blob: 8552ffb70418b1d8a0714a4149e3c47125bb6c20 [file] [log] [blame] [raw]
/* SPDX-License-Identifier: LGPL-2.1+ */
#include <errno.h>
#include "sd-id128.h"
#include "sd-messages.h"
#include "alloc-util.h"
#include "async.h"
#include "dbus-job.h"
#include "dbus.h"
#include "escape.h"
#include "fileio.h"
#include "job.h"
#include "log.h"
#include "macro.h"
#include "parse-util.h"
#include "set.h"
#include "special.h"
#include "stdio-util.h"
#include "string-table.h"
#include "string-util.h"
#include "strv.h"
#include "terminal-util.h"
#include "unit.h"
#include "virt.h"
Job* job_new_raw(Unit *unit) {
Job *j;
/* used for deserialization */
assert(unit);
j = new0(Job, 1);
if (!j)
return NULL;
j->manager = unit->manager;
j->unit = unit;
j->type = _JOB_TYPE_INVALID;
j->reloaded = false;
return j;
}
Job* job_new(Unit *unit, JobType type) {
Job *j;
assert(type < _JOB_TYPE_MAX);
j = job_new_raw(unit);
if (!j)
return NULL;
j->id = j->manager->current_job_id++;
j->type = type;
/* We don't link it here, that's what job_dependency() is for */
return j;
}
void job_unlink(Job *j) {
assert(j);
assert(!j->installed);
assert(!j->transaction_prev);
assert(!j->transaction_next);
assert(!j->subject_list);
assert(!j->object_list);
if (j->in_run_queue) {
LIST_REMOVE(run_queue, j->manager->run_queue, j);
j->in_run_queue = false;
}
if (j->in_dbus_queue) {
LIST_REMOVE(dbus_queue, j->manager->dbus_job_queue, j);
j->in_dbus_queue = false;
}
if (j->in_gc_queue) {
LIST_REMOVE(gc_queue, j->manager->gc_job_queue, j);
j->in_gc_queue = false;
}
j->timer_event_source = sd_event_source_unref(j->timer_event_source);
}
void job_free(Job *j) {
assert(j);
assert(!j->installed);
assert(!j->transaction_prev);
assert(!j->transaction_next);
assert(!j->subject_list);
assert(!j->object_list);
job_unlink(j);
sd_bus_track_unref(j->bus_track);
strv_free(j->deserialized_clients);
free(j);
}
static void job_set_state(Job *j, JobState state) {
assert(j);
assert(state >= 0);
assert(state < _JOB_STATE_MAX);
if (j->state == state)
return;
j->state = state;
if (!j->installed)
return;
if (j->state == JOB_RUNNING)
j->unit->manager->n_running_jobs++;
else {
assert(j->state == JOB_WAITING);
assert(j->unit->manager->n_running_jobs > 0);
j->unit->manager->n_running_jobs--;
if (j->unit->manager->n_running_jobs <= 0)
j->unit->manager->jobs_in_progress_event_source = sd_event_source_unref(j->unit->manager->jobs_in_progress_event_source);
}
}
void job_uninstall(Job *j) {
Job **pj;
assert(j->installed);
job_set_state(j, JOB_WAITING);
pj = (j->type == JOB_NOP) ? &j->unit->nop_job : &j->unit->job;
assert(*pj == j);
/* Detach from next 'bigger' objects */
/* daemon-reload should be transparent to job observers */
if (!MANAGER_IS_RELOADING(j->manager))
bus_job_send_removed_signal(j);
*pj = NULL;
unit_add_to_gc_queue(j->unit);
hashmap_remove(j->manager->jobs, UINT32_TO_PTR(j->id));
j->installed = false;
}
static bool job_type_allows_late_merge(JobType t) {
/* Tells whether it is OK to merge a job of type 't' with an already
* running job.
* Reloads cannot be merged this way. Think of the sequence:
* 1. Reload of a daemon is in progress; the daemon has already loaded
* its config file, but hasn't completed the reload operation yet.
* 2. Edit foo's config file.
* 3. Trigger another reload to have the daemon use the new config.
* Should the second reload job be merged into the first one, the daemon
* would not know about the new config.
