diff options
author | Lennart Poettering <lennart@poettering.net> | 2013-11-19 21:12:59 +0100 |
---|---|---|
committer | Lennart Poettering <lennart@poettering.net> | 2013-11-20 20:52:36 +0100 |
commit | 718db96199eb307751264e4163555662c9a389fa (patch) | |
tree | 9ec8467596ba1acba76bb6273c7797baf68c1a00 /src/core/manager.c | |
parent | 3febea3a0b0a968ea281e7959c1654cbaf95c9bf (diff) |
core: convert PID 1 to libsystemd-bus
This patch converts PID 1 to libsystemd-bus and thus drops the
dependency on libdbus. The only remaining code using libdbus is a test
case that validates our bus marshalling against libdbus' marshalling,
and this dependency can be turned off.
This patch also adds a couple of things to libsystem-bus, that are
necessary to make the port work:
- Synthesizing of "Disconnected" messages when bus connections are
severed.
- Support for attaching multiple vtables for the same interface on the
same path.
This patch also fixes the SetDefaultTarget() and GetDefaultTarget() bus
calls which used an inappropriate signature.
As a side effect we will now generate PropertiesChanged messages which
carry property contents, rather than just invalidation information.
Diffstat (limited to 'src/core/manager.c')
-rw-r--r-- | src/core/manager.c | 652 |
1 files changed, 239 insertions, 413 deletions
diff --git a/src/core/manager.c b/src/core/manager.c index c99a022cd5..86de0e3fbf 100644 --- a/src/core/manager.c +++ b/src/core/manager.c @@ -42,9 +42,9 @@ #include <libaudit.h> #endif -#include "systemd/sd-daemon.h" -#include "systemd/sd-id128.h" -#include "systemd/sd-messages.h" +#include "sd-daemon.h" +#include "sd-id128.h" +#include "sd-messages.h" #include "manager.h" #include "transaction.h" @@ -58,12 +58,9 @@ #include "locale-setup.h" #include "mount-setup.h" #include "unit-name.h" -#include "dbus-unit.h" -#include "dbus-job.h" #include "missing.h" #include "path-lookup.h" #include "special.h" -#include "bus-errors.h" #include "exit-status.h" #include "virt.h" #include "watchdog.h" @@ -72,6 +69,13 @@ #include "audit-fd.h" #include "boot-timestamps.h" #include "env-util.h" +#include "bus-errors.h" +#include "bus-error.h" +#include "bus-util.h" +#include "dbus.h" +#include "dbus-unit.h" +#include "dbus-job.h" +#include "dbus-manager.h" /* As soon as 5s passed since a unit was added to our GC queue, make sure to run a gc sweep */ #define GC_QUEUE_USEC_MAX (10*USEC_PER_SEC) @@ -86,6 +90,12 @@ #define TIME_T_MAX (time_t)((1UL << ((sizeof(time_t) << 3) - 1)) - 1) +static int manager_dispatch_notify_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata); +static int manager_dispatch_signal_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata); +static int manager_dispatch_time_change_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata); +static int manager_dispatch_idle_pipe_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata); +static int manager_dispatch_jobs_in_progress(sd_event_source *source, usec_t usec, void *userdata); + static int manager_setup_notify(Manager *m) { union { struct sockaddr sa; @@ -93,15 +103,10 @@ static int manager_setup_notify(Manager *m) { } sa = { .sa.sa_family = AF_UNIX, }; - struct epoll_event ev = { - .events = EPOLLIN, - .data.ptr = &m->notify_watch, - }; int one = 1, r; - m->notify_watch.type = WATCH_NOTIFY; - m->notify_watch.fd = socket(AF_UNIX, SOCK_DGRAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0); - if (m->notify_watch.fd < 0) { + m->notify_fd = socket(AF_UNIX, SOCK_DGRAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0); + if (m->notify_fd < 0) { log_error("Failed to allocate notification socket: %m"); return -errno; } @@ -110,25 +115,23 @@ static int manager_setup_notify(Manager *m) { snprintf(sa.un.sun_path, sizeof(sa.un.sun_path), NOTIFY_SOCKET "/%llu", random_ull()); else strncpy(sa.un.sun_path, NOTIFY_SOCKET, sizeof(sa.un.sun_path)); - sa.un.sun_path[0] = 0; - r = bind(m->notify_watch.fd, &sa.sa, - offsetof(struct sockaddr_un, sun_path) + 1 + strlen(sa.un.sun_path+1)); + r = bind(m->notify_fd, &sa.sa, offsetof(struct sockaddr_un, sun_path) + 1 + strlen(sa.un.sun_path+1)); if (r < 0) { log_error("bind() failed: %m"); return -errno; } - r = setsockopt(m->notify_watch.fd, SOL_SOCKET, SO_PASSCRED, &one, sizeof(one)); + r = setsockopt(m->notify_fd, SOL_SOCKET, SO_PASSCRED, &one, sizeof(one)); if (r < 0) { log_error("SO_PASSCRED failed: %m"); return -errno; } - r = epoll_ctl(m->epoll_fd, EPOLL_CTL_ADD, m->notify_watch.fd, &ev); + r = sd_event_add_io(m->event, m->notify_fd, EPOLLIN, manager_dispatch_notify_fd, m, &m->notify_event_source); if (r < 0) { - log_error("Failed to add notification socket fd to epoll: %m"); + log_error("Failed to allocate notify event source: %s", strerror(-r)); return -errno; } @@ -142,75 +145,17 @@ static int manager_setup_notify(Manager *m) { return 0; } -static int manager_jobs_in_progress_mod_timer(Manager *m) { - struct itimerspec its = { - .it_value.tv_sec = JOBS_IN_PROGRESS_WAIT_SEC, - .it_interval.tv_sec = JOBS_IN_PROGRESS_PERIOD_SEC, - }; - - if (m->jobs_in_progress_watch.type != WATCH_JOBS_IN_PROGRESS) - return 0; - - if (timerfd_settime(m->jobs_in_progress_watch.fd, 0, &its, NULL) < 0) - return -errno; - - return 0; -} - static int manager_watch_jobs_in_progress(Manager *m) { - struct epoll_event ev = { - .events = EPOLLIN, - .data.ptr = &m->jobs_in_progress_watch, - }; - int r; + assert(m); - if (m->jobs_in_progress_watch.type != WATCH_INVALID) + if (m->jobs_in_progress_event_source) return 0; - m->jobs_in_progress_watch.type = WATCH_JOBS_IN_PROGRESS; - m->jobs_in_progress_watch.fd = timerfd_create(CLOCK_MONOTONIC, TFD_NONBLOCK|TFD_CLOEXEC); - if (m->jobs_in_progress_watch.fd < 0) { - log_error("Failed to create timerfd: %m"); - r = -errno; - goto err; - } - - r = manager_jobs_in_progress_mod_timer(m); - if (r < 0) { - log_error("Failed to set up timer for jobs progress watch: %s", strerror(-r)); - goto err; - } - - if (epoll_ctl(m->epoll_fd, EPOLL_CTL_ADD, m->jobs_in_progress_watch.fd, &ev) < 0) { - log_error("Failed to add jobs progress timer fd to epoll: %m"); - r = -errno; - goto err; - } - - log_debug("Set up jobs progress timerfd."); - - return 0; - -err: - if (m->jobs_in_progress_watch.fd >= 0) - close_nointr_nofail(m->jobs_in_progress_watch.fd); - watch_init(&m->jobs_in_progress_watch); - return r; + return sd_event_add_monotonic(m->event, JOBS_IN_PROGRESS_WAIT_SEC, 0, manager_dispatch_jobs_in_progress, m, &m->jobs_in_progress_event_source); } -static void manager_unwatch_jobs_in_progress(Manager *m) { - if (m->jobs_in_progress_watch.type != WATCH_JOBS_IN_PROGRESS) - return; - - assert_se(epoll_ctl(m->epoll_fd, EPOLL_CTL_DEL, m->jobs_in_progress_watch.fd, NULL) >= 0); - close_nointr_nofail(m->jobs_in_progress_watch.fd); - watch_init(&m->jobs_in_progress_watch); - m->jobs_in_progress_iteration = 0; - - log_debug("Closed jobs progress timerfd."); -} +#define CYLON_BUFFER_EXTRA (2*(sizeof(ANSI_RED_ON)-1) + sizeof(ANSI_HIGHLIGHT_RED_ON)-1 + 2*(sizeof(ANSI_HIGHLIGHT_OFF)-1)) -#define CYLON_BUFFER_EXTRA (2*strlen(ANSI_RED_ON) + strlen(ANSI_HIGHLIGHT_RED_ON) + 2*strlen(ANSI_HIGHLIGHT_OFF)) static void draw_cylon(char buffer[], size_t buflen, unsigned width, unsigned pos) { char *p = buffer; @@ -241,13 +186,15 @@ static void draw_cylon(char buffer[], size_t buflen, unsigned width, unsigned po } static void manager_print_jobs_in_progress(Manager *m) { + _cleanup_free_ char *job_of_n = NULL; Iterator i; Job *j; - char *job_of_n = NULL; unsigned counter = 0, print_nr; char cylon[6 + CYLON_BUFFER_EXTRA + 1]; unsigned cylon_pos; + assert(m); + print_nr = (m->jobs_in_progress_iteration / JOBS_IN_PROGRESS_PERIOD_DIVISOR) % m->n_running_jobs; HASHMAP_FOREACH(j, m->jobs, i) @@ -270,94 +217,69 @@ static void manager_print_jobs_in_progress(Manager *m) { manager_status_printf(m, true, cylon, "%sA %s job is running for %s", strempty(job_of_n), job_type_to_string(j->type), unit_description(j->unit)); - free(job_of_n); m->jobs_in_progress_iteration++; } static int manager_watch_idle_pipe(Manager *m) { - struct epoll_event ev = { - .events = EPOLLIN, - .data.ptr = &m->idle_pipe_watch, - }; int r; - if (m->idle_pipe_watch.type != WATCH_INVALID) + assert(m); + + if (m->idle_pipe_event_source) return 0; if (m->idle_pipe[2] < 0) return 0; - m->idle_pipe_watch.type = WATCH_IDLE_PIPE; - m->idle_pipe_watch.fd = m->idle_pipe[2]; - if (m->idle_pipe_watch.fd < 0) { - log_error("Failed to create timerfd: %m"); - r = -errno; - goto err; - } - - if (epoll_ctl(m->epoll_fd, EPOLL_CTL_ADD, m->idle_pipe_watch.fd, &ev) < 0) { - log_error("Failed to add idle_pipe fd to epoll: %m"); - r = -errno; - goto err; + r = sd_event_add_io(m->event, m->idle_pipe[2], EPOLLIN, manager_dispatch_idle_pipe_fd, m, &m->idle_pipe_event_source); + if (r < 0) { + log_error("Failed to watch idle pipe: %s", strerror(-r)); + return