Copyright 2010 Lennart Poettering
systemd is free software; you can redistribute it and/or modify it
- under the terms of the GNU General Public License as published by
- the Free Software Foundation; either version 2 of the License, or
+ under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 2.1 of the License, or
(at your option) any later version.
systemd is distributed in the hope that it will be useful, but
WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
- General Public License for more details.
+ Lesser General Public License for more details.
- You should have received a copy of the GNU General Public License
+ You should have received a copy of the GNU Lesser General Public License
along with systemd; If not, see <http://www.gnu.org/licenses/>.
***/
#include <sys/mount.h>
#include <linux/fs.h>
#include <linux/oom.h>
+#include <sys/poll.h>
#ifdef HAVE_PAM
#include <security/pam_appl.h>
#include "utmp-wtmp.h"
#include "def.h"
#include "loopback-setup.h"
+#include "path-util.h"
+
+#define IDLE_TIMEOUT_USEC (5*USEC_PER_SEC)
/* This assumes there is a 'tty' group */
#define TTY_MODE 0620
static int setup_pam(
const char *name,
const char *user,
+ uid_t uid,
const char *tty,
char ***pam_env,
int fds[], unsigned n_fds) {
open here that have been opened by PAM. */
close_many(fds, n_fds);
- /* Wait until our parent died. This will most likely
- * not work since the kernel does not allow
- * unprivileged parents kill their privileged children
- * this way. We rely on the control groups kill logic
+ /* Drop privileges - we don't need any to pam_close_session
+ * and this will make PR_SET_PDEATHSIG work in most cases.
+ * If this fails, ignore the error - but expect sd-pam threads
+ * to fail to exit normally */
+ if (setresuid(uid, uid, uid) < 0)
+ log_error("Error: Failed to setresuid() in sd-pam: %s", strerror(-r));
+
+ /* Wait until our parent died. This will only work if
+ * the above setresuid() succeeds, otherwise the kernel
+ * will not allow unprivileged parents kill their privileged
+ * children this way. We rely on the control groups kill logic
* to do the rest for us. */
if (prctl(PR_SET_PDEATHSIG, SIGTERM) < 0)
goto child_finish;
}
#endif
-static int do_capability_bounding_set_drop(uint64_t drop) {
- unsigned long i;
- cap_t old_cap = NULL, new_cap = NULL;
- cap_flag_value_t fv;
- int r;
-
- /* If we are run as PID 1 we will lack CAP_SETPCAP by default
- * in the effective set (yes, the kernel drops that when
- * executing init!), so get it back temporarily so that we can
- * call PR_CAPBSET_DROP. */
-
- old_cap = cap_get_proc();
- if (!old_cap)
- return -errno;
-
- if (cap_get_flag(old_cap, CAP_SETPCAP, CAP_EFFECTIVE, &fv) < 0) {
- r = -errno;
- goto finish;
- }
-
- if (fv != CAP_SET) {
- static const cap_value_t v = CAP_SETPCAP;
-
- new_cap = cap_dup(old_cap);
- if (!new_cap) {
- r = -errno;
- goto finish;
- }
-
- if (cap_set_flag(new_cap, CAP_EFFECTIVE, 1, &v, CAP_SET) < 0) {
- r = -errno;
- goto finish;
- }
-
- if (cap_set_proc(new_cap) < 0) {
- r = -errno;
- goto finish;
- }
- }
-
- for (i = 0; i <= cap_last_cap(); i++)
- if (drop & ((uint64_t) 1ULL << (uint64_t) i)) {
- if (prctl(PR_CAPBSET_DROP, i) < 0) {
- r = -errno;
- goto finish;
- }
- }
-
- r = 0;
-
-finish:
- if (new_cap)
- cap_free(new_cap);
-
- if (old_cap) {
- cap_set_proc(old_cap);
- cap_free(old_cap);
- }
-
- return r;
-}
-
static void rename_process_from_path(const char *path) {
char process_name[11];
const char *p;
/* This resulting string must fit in 10 chars (i.e. the length