* JOB_RESTART jobs on the other hand can be merged, because they get
* patched into JOB_START after stopping the unit. So if we see a
* JOB_RESTART running, it means the unit hasn't stopped yet and at
* this time the merge is still allowed. */
return t != JOB_RELOAD;
}
static void job_merge_into_installed(Job *j, Job *other) {
assert(j->installed);
assert(j->unit == other->unit);
if (j->type != JOB_NOP)
job_type_merge_and_collapse(&j->type, other->type, j->unit);
else
assert(other->type == JOB_NOP);
j->irreversible = j->irreversible || other->irreversible;
j->ignore_order = j->ignore_order || other->ignore_order;
}
Job* job_install(Job *j) {
Job **pj;
Job *uj;
assert(!j->installed);
assert(j->type < _JOB_TYPE_MAX_IN_TRANSACTION);
assert(j->state == JOB_WAITING);
pj = (j->type == JOB_NOP) ? &j->unit->nop_job : &j->unit->job;
uj = *pj;
if (uj) {
if (job_type_is_conflicting(uj->type, j->type))
job_finish_and_invalidate(uj, JOB_CANCELED, false, false);
else {
/* not conflicting, i.e. mergeable */
if (uj->state == JOB_WAITING ||
(job_type_allows_late_merge(j->type) && job_type_is_superset(uj->type, j->type))) {
job_merge_into_installed(uj, j);
log_unit_debug(uj->unit,
"Merged into installed job %s/%s as %u",
uj->unit->id, job_type_to_string(uj->type), (unsigned) uj->id);
return uj;
} else {
/* already running and not safe to merge into */
/* Patch uj to become a merged job and re-run it. */
/* XXX It should be safer to queue j to run after uj finishes, but it is
* not currently possible to have more than one installed job per unit. */
job_merge_into_installed(uj, j);
log_unit_debug(uj->unit,
"Merged into running job, re-running: %s/%s as %u",
uj->unit->id, job_type_to_string(uj->type), (unsigned) uj->id);
job_set_state(uj, JOB_WAITING);
return uj;
}
}
}
/* Install the job */
*pj = j;
j->installed = true;
j->manager->n_installed_jobs++;
log_unit_debug(j->unit,
"Installed new job %s/%s as %u",
j->unit->id, job_type_to_string(j->type), (unsigned) j->id);
job_add_to_gc_queue(j);
return j;
}
int job_install_deserialized(Job *j) {
Job **pj;
assert(!j->installed);
if (j->type < 0 || j->type >= _JOB_TYPE_MAX_IN_TRANSACTION) {
log_debug("Invalid job type %s in deserialization.", strna(job_type_to_string(j->type)));
return -EINVAL;
}
pj = (j->type == JOB_NOP) ? &j->unit->nop_job : &j->unit->job;
if (*pj) {
log_unit_debug(j->unit, "Unit already has a job installed. Not installing deserialized job.");
return -EEXIST;
}
*pj = j;
j->installed = true;
j->reloaded = true;
if (j->state == JOB_RUNNING)
j->unit->manager->n_running_jobs++;
log_unit_debug(j->unit,
"Reinstalled deserialized job %s/%s as %u",
j->unit->id, job_type_to_string(j->type), (unsigned) j->id);
return 0;
}
JobDependency* job_dependency_new(Job *subject, Job *object, bool matters, bool conflicts) {
JobDependency *l;
assert(object);
/* Adds a new job link, which encodes that the 'subject' job
* needs the 'object' job in some way. If 'subject' is NULL
* this means the 'anchor' job (i.e. the one the user
* explicitly asked for) is the requester. */
l = new0(JobDependency, 1);
if (!l)
return NULL;
l->subject = subject;
l->object = object;
l->matters = matters;
l->conflicts = conflicts;
if (subject)
LIST_PREPEND(subject, subject->subject_list, l);
LIST_PREPEND(object, object->object_list, l);
return l;
}
void job_dependency_free(JobDependency *l) {
assert(l);
if (l->subject)
LIST_REMOVE(subject, l->subject->subject_list, l);
LIST_REMOVE(object, l->object->object_list, l);
free(l);
}
void job_dump(Job *j, FILE*f, const char *prefix) {
assert(j);
assert(f);
prefix = strempty(prefix);
fprintf(f,
"%s-> Job %u:\n"
"%s\tAction: %s -> %s\n"
"%s\tState: %s\n"
"%s\tIrreversible: %s\n"
"%s\tMay GC: %s\n",
prefix, j->id,
prefix, j->unit->id, job_type_to_string(j->type),
prefix, job_state_to_string(j->state),
prefix, yes_no(j->irreversible),
prefix, yes_no(job_may_gc(j)));
}
/*
* Merging is commutative, so imagine the matrix as symmetric. We store only
* its lower triangle to avoid duplication. We don't store the main diagonal,
* because A merged with A is simply A.
*
* If the resulting type is collapsed immediately afterwards (to get rid of
* the JOB_RELOAD_OR_START, which lies outside the lookup function's domain),
* the following properties hold:
*
* Merging is associative! A merged with B, and then merged with C is the same
* as A merged with the result of B merged with C.
*
* Mergeability is transitive! If A can be merged with B and B with C then
* A also with C.
*
* Also, if A merged with B cannot be merged with C, then either A or B cannot
* be merged with C either.
*/
static const JobType job_merging_table[] = {
/* What \ With * JOB_START JOB_VERIFY_ACTIVE JOB_STOP JOB_RELOAD */
/*********************************************************************************/
/*JOB_START */
/*JOB_VERIFY_ACTIVE */ JOB_START,
/*JOB_STOP */ -1, -1,
/*JOB_RELOAD */ JOB_RELOAD_OR_START, JOB_RELOAD, -1,
/*JOB_RESTART */ JOB_RESTART, JOB_RESTART, -1, JOB_RESTART,
};
JobType job_type_lookup_merge(JobType a, JobType b) {
assert_cc(ELEMENTSOF(job_merging_table) == _JOB_TYPE_MAX_MERGING * (_JOB_TYPE_MAX_MERGING - 1) / 2);
assert(a >= 0 && a < _JOB_TYPE_MAX_MERGING);
assert(b >= 0 && b < _JOB_TYPE_MAX_MERGING);
if (a == b)
return a;
if (a < b) {
JobType tmp = a;
a = b;
b = tmp;
}
return job_merging_table[(a - 1) * a / 2 + b];
}
bool job_type_is_redundant(JobType a, UnitActiveState b) {
switch (a) {
case JOB_START:
return IN_SET(b, UNIT_ACTIVE, UNIT_RELOADING);
case JOB_STOP:
return IN_SET(b, UNIT_INACTIVE, UNIT_FAILED);
case JOB_VERIFY_ACTIVE:
return IN_SET(b, UNIT_ACTIVE, UNIT_RELOADING);
case JOB_RELOAD:
return
b == UNIT_RELOADING;
case JOB_RESTART:
return