r; } - log_debug("Set up idle_pipe watch."); - return 0; - -err: - if (m->idle_pipe_watch.fd >= 0) - close_nointr_nofail(m->idle_pipe_watch.fd); - watch_init(&m->idle_pipe_watch); - return r; } -static void manager_unwatch_idle_pipe(Manager *m) { - if (m->idle_pipe_watch.type != WATCH_IDLE_PIPE) - return; - - assert_se(epoll_ctl(m->epoll_fd, EPOLL_CTL_DEL, m->idle_pipe_watch.fd, NULL) >= 0); - watch_init(&m->idle_pipe_watch); +static void manager_close_idle_pipe(Manager *m) { + assert(m); - log_debug("Closed idle_pipe watch."); + close_pipe(m->idle_pipe); + close_pipe(m->idle_pipe + 2); } static int manager_setup_time_change(Manager *m) { - struct epoll_event ev = { - .events = EPOLLIN, - .data.ptr = &m->time_change_watch, - }; + int r; /* We only care for the cancellation event, hence we set the * timeout to the latest possible value. */ struct itimerspec its = { .it_value.tv_sec = TIME_T_MAX, }; - assert_cc(sizeof(time_t) == sizeof(TIME_T_MAX)); - assert(m->time_change_watch.type == WATCH_INVALID); + assert(m); + assert_cc(sizeof(time_t) == sizeof(TIME_T_MAX)); /* Uses TFD_TIMER_CANCEL_ON_SET to get notifications whenever * CLOCK_REALTIME makes a jump relative to CLOCK_MONOTONIC */ - m->time_change_watch.type = WATCH_TIME_CHANGE; - m->time_change_watch.fd = timerfd_create(CLOCK_REALTIME, TFD_NONBLOCK|TFD_CLOEXEC); - if (m->time_change_watch.fd < 0) { + m->time_change_fd = timerfd_create(CLOCK_REALTIME, TFD_NONBLOCK|TFD_CLOEXEC); + if (m->time_change_fd < 0) { log_error("Failed to create timerfd: %m"); return -errno; } - if (timerfd_settime(m->time_change_watch.fd, TFD_TIMER_ABSTIME|TFD_TIMER_CANCEL_ON_SET, &its, NULL) < 0) { + if (timerfd_settime(m->time_change_fd, TFD_TIMER_ABSTIME|TFD_TIMER_CANCEL_ON_SET, &its, NULL) < 0) { log_debug("Failed to set up TFD_TIMER_CANCEL_ON_SET, ignoring: %m"); - close_nointr_nofail(m->time_change_watch.fd); - watch_init(&m->time_change_watch); + close_nointr_nofail(m->time_change_fd); + m->time_change_fd = -1; return 0; } - if (epoll_ctl(m->epoll_fd, EPOLL_CTL_ADD, m->time_change_watch.fd, &ev) < 0) { - log_error("Failed to add timer change fd to epoll: %m"); - return -errno; + r = sd_event_add_io(m->event, m->time_change_fd, EPOLLIN, manager_dispatch_time_change_fd, m, &m->time_change_event_source); + if (r < 0) { + log_error("Failed to create time change event source: %s", strerror(-r)); + return r; } log_debug("Set up TFD_TIMER_CANCEL_ON_SET timerfd."); @@ -366,7 +288,7 @@ static int manager_setup_time_change(Manager *m) { } static int enable_special_signals(Manager *m) { - int fd; + _cleanup_close_ int fd = -1; assert(m); @@ -385,23 +307,18 @@ static int enable_special_signals(Manager *m) { /* Enable that we get SIGWINCH on kbrequest */ if (ioctl(fd, KDSIGACCEPT, SIGWINCH) < 0) log_warning("Failed to enable kbrequest handling: %s", strerror(errno)); - - close_nointr_nofail(fd); } return 0; } static int manager_setup_signals(Manager *m) { - sigset_t mask; - struct epoll_event ev = { - .events = EPOLLIN, - .data.ptr = &m->signal_watch, - }; struct sigaction sa = { .sa_handler = SIG_DFL, .sa_flags = SA_NOCLDSTOP|SA_RESTART, }; + sigset_t mask; + int r; assert(m); @@ -442,13 +359,13 @@ static int manager_setup_signals(Manager *m) { -1); assert_se(sigprocmask(SIG_SETMASK, &mask, NULL) == 0); - m->signal_watch.type = WATCH_SIGNAL; - m->signal_watch.fd = signalfd(-1, &mask, SFD_NONBLOCK|SFD_CLOEXEC); - if (m->signal_watch.fd < 0) + m->signal_fd = signalfd(-1, &mask, SFD_NONBLOCK|SFD_CLOEXEC); + if (m->signal_fd < 0) return -errno; - if (epoll_ctl(m->epoll_fd, EPOLL_CTL_ADD, m->signal_watch.fd, &ev) < 0) - return -errno; + r = sd_event_add_io(m->event, m->signal_fd, EPOLLIN, manager_dispatch_signal_fd, m, &m->signal_event_source); + if (r < 0) + return r; if (m->running_as == SYSTEMD_SYSTEM) return enable_special_signals(m); @@ -502,44 +419,39 @@ int manager_new(SystemdRunningAs running_as, bool reexecuting, Manager **_m) { #endif m->running_as = running_as; - m->name_data_slot = m->conn_data_slot = m->subscribed_data_slot = -1; m->exit_code = _MANAGER_EXIT_CODE_INVALID; - m->pin_cgroupfs_fd = -1; - m->idle_pipe[0] = m->idle_pipe[1] = m->idle_pipe[2] = m->idle_pipe[3] = -1; - watch_init(&m->signal_watch); - watch_init(&m->mount_watch); - watch_init(&m->swap_watch); - watch_init(&m->udev_watch); - watch_init(&m->time_change_watch); - watch_init(&m->jobs_in_progress_watch); + m->idle_pipe[0] = m->idle_pipe[1] = m->idle_pipe[2] = m->idle_pipe[3] = -1; - m->epoll_fd = m->dev_autofs_fd = -1; + m->pin_cgroupfs_fd = m->notify_fd = m->signal_fd = m->time_change_fd = m->dev_autofs_fd = m->private_listen_fd = -1; m->current_job_id = 1; /* start as id #1, so that we can leave #0 around as "null-like" value */ r = manager_default_environment(m); if (r < 0) goto fail; - if (!