* of "/sbin/init") to look pretty in /bin/ps */
- p = file_name_from_path(path);
+ p = path_get_file_name(path);
if (isempty(p)) {
rename_process("(...)");
return;
bool confirm_spawn,
CGroupBonding *cgroup_bondings,
CGroupAttribute *cgroup_attributes,
+ const char *cgroup_suffix,
+ int idle_pipe[2],
pid_t *ret) {
pid_t pid;
goto fail_child;
}
+ if (idle_pipe) {
+ if (idle_pipe[1] >= 0)
+ close_nointr_nofail(idle_pipe[1]);
+ if (idle_pipe[0] >= 0) {
+ fd_wait_for_event(idle_pipe[0], POLLHUP, IDLE_TIMEOUT_USEC);
+ close_nointr_nofail(idle_pipe[0]);
+ }
+ }
+
/* Close sockets very early to make sure we don't
* block init reexecution because it cannot bind its
* sockets */
}
if (!keep_stdout) {
- err = setup_output(context, socket_fd, file_name_from_path(command->path), apply_tty_stdin);
+ err = setup_output(context, socket_fd, path_get_file_name(command->path), apply_tty_stdin);
if (err < 0) {
r = EXIT_STDOUT;
goto fail_child;
}
}
- err = setup_error(context, socket_fd, file_name_from_path(command->path), apply_tty_stdin);
+ err = setup_error(context, socket_fd, path_get_file_name(command->path), apply_tty_stdin);
if (err < 0) {
r = EXIT_STDERR;
goto fail_child;
}
if (cgroup_bondings) {
- err = cgroup_bonding_install_list(cgroup_bondings, 0);
+ err = cgroup_bonding_install_list(cgroup_bondings, 0, cgroup_suffix);
if (err < 0) {
r = EXIT_CGROUP;
goto fail_child;
char_array_0(t);
if (write_one_line_file("/proc/self/oom_score_adj", t) < 0) {
- /* Compatibility with Linux <= 2.6.35 */
-
- int adj;
-
- adj = (context->oom_score_adjust * -OOM_DISABLE) / OOM_SCORE_ADJ_MAX;
- adj = CLAMP(adj, OOM_DISABLE, OOM_ADJUST_MAX);
-
- snprintf(t, sizeof(t), "%i", adj);
- char_array_0(t);
-
- if (write_one_line_file("/proc/self/oom_adj", t) < 0
- && errno != EACCES) {
- err = -errno;
- r = EXIT_OOM_ADJUST;
- goto fail_child;
- }
+ err = -errno;
+ r = EXIT_OOM_ADJUST;
+ goto fail_child;
}
}
goto fail_child;
}
- if (context->timer_slack_nsec_set)
+ if (context->timer_slack_nsec != (nsec_t) -1)
if (prctl(PR_SET_TIMERSLACK, context->timer_slack_nsec) < 0) {
err = -errno;
r = EXIT_TIMERSLACK;
#ifdef HAVE_PAM
if (context->pam_name && username) {
- err = setup_pam(context->pam_name, username, context->tty_path, &pam_env, fds, n_fds);
+ err = setup_pam(context->pam_name, username, uid, context->tty_path, &pam_env, fds, n_fds);
if (err < 0) {
r = EXIT_PAM;
goto fail_child;
if (!context->rlimit[i])
continue;
- if (setrlimit(i, context->rlimit[i]) < 0) {
+ if (setrlimit_closest(i, context->rlimit[i]) < 0) {
err = -errno;
r = EXIT_LIMITS;
goto fail_child;
}
if (context->capability_bounding_set_drop) {
- err = do_capability_bounding_set_drop(context->capability_bounding_set_drop);
+ err = capability_bounding_set_drop(context->capability_bounding_set_drop, false);
if (err < 0) {
r = EXIT_CAPABILITIES;
goto fail_child;
* sure that when we kill the cgroup the process will be
* killed too). */
if (cgroup_bondings)
- cgroup_bonding_install_list(cgroup_bondings, pid);
+ cgroup_bonding_install_list(cgroup_bondings, pid, cgroup_suffix);
log_debug("Forked %s as %lu", command->path, (unsigned long) pid);
c->send_sigkill = true;
c->control_group_persistent = -1;
c->ignore_sigpipe = true;
+ c->timer_slack_nsec = (nsec_t) -1;
}
void exec_context_done(ExecContext *c) {
fputs("\n", f);
}
- if (c->timer_slack_nsec_set)
- fprintf(f, "%sTimerSlackNSec: %lu\n", prefix, c->timer_slack_nsec);
+ if (c->timer_slack_nsec != (nsec_t) -1)
+ fprintf(f, "%sTimerSlackNSec: %lu\n", prefix, (unsigned long)c->timer_slack_nsec);
fprintf(f,
"%sStandardInput: %s\n"