b == UNIT_ACTIVATING;
case JOB_NOP:
return true;
default:
assert_not_reached("Invalid job type");
}
}
JobType job_type_collapse(JobType t, Unit *u) {
UnitActiveState s;
switch (t) {
case JOB_TRY_RESTART:
s = unit_active_state(u);
if (UNIT_IS_INACTIVE_OR_DEACTIVATING(s))
return JOB_NOP;
return JOB_RESTART;
case JOB_TRY_RELOAD:
s = unit_active_state(u);
if (UNIT_IS_INACTIVE_OR_DEACTIVATING(s))
return JOB_NOP;
return JOB_RELOAD;
case JOB_RELOAD_OR_START:
s = unit_active_state(u);
if (UNIT_IS_INACTIVE_OR_DEACTIVATING(s))
return JOB_START;
return JOB_RELOAD;
default:
return t;
}
}
int job_type_merge_and_collapse(JobType *a, JobType b, Unit *u) {
JobType t;
t = job_type_lookup_merge(*a, b);
if (t < 0)
return -EEXIST;
*a = job_type_collapse(t, u);
return 0;
}
static bool job_is_runnable(Job *j) {
Iterator i;
Unit *other;
void *v;
assert(j);
assert(j->installed);
/* Checks whether there is any job running for the units this
* job needs to be running after (in the case of a 'positive'
* job type) or before (in the case of a 'negative' job
* type. */
/* Note that unit types have a say in what is runnable,
* too. For example, if they return -EAGAIN from
* unit_start() they can indicate they are not
* runnable yet. */
/* First check if there is an override */
if (j->ignore_order)
return true;
if (j->type == JOB_NOP)
return true;
if (IN_SET(j->type, JOB_START, JOB_VERIFY_ACTIVE, JOB_RELOAD)) {
/* Immediate result is that the job is or might be
* started. In this case let's wait for the
* dependencies, regardless whether they are
* starting or stopping something. */
HASHMAP_FOREACH_KEY(v, other, j->unit->dependencies[UNIT_AFTER], i)
if (other->job)
return false;
}
/* Also, if something else is being stopped and we should
* change state after it, then let's wait. */
HASHMAP_FOREACH_KEY(v, other, j->unit->dependencies[UNIT_BEFORE], i)
if (other->job &&
IN_SET(other->job->type, JOB_STOP, JOB_RESTART))
return false;
/* This means that for a service a and a service b where b
* shall be started after a:
*
* start a + start b → 1st step start a, 2nd step start b
* start a + stop b → 1st step stop b, 2nd step start a
* stop a + start b → 1st step stop a, 2nd step start b
* stop a + stop b → 1st step stop b, 2nd step stop a
*
* This has the side effect that restarts are properly
* synchronized too. */
return true;
}
static void job_change_type(Job *j, JobType newtype) {
assert(j);
log_unit_debug(j->unit,
"Converting job %s/%s -> %s/%s",
j->unit->id, job_type_to_string(j->type),
j->unit->id, job_type_to_string(newtype));
j->type = newtype;
}
static int job_perform_on_unit(Job **j) {
uint32_t id;
Manager *m;
JobType t;
Unit *u;
int r;
/* While we execute this operation the job might go away (for
* example: because it finishes immediately or is replaced by
* a new, conflicting job.) To make sure we don't access a
* freed job later on we store the id here, so that we can
* verify the job is still valid. */
assert(j);
assert(*j);
m = (*j)->manager;
u = (*j)->unit;
t = (*j)->type;
id = (*j)->id;
switch (t) {
case JOB_START:
r = unit_start(u);
break;
case JOB_RESTART:
t = JOB_STOP;
_fallthrough_;
case JOB_STOP:
r = unit_stop(u);
break;
case JOB_RELOAD:
r = unit_reload(u);
break;
default:
assert_not_reached("Invalid job type");
}
/* Log if the job still exists and the start/stop/reload function
* actually did something. */
*j = manager_get_job(m, id);
if (*j && r > 0)
unit_status_emit_starting_stopping_reloading(u, t);
return r;
}
int job_run_and_invalidate(Job *j) {
int r;
assert(j);
assert(j->installed);
assert(j->type < _JOB_TYPE_MAX_IN_TRANSACTION);
assert(j->in_run_queue);
LIST_REMOVE(run_queue, j->manager->run_queue, j);
j->in_run_queue = false;
if (j->state != JOB_WAITING)
return 0;
if (!job_is_runnable(j))
return -EAGAIN;
job_start_timer(j, true);
job_set_state(j, JOB_RUNNING);
job_add_to_dbus_queue(j);
switch (j->type) {
case JOB_VERIFY_ACTIVE: {
UnitActiveState t = unit_active_state(j->unit);
if (UNIT_IS_ACTIVE_OR_RELOADING(t))
r = -EALREADY;
else if (t == UNIT_ACTIVATING)
r = -EAGAIN;
else
r = -EBADR;
break;
}
case JOB_START:
case JOB_STOP:
case JOB_RESTART:
r = job_perform_on_unit(&j);
/* If the unit type does not support starting/stopping,
* then simply wait. */
if (r == -EBADR)
r = 0;
break;
case JOB_RELOAD:
r = job_perform_on_unit(&j);
break;
case JOB_NOP:
r = -EALREADY;
break;
default:
assert_not_reached("Unknown job type");
}
if (j) {
if (r == -EALREADY)
r = job_finish_and_invalidate(j, JOB_DONE, true, true);
else if (r == -EBADR)
r = job_finish_and_invalidate(j, JOB_SKIPPED, true, false);
else if (r == -ENOEXEC)
r = job_finish_and_invalidate(j, JOB_INVALID, true, false);
else if (r == -EPROTO)
r = job_finish_and_invalidate(j, JOB_ASSERT, true, false);
else if (r == -EOPNOTSUPP)
r = job_finish_and_invalidate(j, JOB_UNSUPPORTED, true, false);
else if (r == -ENOLINK)
r = job_finish_and_invalidate(j, JOB_DEPENDENCY, true, false);
else if (r == -ESTALE)
r = job_finish_and_invalidate(j, JOB_ONCE, true, false);
else if (r == -EAGAIN)
job_set_state(j, JOB_WAITING);
else if (r < 0)
r = job_finish_and_invalidate(j, JOB_FAILED, true, false);
}
return r;
}
_pure_ static const char *job_get_status_message_format(Unit *u, JobType t, JobResult result) {
static const char *const generic_finished_start_job[_JOB_RESULT_MAX] = {
[JOB_DONE] = "Started %s.",
[JOB_TIMEOUT] = "Timed out starting %s.",
[JOB_FAILED] = "Failed to start %s.",
[JOB_DEPENDENCY] = "Dependency failed for %s.",
[JOB_ASSERT] = "Assertion failed for %s.",
[JOB_UNSUPPORTED] = "Starting of %s not supported.",
[JOB_COLLECTED] = "Unnecessary job for %s was removed.",
[JOB_ONCE] = "Unit %s has been started before and cannot be started again."