(m->units = hashmap_new(string_hash_func, string_compare_func))) + r = hashmap_ensure_allocated(&m->units, string_hash_func, string_compare_func); + if (r < 0) goto fail; - if (!(m->jobs = hashmap_new(trivial_hash_func, trivial_compare_func))) + r = hashmap_ensure_allocated(&m->jobs, trivial_hash_func, trivial_compare_func); + if (r < 0) goto fail; - if (!(m->watch_pids = hashmap_new(trivial_hash_func, trivial_compare_func))) + r = hashmap_ensure_allocated(&m->cgroup_unit, string_hash_func, string_compare_func); + if (r < 0) goto fail; - m->cgroup_unit = hashmap_new(string_hash_func, string_compare_func); - if (!m->cgroup_unit) + r = hashmap_ensure_allocated(&m->watch_pids, trivial_hash_func, trivial_compare_func); + if (r < 0) goto fail; - m->watch_bus = hashmap_new(string_hash_func, string_compare_func); - if (!m->watch_bus) + r = hashmap_ensure_allocated(&m->watch_bus, string_hash_func, string_compare_func); + if (r < 0) goto fail; - m->epoll_fd = epoll_create1(EPOLL_CLOEXEC); - if (m->epoll_fd < 0) + r = sd_event_default(&m->event); + if (r < 0) goto fail; r = manager_setup_signals(m); @@ -718,11 +630,6 @@ static void manager_clear_jobs_and_units(Manager *m) { m->n_running_jobs = 0; } -static void close_idle_pipe(Manager *m) { - close_pipe(m->idle_pipe); - close_pipe(m->idle_pipe + 2); -} - void manager_free(Manager *m) { UnitType c; int i; @@ -748,16 +655,22 @@ void manager_free(Manager *m) { hashmap_free(m->watch_pids); hashmap_free(m->watch_bus); - if (m->epoll_fd >= 0) - close_nointr_nofail(m->epoll_fd); - if (m->signal_watch.fd >= 0) - close_nointr_nofail(m->signal_watch.fd); - if (m->notify_watch.fd >= 0) - close_nointr_nofail(m->notify_watch.fd); - if (m->time_change_watch.fd >= 0) - close_nointr_nofail(m->time_change_watch.fd); - if (m->jobs_in_progress_watch.fd >= 0) - close_nointr_nofail(m->jobs_in_progress_watch.fd); + sd_event_source_unref(m->signal_event_source); + sd_event_source_unref(m->notify_event_source); + sd_event_source_unref(m->time_change_event_source); + sd_event_source_unref(m->jobs_in_progress_event_source); + sd_event_source_unref(m->idle_pipe_event_source); + + if (m->signal_fd >= 0) + close_nointr_nofail(m->signal_fd); + if (m->notify_fd >= 0) + close_nointr_nofail(m->notify_fd); + if (m->time_change_fd >= 0) + close_nointr_nofail(m->time_change_fd); + + manager_close_idle_pipe(m); + + sd_event_unref(m->event); free(m->notify_socket); @@ -767,8 +680,6 @@ void manager_free(Manager *m) { hashmap_free(m->cgroup_unit); set_free_free(m->unit_path_cache); - close_idle_pipe(m); - free(m->switch_root); free(m->switch_root_init); @@ -929,9 +840,9 @@ int manager_startup(Manager *m, FILE *serialization, FDSet *fds) { m->n_reloading ++; /* First, enumerate what we can from all config files */ - dual_timestamp_get(&m->unitsload_start_timestamp); + dual_timestamp_get(&m->units_load_start_timestamp); r = manager_enumerate(m); - dual_timestamp_get(&m->unitsload_finish_timestamp); + dual_timestamp_get(&m->units_load_finish_timestamp); /* Second, deserialize if there is something to deserialize */ if (serialization) { @@ -968,7 +879,7 @@ int manager_startup(Manager *m, FILE *serialization, FDSet *fds) { return r; } -int manager_add_job(Manager *m, JobType type, Unit *unit, JobMode mode, bool override, DBusError *e, Job **_ret) { +int manager_add_job(Manager *m, JobType type, Unit *unit, JobMode mode, bool override, sd_bus_error *e, Job **_ret) { int r; Transaction *tr; @@ -978,12 +889,12 @@ int manager_add_job(Manager *m, JobType type, Unit *unit, JobMode mode, bool ove assert(mode < _JOB_MODE_MAX); if (mode == JOB_ISOLATE && type != JOB_START) { - dbus_set_error(e, BUS_ERROR_INVALID_JOB_MODE, "Isolate is only valid for start."); + sd_bus_error_setf(e, SD_BUS_ERROR_INVALID_ARGS, "Isolate is only valid for start."); return -EINVAL; } if (mode == JOB_ISOLATE && !unit->allow_isolate) { - dbus_set_error(e, BUS_ERROR_NO_ISOLATION, "Operation refused, unit may not be isolated."); + sd_bus_error_setf(e, BUS_ERROR_NO_ISOLATION, "Operation refused, unit may