at least 9 bytes + * A uint32 header is at least 8 bytes + * + * Hence the minimum message size of a valid message + * is header + 8 bytes */ + + return 0; + } + + a = ((const uint32_t*) bus->rbuffer)[1]; + b = ((const uint32_t*) bus->rbuffer)[3]; + + e = ((const uint8_t*) bus->rbuffer)[0]; + if (e == BUS_LITTLE_ENDIAN) { + a = le32toh(a); + b = le32toh(b); + } else if (e == BUS_BIG_ENDIAN) { + a = be32toh(a); + b = be32toh(b); + } else + return -EBADMSG; + + sum = (uint64_t) sizeof(struct bus_header) + (uint64_t) ALIGN8(b) + (uint64_t) a; + if (sum >= BUS_MESSAGE_SIZE_MAX) + return -ENOBUFS; + + *need = (size_t) sum; + return 0; +} + +static int bus_socket_make_message(sd_bus *bus, size_t size) { + sd_bus_message *t = NULL; + void *b; + int r; + + assert(bus); + assert(bus->rbuffer_size >= size); + assert(IN_SET(bus->state, BUS_RUNNING, BUS_HELLO)); + + r = bus_rqueue_make_room(bus); + if (r < 0) + return r; + + if (bus->rbuffer_size > size) { + b = memdup((const uint8_t*) bus->rbuffer + size, + bus->rbuffer_size - size); + if (!b) + return -ENOMEM; + } else + b = NULL; + + r = bus_message_from_malloc(bus, + bus->rbuffer, size, + bus->fds, bus->n_fds, + NULL, + &t); + if (r == -EBADMSG) { + log_debug_errno(r, "Received invalid message from connection %s, dropping.", strna(bus->description)); + free(bus->rbuffer); /* We want to drop current rbuffer and proceed with whatever remains in b */ + } else if (r < 0) { + free(b); + return r; + } + + /* rbuffer ownership was either transferred to t, or we got EBADMSG and dropped it. */ + bus->rbuffer = b; + bus->rbuffer_size -= size; + + bus->fds = NULL; + bus->n_fds = 0; + + if (t) { + t->read_counter = ++bus->read_counter; + bus->rqueue[bus->rqueue_size++] = bus_message_ref_queued(t, bus); + sd_bus_message_unref(t); + } + + return 1; +} + +int bus_socket_read_message(sd_bus *bus) { + struct msghdr mh; + struct iovec iov = {}; + ssize_t k; + size_t need; + int r; + void *b; + CMSG_BUFFER_TYPE(CMSG_SPACE(sizeof(int) * BUS_FDS_MAX)) control; + bool handle_cmsg = false; + + assert(bus); + assert(IN_SET(bus->state, BUS_RUNNING, BUS_HELLO)); + + r = bus_socket_read_message_need(bus, &need); + if (r < 0) + return r; + + if (bus->rbuffer_size >= need) + return bus_socket_make_message(bus, need); + + b = realloc(bus->rbuffer, need); + if (!b) + return -ENOMEM; + + bus->rbuffer = b; + + iov = IOVEC_MAKE((uint8_t *)bus->rbuffer + bus->rbuffer_size, need - bus->rbuffer_size); + + if (bus->prefer_readv) { + k = readv(bus->input_fd, &iov, 1); + if (k < 0) + k = -errno; + } else { + mh = (struct msghdr) { + .msg_iov = &iov, + .msg_iovlen = 1, + .msg_control = &control, + .msg_controllen = sizeof(control), + }; + + k = recvmsg_safe(bus->input_fd, &mh, MSG_DONTWAIT|MSG_CMSG_CLOEXEC); + if (k == -ENOTSOCK) { + bus->prefer_readv = true; + k = readv(bus->input_fd, &iov, 1); + if (k < 0) + k = -errno; + } else + handle_cmsg = true; + } + if (k < 0) { + if (ERRNO_IS_TRANSIENT(k)) + return 0; + return (int) k; + } + if (k == 0) { + if (handle_cmsg) + cmsg_close_all(&mh); /* On EOF we shouldn't have gotten an fd, but let's make sure */ + return -ECONNRESET; + } + + bus->rbuffer_size += k; + + if (handle_cmsg) { + struct cmsghdr *cmsg; + + CMSG_FOREACH(cmsg, &mh) + if (cmsg->cmsg_level == SOL_SOCKET && + cmsg->cmsg_type == SCM_RIGHTS) { + int n, *f, i; + + n = (cmsg->cmsg_len - CMSG_LEN(0)) / sizeof(int); + + if (!bus->can_fds) { + /* Whut? We received fds but this + * isn't actually enabled? Close them, + * and fail */ + + close_many((int*) CMSG_DATA(cmsg), n); + return -EIO; + } + + f = reallocarray(bus->fds, bus->n_fds + n, sizeof(int)); + if (!f) { + close_many((int*) CMSG_DATA(cmsg), n); + return -ENOMEM; + } + + for (i = 0; i < n; i++) + f[bus->n_fds++] = fd_move_above_stdio(((int*) CMSG_DATA(cmsg))[i]); + bus->fds = f; + } else + log_debug("Got unexpected auxiliary data with level=%d and type=%d", + cmsg->cmsg_level, cmsg->cmsg_type); + } + + r = bus_socket_read_message_need(bus, &need); + if (r < 0) + return r; + + if (bus->rbuffer_size >= need) + return bus_socket_make_message(bus, need); + + return 1; +} + +int bus_socket_process_opening(sd_bus *b) { + int error = 0, events, r; + socklen_t slen = sizeof(error); + + assert(b->state == BUS_OPENING); + + events = fd_wait_for_event(b->output_fd, POLLOUT, 0); + if (events < 0) + return events; + if (!(events & (POLLOUT|POLLERR|POLLHUP))) + return 0; + + r = getsockopt(b->output_fd, SOL_SOCKET, SO_ERROR, &error, &slen); + if (r < 0) + b->last_connect_error = errno; + else if (error != 0) + b->last_connect_error = error; + else if (events & (POLLERR|POLLHUP)) + b->last_connect_error = ECONNREFUSED; + else + return bus_socket_start_auth(b); + + return bus_next_address(b); +} + +int bus_socket_process_authenticating(sd_bus *b) { + int r; + + assert(b); + assert(b->state == BUS_AUTHENTICATING); + + if (now(CLOCK_MONOTONIC) >= b->auth_timeout) + return -ETIMEDOUT; + + r = bus_socket_write_auth(b); + if (r != 0) + return r; + + return bus_socket_read_auth(b); +} + +int bus_socket_process_watch_bind(sd_bus *b) { + int r, q; + + assert(b); + assert(b->state == BUS_WATCH_BIND); + assert(b->inotify_fd >= 0); + + r = flush_fd(b->inotify_fd); + if (r <= 0) + return r; + + log_debug("Got inotify event on bus %s.", strna(b->description)); + + /* We flushed events out of the inotify fd. In that case, maybe the socket is valid now? Let's try to connect + * to it again */ + + r = bus_socket_connect(b); + if (r < 0) + return r; + + q = bus_attach_io_events(b); + if (q < 0) + return q; + + q = bus_attach_inotify_event(b); + if (q < 0) + return q; + + return r; +} |