};
static const char *const generic_finished_stop_job[_JOB_RESULT_MAX] = {
[JOB_DONE] = "Stopped %s.",
[JOB_FAILED] = "Stopped (with error) %s.",
[JOB_TIMEOUT] = "Timed out stopping %s.",
};
static const char *const generic_finished_reload_job[_JOB_RESULT_MAX] = {
[JOB_DONE] = "Reloaded %s.",
[JOB_FAILED] = "Reload failed for %s.",
[JOB_TIMEOUT] = "Timed out reloading %s.",
};
/* When verify-active detects the unit is inactive, report it.
* Most likely a DEPEND warning from a requisiting unit will
* occur next and it's nice to see what was requisited. */
static const char *const generic_finished_verify_active_job[_JOB_RESULT_MAX] = {
[JOB_SKIPPED] = "%s is not active.",
};
const UnitStatusMessageFormats *format_table;
const char *format;
assert(u);
assert(t >= 0);
assert(t < _JOB_TYPE_MAX);
if (IN_SET(t, JOB_START, JOB_STOP, JOB_RESTART)) {
format_table = &UNIT_VTABLE(u)->status_message_formats;
if (format_table) {
format = t == JOB_START ? format_table->finished_start_job[result] :
format_table->finished_stop_job[result];
if (format)
return format;
}
}
/* Return generic strings */
if (t == JOB_START)
return generic_finished_start_job[result];
else if (IN_SET(t, JOB_STOP, JOB_RESTART))
return generic_finished_stop_job[result];
else if (t == JOB_RELOAD)
return generic_finished_reload_job[result];
else if (t == JOB_VERIFY_ACTIVE)
return generic_finished_verify_active_job[result];
return NULL;
}
static const struct {
const char *color, *word;
} job_print_status_messages [_JOB_RESULT_MAX] = {
[JOB_DONE] = { ANSI_OK_COLOR, " OK " },
[JOB_TIMEOUT] = { ANSI_HIGHLIGHT_RED, " TIME " },
[JOB_FAILED] = { ANSI_HIGHLIGHT_RED, "FAILED" },
[JOB_DEPENDENCY] = { ANSI_HIGHLIGHT_YELLOW, "DEPEND" },
[JOB_SKIPPED] = { ANSI_HIGHLIGHT, " INFO " },
[JOB_ASSERT] = { ANSI_HIGHLIGHT_YELLOW, "ASSERT" },
[JOB_UNSUPPORTED] = { ANSI_HIGHLIGHT_YELLOW, "UNSUPP" },
/* JOB_COLLECTED */
[JOB_ONCE] = { ANSI_HIGHLIGHT_RED, " ONCE " },
};
static void job_print_status_message(Unit *u, JobType t, JobResult result) {
const char *format;
const char *status;
assert(u);
assert(t >= 0);
assert(t < _JOB_TYPE_MAX);
/* Reload status messages have traditionally not been printed to console. */
if (t == JOB_RELOAD)
return;
if (!job_print_status_messages[result].word)
return;
format = job_get_status_message_format(u, t, result);
if (!format)
return;
if (log_get_show_color())
status = strjoina(job_print_status_messages[result].color,
job_print_status_messages[result].word,
ANSI_NORMAL);
else
status = job_print_status_messages[result].word;
if (result != JOB_DONE)
manager_flip_auto_status(u->manager, true);
DISABLE_WARNING_FORMAT_NONLITERAL;
unit_status_printf(u, status, format);
REENABLE_WARNING;
if (t == JOB_START && result == JOB_FAILED) {
_cleanup_free_ char *quoted;
quoted = shell_maybe_quote(u->id, ESCAPE_BACKSLASH);
manager_status_printf(u->manager, STATUS_TYPE_NORMAL, NULL, "See 'systemctl status %s' for details.", strna(quoted));
}
}
static void job_log_status_message(Unit *u, JobType t, JobResult result) {
const char *format, *mid;
char buf[LINE_MAX];
static const int job_result_log_level[_JOB_RESULT_MAX] = {
[JOB_DONE] = LOG_INFO,
[JOB_CANCELED] = LOG_INFO,
[JOB_TIMEOUT] = LOG_ERR,
[JOB_FAILED] = LOG_ERR,
[JOB_DEPENDENCY] = LOG_WARNING,
[JOB_SKIPPED] = LOG_NOTICE,
[JOB_INVALID] = LOG_INFO,
[JOB_ASSERT] = LOG_WARNING,
[JOB_UNSUPPORTED] = LOG_WARNING,
[JOB_COLLECTED] = LOG_INFO,
[JOB_ONCE] = LOG_ERR,
};
assert(u);
assert(t >= 0);
assert(t < _JOB_TYPE_MAX);
/* Skip printing if output goes to the console, and job_print_status_message()
will actually print something to the console. */
if (log_on_console() && job_print_status_messages[result].word)
return;
format = job_get_status_message_format(u, t, result);
if (!format)
return;
/* The description might be longer than the buffer, but that's OK,
* we'll just truncate it here. Note that we use snprintf() rather than
* xsprintf() on purpose here: we are fine with truncation and don't
* consider that an error. */
DISABLE_WARNING_FORMAT_NONLITERAL;
(void) snprintf(buf, sizeof(buf), format, unit_description(u));
REENABLE_WARNING;
switch (t) {
case JOB_START:
if (result == JOB_DONE)
mid = "MESSAGE_ID=" SD_MESSAGE_UNIT_STARTED_STR;
else
mid = "MESSAGE_ID=" SD_MESSAGE_UNIT_FAILED_STR;
break;
case JOB_RELOAD:
mid = "MESSAGE_ID=" SD_MESSAGE_UNIT_RELOADED_STR;
break;
case JOB_STOP:
case JOB_RESTART:
mid = "MESSAGE_ID=" SD_MESSAGE_UNIT_STOPPED_STR;
break;
default:
log_struct(job_result_log_level[result],