not be isolated."); return -EPERM; } @@ -1029,7 +940,7 @@ tr_abort: return r; } -int manager_add_job_by_name(Manager *m, JobType type, const char *name, JobMode mode, bool override, DBusError *e, Job **_ret) { +int manager_add_job_by_name(Manager *m, JobType type, const char *name, JobMode mode, bool override, sd_bus_error *e, Job **_ret) { Unit *unit; int r; @@ -1088,7 +999,7 @@ int manager_load_unit_prepare( Manager *m, const char *name, const char *path, - DBusError *e, + sd_bus_error *e, Unit **_ret) { Unit *ret; @@ -1101,20 +1012,16 @@ int manager_load_unit_prepare( /* This will prepare the unit for loading, but not actually * load anything from disk. */ - if (path && !is_path(path)) { - dbus_set_error(e, BUS_ERROR_INVALID_PATH, "Path %s is not absolute.", path); - return -EINVAL; - } + if (path && !is_path(path)) + return sd_bus_error_setf(e, SD_BUS_ERROR_INVALID_ARGS, "Path %s is not absolute.", path); if (!name) name = path_get_file_name(path); t = unit_name_to_type(name); - if (t == _UNIT_TYPE_INVALID || !unit_name_is_valid(name, false)) { - dbus_set_error(e, BUS_ERROR_INVALID_NAME, "Unit name %s is not valid.", name); - return -EINVAL; - } + if (t == _UNIT_TYPE_INVALID || !unit_name_is_valid(name, false)) + return sd_bus_error_setf(e, SD_BUS_ERROR_INVALID_ARGS, "Unit name %s is not valid.", name); ret = manager_get_unit(m, name); if (ret) { @@ -1154,7 +1061,7 @@ int manager_load_unit( Manager *m, const char *name, const char *path, - DBusError *e, + sd_bus_error *e, Unit **_ret) { int r; @@ -1269,16 +1176,26 @@ static unsigned manager_dispatch_dbus_queue(Manager *m) { if (m->send_reloading_done) { m->send_reloading_done = false; - bus_broadcast_reloading(m, false); + bus_manager_send_reloading(m, false); } + if (m->queued_message) + bus_send_queued_message(m); + return n; } -static int manager_process_notify_fd(Manager *m) { +static int manager_dispatch_notify_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + Manager *m = userdata; ssize_t n; assert(m); + assert(m->notify_fd == fd); + + if (revents != EPOLLIN) { + log_warning("Got unexpected poll event for notify fd."); + return 0; + } for (;;) { char buf[4096]; @@ -1302,7 +1219,7 @@ static int manager_process_notify_fd(Manager *m) { Unit *u; _cleanup_strv_free_ char **tags = NULL; - n = recvmsg(m->notify_watch.fd, &msghdr, MSG_DONTWAIT); + n = recvmsg(m->notify_fd, &msghdr, MSG_DONTWAIT); if (n <= 0) { if (n == 0) return -EIO; @@ -1383,7 +1300,7 @@ static int manager_dispatch_sigchld(Manager *m) { * have queued for us. This ensures that the process * still exists in /proc so that we can figure out * which cgroup and hence unit it belongs to. */ - r = manager_process_notify_fd(m); + r = manager_dispatch_notify_fd(m->notify_event_source, m->notify_fd, EPOLLIN, m); if (r < 0) return r; @@ -1425,32 +1342,34 @@ static int manager_dispatch_sigchld(Manager *m) { } static int manager_start_target(Manager *m, const char *name, JobMode mode) { + _cleanup_bus_error_free_ sd_bus_error error = SD_BUS_ERROR_NULL; int r; - DBusError error; - - dbus_error_init(&error); log_debug_unit(name, "Activating special unit %s", name); r = manager_add_job_by_name(m, JOB_START, name, mode, true, &error, NULL); if (r < 0) - log_error_unit(name, - "Failed to enqueue %s job: %s", name, bus_error(&error, r)); - - dbus_error_free(&error); + log_error_unit(name, "Failed to enqueue %s job: %s", name, bus_error_message(&error, r)); return r; } -static int manager_process_signal_fd(Manager *m) { +static int manager_dispatch_signal_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + Manager *m = userdata; ssize_t n; struct signalfd_siginfo sfsi; bool sigchld = false; assert(m); + assert(m->signal_fd == fd); + + if (revents != EPOLLIN) { + log_warning("Got unexpected events from signal file descriptor."); + return 0; + } for (;;) { - n = read(m->signal_watch.fd, &sfsi, sizeof(sfsi)); + n = read(m->signal_fd, &sfsi, sizeof(sfsi)); if (n != sizeof(sfsi)) { if (n >= 0) @@ -1537,11 +1456,12 @@ static int manager_process_signal_fd(Manager *m) { } case SIGUSR2: { - FILE *f; - char *dump = NULL; + _cleanup_free_ char *dump = NULL; + _cleanup_fclose_ FILE *f = NULL; size_t size; - if (!(f = open_memstream(&dump, &size))) { + f = open_memstream(&dump, &size); + if (!f) { log_warning("Failed to allocate memory stream."); break; } @@ -1550,16 +1470,11 @@ static int manager_process_signal_fd(Manager *m) { manager_dump_jobs(m, f, "\t"); if (ferror(f)) { - fclose(f); - free(dump); log_warning("Failed to write status stream"); break; } - fclose(f); log_dump(LOG_INFO, dump); - free(dump); - break; } @@ -1666,141 +1581,54 @@ static int manager_process_signal_fd(Manager *m) { return 0; } -static int process_event(Manager *m, struct epoll_event *ev) { - int r; - Watch *w; +static int manager_dispatch_time_change_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + Manager *m = userdata; + Iterator i; + Unit *u; assert(m); - assert(ev); - - assert_se(w = ev->data.ptr); - - if (w->type == WATCH_INVALID) - return 0; - - switch (w->type) { - - case WATCH_SIGNAL: - - /* An incoming signal? */ - if (ev->events != EPOLLIN) - return -EINVAL; - - if ((r = manager_process_signal_fd(m)) < 0) - return r; - - break; - - case WATCH_NOTIFY: - - /* An incoming daemon notification event? */ - if (ev->events != EPOLLIN) - return -EINVAL; - - if ((r = manager_process_notify_fd(m)) < 0) - return r; - - break; - - case WATCH_FD: - - /* Some fd event, to be dispatched to the units */ - UNIT_VTABLE(w->data.unit)->fd_event(w->data.unit, w->fd, ev->events, w); - break; - - case WATCH_UNIT_TIMER: - case WATCH_JOB_TIMER: { - uint64_t v; - ssize_t k; - - /* Some timer event, to be dispatched to the units */ - k = read(w->fd, &v, sizeof(v)); - if (k != sizeof(v)) { + assert(m->time_change_fd == fd); - if (k < 0 && (errno == EINTR || errno == EAGAIN)) - break; - - log_error("Failed to read timer event counter: %s", k < 0 ? strerror(-k) : "Short read"); - return k < 0 ? -errno : -EIO; - } + log_struct(LOG_INFO, + MESSAGE_ID(SD_MESSAGE_TIME_CHANGE), + "MESSAGE=Time has been changed", + NULL); - if (w->type == WATCH_UNIT_TIMER) - UNIT_VTABLE(w->data.unit)->timer_event(w->data.unit, v, w); - else - job_timer_event(w->data.job, v, w); - break; - } + /* Restart the watch */ + m->time_change_event_source = sd_event_source_unref(m->time_change_event_source); - case WATCH_MOUNT: - /* Some mount table change, intended for the mount subsystem */ - mount_fd_event(m, ev->events); - break; + close_nointr_nofail(m->time_change_fd); + m->time_change_fd = -1; - case WATCH_SWAP: - /* Some swap table change, intended for the swap subsystem */ - swap_fd_event(m, ev->events); - break; + manager_setup_time_change(m); - case WATCH_UDEV: - /* Some notification from udev, intended for the device subsystem */ - device_fd_event(m, ev->events); - break; - - case WATCH_DBUS_WATCH: - bus_watch_event(m, w, ev->events); - break; + HASHMAP_FOREACH(u, m->units, i) + if (UNIT_VTABLE(u)->time_change) + UNIT_VTABLE(u)->time_change(u); - case WATCH_DBUS_TIMEOUT: - bus_timeout_event(m, w, ev->events); - break; + return 0; +} - case WATCH_TIME_CHANGE: { - Unit *u; - Iterator i; - - log_struct(LOG_INFO, - MESSAGE_ID(SD_MESSAGE_TIME_CHANGE), - "MESSAGE=Time has been changed", - NULL); - - /* Restart the watch */ - epoll_ctl(m->epoll_fd, EPOLL_CTL_DEL, m->time_change_watch.fd, - NULL); - close_nointr_nofail(m->time_change_watch.fd); - watch_init(&m->time_change_watch); - manager_setup_time_change(m); - - HASHMAP_FOREACH(u, m->units, i) { - if (UNIT_VTABLE(u)->time_change) - UNIT_VTABLE(u)->time_change(u); - } +static int manager_dispatch_idle_pipe_fd(sd_event_source *source, int fd, uint32_t revents, void *userdata) { + Manager *m = userdata; - break; - } + assert(m); + assert(m->idle_pipe[2] == fd); - case WATCH_JOBS_IN_PROGRESS: { - uint64_t v; + m->no_console_output = m->n_on_console > 0; - /* not interested in the data */ - read(w->fd, &v, sizeof(v)); + m->idle_pipe_event_source = sd_event_source_unref(m->idle_pipe_event_source); + manager_close_idle_pipe(m); - manager_print_jobs_in_progress(m); - break; - } - - case WATCH_IDLE_PIPE: { - m->no_console_output = m->n_on_console > 0; + return 0; +} - manager_unwatch_idle_pipe(m); - close_idle_pipe(m); - break; - } +static int manager_dispatch_jobs_in_progress(sd_event_source *source, usec_t usec, void *userdata) { + Manager *m = userdata; - default: - log_error("event type=%i", w->type); - assert_not_reached("Unknown epoll event type."); - } + assert(m); + manager_print_jobs_in_progress(m); return 0; } @@ -1825,9 +1653,7 @@ int manager_loop(Manager *m) { return r; while (m->exit_code == MANAGER_RUNNING) { - struct epoll_event event; - int n; - int wait_msec = -1; + usec_t wait_usec; if (m->runtime_watchdog > 0 && m->running_as == SYSTEMD_SYSTEM) watchdog_ping(); @@ -1854,44 +1680,28 @@ int manager_loop(Manager *m) { if (manager_dispatch_run_queue(m) > 0) continue; - if (bus_dispatch(m) > 0) - continue; - if (manager_dispatch_dbus_queue(m) > 0) continue; - if (swap_dispatch_reload(m) > 0) - continue; - /* Sleep for half the watchdog time */ if (m->runtime_watchdog > 0 && m->running_as == SYSTEMD_SYSTEM) { - wait_msec = (int) (m->runtime_watchdog / 2 / USEC_PER_MSEC); - if (wait_msec <= 0) - wait_msec = 1; + wait_usec = m->runtime_watchdog / 2; + if (wait_usec <= 0) + wait_usec = 1; } else - wait_msec = -1; - - n = epoll_wait(m->epoll_fd, &event, 1, wait_msec); - if (n < 0) { + wait_usec = (usec_t) -1; - if (errno == EINTR) - continue; - - return -errno; - } else if (n == 0) - continue; - - assert(n == 1); - - r = process_event(m, &event); - if (r < 0) + r = sd_event_run(m->event, wait_usec); + if (r < 0) { + log_error("Failed to run event loop: %s", strerror(-r)); return r; + } } return m->exit_code; } -int manager_load_unit_from_dbus_path(Manager *m, const char *s, DBusError *e, Unit **_u) { +int manager_load_unit_from_dbus_path(Manager *m, const char *s, sd_bus_error *e, Unit **_u) { _cleanup_free_ char *n = NULL; Unit *u; int r; @@ -1914,18 +1724,20 @@ int manager_load_unit_from_dbus_path(Manager *m, const char *s, DBusError *e, Un } int manager_get_job_from_dbus_path(Manager *m, const char *s, Job **_j) { - Job *j; + const char *p; unsigned id; + Job *j; int r; assert(m); assert(s); assert(_j); - if (!startswith(s, "/org/freedesktop/systemd1/job/")) + p = startswith(s, "/org/freedesktop/systemd1/job/"); + if (!p) return -EINVAL; - r = safe_atou(s + 30, &id); + r = safe_atou(p, &id); if (r < 0) return r; @@ -2060,29 +1872,13 @@ void manager_dispatch_bus_name_owner_changed( assert(m); assert(name); - if (!(u = hashmap_get(m->watch_bus, name))) + u = hashmap_get(m->watch_bus, name); + if (!u) return; UNIT_VTABLE(u)->bus_name_owner_change(u, name, old_owner, new_owner); } -void manager_dispatch_bus_query_pid_done( - Manager *m, - const char *name, - pid_t pid) { - - Unit *u; - - assert(m); - assert(name); - assert(pid >= 1); - - if (!(u = hashmap_get(m->watch_bus, name))) - return; - - UNIT_VTABLE(u)->bus_query_pid_done(u, name, pid); -} - int manager_open_serialization(Manager *m, FILE **_f) { char *path = NULL; int fd = -1; @@ -2140,13 +1936,19 @@ int manager_serialize(Manager *m, FILE *f, FDSet *fds, bool switching_root) { fprintf(f, "n-failed-jobs=%u\n", m->n_failed_jobs); dual_timestamp_serialize(f, "firmware-timestamp", &m->firmware_timestamp); - dual_timestamp_serialize(f, "kernel-timestamp", &m->kernel_timestamp); dual_timestamp_serialize(f, "loader-timestamp", &m->loader_timestamp); + dual_timestamp_serialize(f, "kernel-timestamp", &m->kernel_timestamp); dual_timestamp_serialize(f, "initrd-timestamp", &m->initrd_timestamp); if (!in_initrd()) { dual_timestamp_serialize(f, "userspace-timestamp", &m->userspace_timestamp); dual_timestamp_serialize(f, "finish-timestamp", &m->finish_timestamp); + dual_timestamp_serialize(f, "security-start-timestamp", &m->security_start_timestamp); + dual_timestamp_serialize(f, "security-finish-timestamp", &m->security_finish_timestamp); + dual_timestamp_serialize(f, "generators-start-timestamp", &m->generators_start_timestamp); + dual_timestamp_serialize(f, "generators-finish-timestamp", &m->generators_finish_timestamp); + dual_timestamp_serialize(f, "units-load-start-timestamp", &m->units_load_start_timestamp); + dual_timestamp_serialize(f, "units-load-finish-timestamp", &m->units_load_finish_timestamp); } if (!switching_root) { @@ -2229,6 +2031,7 @@ int manager_deserialize(Manager *m, FILE *f, FDSet *fds) { log_debug("Failed to parse current job id value %s", l+15); else m->current_job_id = MAX(m->current_job_id, id); + } else if (startswith(l, "n-installed-jobs=")) { uint32_t n; @@ -2236,6 +2039,7 @@ int manager_deserialize(Manager *m, FILE *f, FDSet *fds) { log_debug("Failed to parse installed jobs counter %s", l+17); else m->n_installed_jobs += n; + } else if (startswith(l, "n-failed-jobs=")) { uint32_t n; @@ -2243,6 +2047,7 @@ int manager_deserialize(Manager *m, FILE *f, FDSet *fds) { log_debug("Failed to parse failed jobs counter %s", l+14); else m->n_failed_jobs += n; + } else if (startswith(l, "taint-usr=")) { int b; @@ -2250,6 +2055,7 @@ int manager_deserialize(Manager *m, FILE *f, FDSet *fds) { log_debug("Failed to parse taint /usr flag %s", l+10); else