LOG_MESSAGE("%s", buf),
"JOB_TYPE=%s", job_type_to_string(t),
"JOB_RESULT=%s", job_result_to_string(result),
LOG_UNIT_ID(u),
LOG_UNIT_INVOCATION_ID(u));
return;
}
log_struct(job_result_log_level[result],
LOG_MESSAGE("%s", buf),
"JOB_TYPE=%s", job_type_to_string(t),
"JOB_RESULT=%s", job_result_to_string(result),
LOG_UNIT_ID(u),
LOG_UNIT_INVOCATION_ID(u),
mid);
}
static void job_emit_status_message(Unit *u, JobType t, JobResult result) {
assert(u);
/* No message if the job did not actually do anything due to failed condition. */
if (t == JOB_START && result == JOB_DONE && !u->condition_result)
return;
job_log_status_message(u, t, result);
job_print_status_message(u, t, result);
}
static void job_fail_dependencies(Unit *u, UnitDependency d) {
Unit *other;
Iterator i;
void *v;
assert(u);
HASHMAP_FOREACH_KEY(v, other, u->dependencies[d], i) {
Job *j = other->job;
if (!j)
continue;
if (!IN_SET(j->type, JOB_START, JOB_VERIFY_ACTIVE))
continue;
job_finish_and_invalidate(j, JOB_DEPENDENCY, true, false);
}
}
static int job_save_pending_finished_job(Job *j) {
int r;
assert(j);
r = set_ensure_allocated(&j->manager->pending_finished_jobs, NULL);
if (r < 0)
return r;
job_unlink(j);
return set_put(j->manager->pending_finished_jobs, j);
}
int job_finish_and_invalidate(Job *j, JobResult result, bool recursive, bool already) {
Unit *u;
Unit *other;
JobType t;
Iterator i;
void *v;
assert(j);
assert(j->installed);
assert(j->type < _JOB_TYPE_MAX_IN_TRANSACTION);
u = j->unit;
t = j->type;
j->result = result;
log_unit_debug(u, "Job %s/%s finished, result=%s", u->id, job_type_to_string(t), job_result_to_string(result));
/* If this job did nothing to respective unit we don't log the status message */
if (!already)
job_emit_status_message(u, t, result);
/* Patch restart jobs so that they become normal start jobs */
if (result == JOB_DONE && t == JOB_RESTART) {
job_change_type(j, JOB_START);
job_set_state(j, JOB_WAITING);
job_add_to_dbus_queue(j);
job_add_to_run_queue(j);
job_add_to_gc_queue(j);
goto finish;
}
if (IN_SET(result, JOB_FAILED, JOB_INVALID))
j->manager->n_failed_jobs++;
job_uninstall(j);
/* Keep jobs started before the reload to send singal later, free all others */
if (!MANAGER_IS_RELOADING(j->manager) ||
!j->reloaded ||
job_save_pending_finished_job(j) < 0)
job_free(j);
/* Fail depending jobs on failure */
if (result != JOB_DONE && recursive) {
if (IN_SET(t, JOB_START, JOB_VERIFY_ACTIVE)) {
job_fail_dependencies(u, UNIT_REQUIRED_BY);
job_fail_dependencies(u, UNIT_REQUISITE_OF);
job_fail_dependencies(u, UNIT_BOUND_BY);
} else if (t == JOB_STOP)
job_fail_dependencies(u, UNIT_CONFLICTED_BY);
}
/* Trigger OnFailure dependencies that are not generated by
* the unit itself. We don't treat JOB_CANCELED as failure in
* this context. And JOB_FAILURE is already handled by the
* unit itself. */
if (IN_SET(result, JOB_TIMEOUT, JOB_DEPENDENCY)) {
log_struct(LOG_NOTICE,
"JOB_TYPE=%s", job_type_to_string(t),
"JOB_RESULT=%s", job_result_to_string(result),
LOG_UNIT_ID(u),
LOG_UNIT_MESSAGE(u, "Job %s/%s failed with result '%s'.",
u->id,
job_type_to_string(t),
job_result_to_string(result)));
unit_start_on_failure(u);
}
unit_trigger_notify(u);
finish:
/* Try to start the next jobs that can be started */
HASHMAP_FOREACH_KEY(v, other, u->dependencies[UNIT_AFTER], i)
if (other->job) {
job_add_to_run_queue(other->job);
job_add_to_gc_queue(other->job);
}
HASHMAP_FOREACH_KEY(v, other, u->dependencies[UNIT_BEFORE], i)
if (other->job) {
job_add_to_run_queue(other->job);
job_add_to_gc_queue(other->job);
}
manager_check_finished(u->manager);
return 0;
}
static int job_dispatch_timer(sd_event_source *s, uint64_t monotonic, void *userdata) {
Job *j = userdata;
Unit *u;
assert(j);
assert(s == j->timer_event_source);
log_unit_warning(j->unit, "Job %s/%s timed out.", j->unit->id, job_type_to_string(j->type));
u = j->unit;
job_finish_and_invalidate(j, JOB_TIMEOUT, true, false);
emergency_action(u->manager, u->job_timeout_action, u->job_timeout_reboot_arg, "job timed out");
return 0;
}
int job_start_timer(Job *j, bool job_running) {
int r;
usec_t timeout_time, old_timeout_time;
if (job_running) {
j->begin_running_usec = now(CLOCK_MONOTONIC);
if (j->unit->job_running_timeout == USEC_INFINITY)
return 0;
timeout_time = usec_add(j->begin_running_usec, j->unit->job_running_timeout);
if (j->timer_event_source) {
/* Update only if JobRunningTimeoutSec= results in earlier timeout */
r = sd_event_source_get_time(j->timer_event_source, &old_timeout_time);
if (r < 0)