m->taint_usr = m->taint_usr || b; + } else if (startswith(l, "firmware-timestamp=")) dual_timestamp_deserialize(l+19, &m->firmware_timestamp); else if (startswith(l, "loader-timestamp=")) @@ -2262,6 +2068,18 @@ int manager_deserialize(Manager *m, FILE *f, FDSet *fds) { dual_timestamp_deserialize(l+20, &m->userspace_timestamp); else if (startswith(l, "finish-timestamp=")) dual_timestamp_deserialize(l+17, &m->finish_timestamp); + else if (startswith(l, "security-start-timestamp=")) + dual_timestamp_deserialize(l+25, &m->security_start_timestamp); + else if (startswith(l, "security-finish-timestamp=")) + dual_timestamp_deserialize(l+26, &m->security_finish_timestamp); + else if (startswith(l, "generators-start-timestamp=")) + dual_timestamp_deserialize(l+27, &m->generators_start_timestamp); + else if (startswith(l, "generators-finish-timestamp=")) + dual_timestamp_deserialize(l+28, &m->generators_finish_timestamp); + else if (startswith(l, "units-load-start-timestamp=")) + dual_timestamp_deserialize(l+27, &m->units_load_start_timestamp); + else if (startswith(l, "units-load-finish-timestamp=")) + dual_timestamp_deserialize(l+28, &m->units_load_finish_timestamp); else if (startswith(l, "env=")) { _cleanup_free_ char *uce = NULL; char **e; @@ -2333,7 +2151,7 @@ int manager_reload(Manager *m) { return r; m->n_reloading ++; - bus_broadcast_reloading(m, true); + bus_manager_send_reloading(m, true); fds = fdset_new(); if (!fds) { @@ -2450,16 +2268,16 @@ void manager_check_finished(Manager *m) { assert(m); if (m->n_running_jobs == 0) - manager_unwatch_jobs_in_progress(m); + m->jobs_in_progress_event_source = sd_event_source_unref(m->jobs_in_progress_event_source); - if (hashmap_size(m->jobs) > 0) { - manager_jobs_in_progress_mod_timer(m); + if (hashmap_size(m->jobs) > 0 && m->jobs_in_progress_event_source) { + sd_event_source_set_time(m->jobs_in_progress_event_source, JOBS_IN_PROGRESS_PERIOD_SEC); return; } /* Notify Type=idle units that we are done now */ - manager_unwatch_idle_pipe(m); - close_idle_pipe(m); + m->idle_pipe_event_source = sd_event_source_unref(m->idle_pipe_event_source); + manager_close_idle_pipe(m); /* Turn off confirm spawn now */ m->confirm_spawn = false; @@ -2526,7 +2344,7 @@ void manager_check_finished(Manager *m) { NULL); } - bus_broadcast_finished(m, firmware_usec, loader_usec, kernel_usec, initrd_usec, userspace_usec, total_usec); + bus_manager_send_finished(m, firmware_usec, loader_usec, kernel_usec, initrd_usec, userspace_usec, total_usec); sd_notifyf(false, "READY=1\nSTATUS=Startup finished in %s.", @@ -2590,7 +2408,7 @@ static void trim_generator_dir(Manager *m, char **generator) { } void manager_run_generators(Manager *m) { - DIR *d = NULL; + _cleanup_closedir_ DIR *d = NULL; const char *generator_path; const char *argv[5]; int r; @@ -2626,17 +2444,13 @@ void manager_run_generators(Manager *m) { argv[3] = m->generator_unit_path_late; argv[4] = NULL; - RUN_WITH_UMASK(0022) { + RUN_WITH_UMASK(0022) execute_directory(generator_path, d, (char**) argv); - } +finish: trim_generator_dir(m, &m->generator_unit_path); trim_generator_dir(m, &m->generator_unit_path_early); trim_generator_dir(m, &m->generator_unit_path_late); - -finish: - if (d) - closedir(d); } static void remove_generator_dir(Manager *m, char **generator) { @@ -2661,17 +2475,36 @@ void manager_undo_generators(Manager *m) { remove_generator_dir(m, &m->generator_unit_path_late); } -int manager_environment_add(Manager *m, char **environment) { - char **e = NULL; +int manager_environment_add(Manager *m, char **minus, char **plus) { + char **a = NULL, **b = NULL, **l; assert(m); - e = strv_env_merge(2, m->environment, environment); - if (!e) - return -ENOMEM; + l = m->environment; - strv_free(m->environment); - m->environment = e; + if (!strv_isempty(minus)) { + a = strv_env_delete(l, 1, minus); + if (!a) + return -ENOMEM; + + l = a; + } + + if (!strv_isempty(plus)) { + b = strv_env_merge(2, l, plus); + if (!b) + return -ENOMEM; + l = b; + } + + if (m->environment != l) + strv_free(m->environment); + if (a != l) + strv_free(a); + if (b != l) + strv_free(b); + + m->environment = l; return 0; } @@ -2802,10 +2635,3 @@ Set *manager_get_units_requiring_mounts_for(Manager *m, const char *path) { return hashmap_get(m->units_requiring_mounts_for, streq(p, "/") ? "" : p); } - -void watch_init(Watch *w) { - assert(w); - - w->type = WATCH_INVALID; - w->fd = -1; -} |