return r;
if (old_timeout_time <= timeout_time)
return 0;
return sd_event_source_set_time(j->timer_event_source, timeout_time);
}
} else {
if (j->timer_event_source)
return 0;
j->begin_usec = now(CLOCK_MONOTONIC);
if (j->unit->job_timeout == USEC_INFINITY)
return 0;
timeout_time = usec_add(j->begin_usec, j->unit->job_timeout);
}
r = sd_event_add_time(
j->manager->event,
&j->timer_event_source,
CLOCK_MONOTONIC,
timeout_time, 0,
job_dispatch_timer, j);
if (r < 0)
return r;
(void) sd_event_source_set_description(j->timer_event_source, "job-start");
return 0;
}
void job_add_to_run_queue(Job *j) {
assert(j);
assert(j->installed);
if (j->in_run_queue)
return;
if (!j->manager->run_queue)
sd_event_source_set_enabled(j->manager->run_queue_event_source, SD_EVENT_ONESHOT);
LIST_PREPEND(run_queue, j->manager->run_queue, j);
j->in_run_queue = true;
}
void job_add_to_dbus_queue(Job *j) {
assert(j);
assert(j->installed);
if (j->in_dbus_queue)
return;
/* We don't check if anybody is subscribed here, since this
* job might just have been created and not yet assigned to a
* connection/client. */
LIST_PREPEND(dbus_queue, j->manager->dbus_job_queue, j);
j->in_dbus_queue = true;
}
char *job_dbus_path(Job *j) {
char *p;
assert(j);
if (asprintf(&p, "/org/freedesktop/systemd1/job/%"PRIu32, j->id) < 0)
return NULL;
return p;
}
int job_serialize(Job *j, FILE *f) {
assert(j);
assert(f);
fprintf(f, "job-id=%u\n", j->id);
fprintf(f, "job-type=%s\n", job_type_to_string(j->type));
fprintf(f, "job-state=%s\n", job_state_to_string(j->state));
fprintf(f, "job-irreversible=%s\n", yes_no(j->irreversible));
fprintf(f, "job-sent-dbus-new-signal=%s\n", yes_no(j->sent_dbus_new_signal));
fprintf(f, "job-ignore-order=%s\n", yes_no(j->ignore_order));
if (j->begin_usec > 0)
fprintf(f, "job-begin="USEC_FMT"\n", j->begin_usec);
if (j->begin_running_usec > 0)
fprintf(f, "job-begin-running="USEC_FMT"\n", j->begin_running_usec);
bus_track_serialize(j->bus_track, f, "subscribed");
/* End marker */
fputc('\n', f);
return 0;
}
int job_deserialize(Job *j, FILE *f) {
int r;
assert(j);
assert(f);
for (;;) {
_cleanup_free_ char *line = NULL;
char *l, *v;
size_t k;
r = read_line(f, LONG_LINE_MAX, &line);
if (r < 0)
return log_error_errno(r, "Failed to read serialization line: %m");
if (r == 0)
return 0;
l = strstrip(line);
/* End marker */
if (isempty(l))
return 0;
k = strcspn(l, "=");
if (l[k] == '=') {
l[k] = 0;
v = l+k+1;
} else
v = l+k;
if (streq(l, "job-id")) {
if (safe_atou32(v, &j->id) < 0)
log_debug("Failed to parse job id value %s", v);
} else if (streq(l, "job-type")) {
JobType t;
t = job_type_from_string(v);
if (t < 0)
log_debug("Failed to parse job type %s", v);
else if (t >= _JOB_TYPE_MAX_IN_TRANSACTION)
log_debug("Cannot deserialize job of type %s", v);
else
j->type = t;
} else if (streq(l, "job-state")) {
JobState s;
s = job_state_from_string(v);
if (s < 0)
log_debug("Failed to parse job state %s", v);
else
job_set_state(j, s);
} else if (streq(l, "job-irreversible")) {
int b;
b = parse_boolean(v);
if (b < 0)
log_debug("Failed to parse job irreversible flag %s", v);
else
j->irreversible = j->irreversible || b;
} else if (streq(l, "job-sent-dbus-new-signal")) {
int b;
b = parse_boolean(v);
if (b < 0)
log_debug("Failed to parse job sent_dbus_new_signal flag %s", v);
else
j->sent_dbus_new_signal = j->sent_dbus_new_signal || b;
} else if (streq(l, "job-ignore-order")) {
int b;
b = parse_boolean(v);
if (b < 0)
log_debug("Failed to parse job ignore_order flag %s", v);
else
j->ignore_order = j->ignore_order || b;
} else if (streq(l, "job-begin")) {
unsigned long long ull;
if (sscanf(v, "%llu", &ull) != 1)
log_debug("Failed to parse job-begin value %s", v);
else
j->begin_usec = ull;
} else if (streq(l, "job-begin-running")) {
unsigned long long ull;
if (sscanf(v, "%llu", &ull) != 1)
log_debug("Failed to parse job-begin-running value %s", v);
else
j->begin_running_usec = ull;
} else if (streq(l, "subscribed")) {
if (strv_extend(&j->deserialized_clients, v) < 0)
log_oom();
}
}
}
int job_coldplug(Job *j) {
int r;
usec_t timeout_time = USEC_INFINITY;
assert(j);
/* After deserialization is complete and the bus connection
* set up again, let's start watching our subscribers again */
(void) bus_job_coldplug_bus_track(j);
if (j->state == JOB_WAITING)
job_add_to_run_queue(j);
/* Maybe due to new dependencies we don't actually need this job anymore? */
job_add_to_gc_queue(j);
/* Create timer only when job began or began running and the respective timeout is finite.
* Follow logic of job_start_timer() if both timeouts are finite */
if (j->begin_usec == 0)
return 0;
if (j->unit->job_timeout != USEC_INFINITY)
timeout_time = usec_add(j->begin_usec, j->unit->job_timeout);
if (j->begin_running_usec > 0 && j->unit->job_running_timeout != USEC_INFINITY)
timeout_time = MIN(timeout_time, usec_add(j->begin_running_usec, j->unit->job_running_timeout));
if (timeout_time == USEC_INFINITY)
return 0;
j->timer_event_source = sd_event_source_unref(j->timer_event_source);
r = sd_event_add_time(
j->manager->event,
&j->timer_event_source,
CLOCK_MONOTONIC,
timeout_time, 0,
job_dispatch_timer, j);
if (r < 0)
log_debug_errno(r, "Failed to restart timeout for job: %m");
(void) sd_event_source_set_description(j->timer_event_source, "job-timeout");
return r;
}
void job_shutdown_magic(Job *j) {
assert(j);
/* The shutdown target gets some special treatment here: we
* tell the kernel to begin with flushing its disk caches, to
* optimize shutdown time a bit. Ideally we wouldn't hardcode
* this magic into PID 1. However all other processes aren't
* options either since they'd exit much sooner than PID 1 and
* asynchronous sync() would cause their exit to be
* delayed. */
if (j->type != JOB_START)
return;
if (!MANAGER_IS_SYSTEM(j->unit->manager))
return;
if (!unit_has_name(j->unit, SPECIAL_SHUTDOWN_TARGET))
return;
/* In case messages on console has been disabled on boot */
j->unit->manager->no_console_output = false;
if (detect_container() > 0)
return;
(void) asynchronous_sync(NULL);
}
int job_get_timeout(Job *j, usec_t *timeout) {
usec_t x = USEC_INFINITY, y = USEC_INFINITY;
Unit *u = j->unit;
int r;
assert(u);
if (j->timer_event_source) {
r = sd_event_source_get_time(j->timer_event_source, &x);
if (r < 0)
return r;
}
if (UNIT_VTABLE(u)->get_timeout) {
r = UNIT_VTABLE(u)->get_timeout(u, &y);
if (r < 0)
return r;
}
if (x == USEC_INFINITY && y == USEC_INFINITY)
return 0;
*timeout = MIN(x, y);
return 1;
}
bool job_may_gc(Job *j) {
Unit *other;
Iterator i;
void *v;
assert(j);
/* Checks whether this job should be GC'ed away. We only do this for jobs of units that have no effect on their
* own and just track external state. For now the only unit type that qualifies for this are .device units.
* Returns true if the job can be collected. */
if (!UNIT_VTABLE(j->unit)->gc_jobs)
return false;
if (sd_bus_track_count(j->bus_track) > 0)
return false;
/* FIXME: So this is a bit ugly: for now we don't properly track references made via private bus connections
* (because it's nasty, as sd_bus_track doesn't apply to it). We simply remember that the job was once
* referenced by one, and reset this whenever we notice that no private bus connections are around. This means
* the GC is a bit too conservative when it comes to jobs created by private bus connections. */
if (j->ref_by_private_bus) {
if (set_isempty(j->unit->manager->private_buses))
j->ref_by_private_bus = false;
else
return false;
}
if (j->type == JOB_NOP)
return false;
/* If a job is ordered after ours, and is to be started, then it needs to wait for us, regardless if we stop or
* start, hence let's not GC in that case. */
HASHMAP_FOREACH_KEY(v, other, j->unit->dependencies[UNIT_BEFORE], i) {
if (!other->job)
continue;
if (other->job->ignore_order)
continue;
if (IN_SET(other->job->type, JOB_START, JOB_VERIFY_ACTIVE, JOB_RELOAD))
return false;
}
/* If we are going down, but something else is ordered After= us, then it needs to wait for us */
if (IN_SET(j->type, JOB_STOP, JOB_RESTART))
HASHMAP_FOREACH_KEY(v, other, j->unit->dependencies[UNIT_AFTER], i) {
if (!other->job)
continue;
if (other->job->ignore_order)
continue;
return false;
}
/* The logic above is kinda the inverse of the job_is_runnable() logic. Specifically, if the job "we" is
* ordered before the job "other":
*
* we start + other start → stay
* we start + other stop → gc
* we stop + other start → stay
* we stop + other stop → gc
*
* "we" are ordered after "other":
*
* we start + other start → gc
* we start + other stop → gc
* we stop + other start → stay
* we stop + other stop → stay
*
*/
return true;
}
void job_add_to_gc_queue(Job *j) {
assert(j);
if (j->in_gc_queue)
return;
if (!job_may_gc(j))
return;
LIST_PREPEND(gc_queue, j->unit->manager->gc_job_queue, j);
j->in_gc_queue = true;
}
static int job_compare(const void *a, const void *b) {
Job *x = *(Job**) a, *y = *(Job**) b;
if (x->id < y->id)
return -1;
if (x->id > y->id)
return 1;
return 0;
}
static size_t sort_job_list(Job **list, size_t n) {
Job *previous = NULL;
size_t a, b;
/* Order by numeric IDs */
qsort_safe(list, n, sizeof(Job*), job_compare);
/* Filter out duplicates */
for (a = 0, b = 0; a < n; a++) {
if (previous == list[a])
continue;
previous = list[b++] = list[a];
}
return b;
}
int job_get_before(Job *j, Job*** ret) {
_cleanup_free_ Job** list = NULL;
size_t n = 0, n_allocated = 0;
Unit *other = NULL;
Iterator i;
void *v;
/* Returns a list of all pending jobs that need to finish before this job may be started. */
assert(j);
assert(ret);
if (j->ignore_order) {
*ret = NULL;
return 0;
}
if (IN_SET(j->type, JOB_START, JOB_VERIFY_ACTIVE, JOB_RELOAD)) {
HASHMAP_FOREACH_KEY(v, other, j->unit->dependencies[UNIT_AFTER], i) {
if (!other->job)
continue;
if (!GREEDY_REALLOC(list, n_allocated, n+1))
return -ENOMEM;
list[n++] = other->job;
}
}
HASHMAP_FOREACH_KEY(v, other, j->unit->dependencies[UNIT_BEFORE], i) {
if (!other->job)
continue;
if (!IN_SET(other->job->type, JOB_STOP, JOB_RESTART))
continue;
if (!GREEDY_REALLOC(list, n_allocated, n+1))
return -ENOMEM;
list[n++] = other->job;
}
n = sort_job_list(list, n);
*ret = TAKE_PTR(list);
return (int) n;
}
int job_get_after(Job *j, Job*** ret) {
_cleanup_free_ Job** list = NULL;
size_t n = 0, n_allocated = 0;
Unit *other = NULL;
void *v;
Iterator i;
assert(j);
assert(ret);
/* Returns a list of all pending jobs that are waiting for this job to finish. */
HASHMAP_FOREACH_KEY(v, other, j->unit->dependencies[UNIT_BEFORE], i) {
if (!other->job)
continue;
if (other->job->ignore_order)
continue;
if (!IN_SET(other->job->type, JOB_START, JOB_VERIFY_ACTIVE, JOB_RELOAD))
continue;
if (!GREEDY_REALLOC(list, n_allocated, n+1))
return -ENOMEM;
list[n++] = other->job;
}
if (IN_SET(j->type, JOB_STOP, JOB_RESTART)) {
HASHMAP_FOREACH_KEY(v, other, j->unit->dependencies[UNIT_AFTER], i) {
if (!other->job)
continue;
if (other->job->ignore_order)
continue;
if (!GREEDY_REALLOC(list, n_allocated, n+1))
return -ENOMEM;
list[n++] = other->job;
}
}
n = sort_job_list(list, n);
*ret = TAKE_PTR(list);
return (int) n;
}
static const char* const job_state_table[_JOB_STATE_MAX] = {
[JOB_WAITING] = "waiting",
[JOB_RUNNING] = "running",
};
DEFINE_STRING_TABLE_LOOKUP(job_state, JobState);
static const char* const job_type_table[_JOB_TYPE_MAX] = {
[JOB_START] = "start",
[JOB_VERIFY_ACTIVE] = "verify-active",
[JOB_STOP] = "stop",
[JOB_RELOAD] = "reload",
[JOB_RELOAD_OR_START] = "reload-or-start",
[JOB_RESTART] = "restart",
[JOB_TRY_RESTART] = "try-restart",
[JOB_TRY_RELOAD] = "try-reload",
[JOB_NOP] = "nop",
};
DEFINE_STRING_TABLE_LOOKUP(job_type, JobType);
static const char* const job_mode_table[_JOB_MODE_MAX] = {
[JOB_FAIL] = "fail",
[JOB_REPLACE] = "replace",
[JOB_REPLACE_IRREVERSIBLY] = "replace-irreversibly",
[JOB_ISOLATE] = "isolate",
[JOB_FLUSH] = "flush",
[JOB_IGNORE_DEPENDENCIES] = "ignore-dependencies",
[JOB_IGNORE_REQUIREMENTS] = "ignore-requirements",
};
DEFINE_STRING_TABLE_LOOKUP(job_mode, JobMode);
static const char* const job_result_table[_JOB_RESULT_MAX] = {
[JOB_DONE] = "done",
[JOB_CANCELED] = "canceled",
[JOB_TIMEOUT] = "timeout",
[JOB_FAILED] = "failed",
[JOB_DEPENDENCY] = "dependency",
[JOB_SKIPPED] = "skipped",
[JOB_INVALID] = "invalid",
[JOB_ASSERT] = "assert",
[JOB_UNSUPPORTED] = "unsupported",
[JOB_COLLECTED] = "collected",
[JOB_ONCE] = "once",
};
DEFINE_STRING_TABLE_LOOKUP(job_result, JobResult);
const char* job_type_to_access_method(JobType t) {
assert(t >= 0);
assert(t < _JOB_TYPE_MAX);
if (IN_SET(t, JOB_START, JOB_RESTART, JOB_TRY_RESTART))
return "start";
else if (t == JOB_STOP)
return "stop";
else
return "reload";
}