summaryrefslogtreecommitdiff
path: root/src/core
diff options
context:
space:
mode:
authorLennart Poettering <lennart@poettering.net>2015-09-23 01:06:56 +0200
committerLennart Poettering <lennart@poettering.net>2015-09-29 21:08:37 +0200
commit92ca4cac43055689696ea63ca7b7225e85707040 (patch)
treee9eedd458e8938dcf69cda8e5e83cb3284c51921 /src/core
parent2feceb5eb945d430999a8130e14c512af35b21d9 (diff)
main: minor clean-ups
Add (void) casting for a couple of functions where we knowingly ignore the returning error code. Use EXIT_FAILURE where appropriate. Try to initialize structures at declaration time, or at once.
Diffstat (limited to 'src/core')
-rw-r--r--src/core/main.c54
1 files changed, 28 insertions, 26 deletions
diff --git a/src/core/main.c b/src/core/main.c
index 815dbe7bd4..62bf94b7cb 100644
--- a/src/core/main.c
+++ b/src/core/main.c
@@ -129,7 +129,7 @@ noreturn static void crash(int sig) {
if (getpid() != 1)
/* Pass this on immediately, if this is not PID 1 */
- raise(sig);
+ (void) raise(sig);
else if (!arg_dump_core)
log_emergency("Caught <%s>, not dumping core.", signal_to_string(sig));
else {
@@ -140,34 +140,35 @@ noreturn static void crash(int sig) {
pid_t pid;
/* We want to wait for the core process, hence let's enable SIGCHLD */
- sigaction(SIGCHLD, &sa, NULL);
+ (void) sigaction(SIGCHLD, &sa, NULL);
pid = raw_clone(SIGCHLD, NULL);
if (pid < 0)
log_emergency_errno(errno, "Caught <%s>, cannot fork for core dump: %m", signal_to_string(sig));
-
else if (pid == 0) {
- struct rlimit rl = {};
+ struct rlimit rl = {
+ .rlim_cur = RLIM_INFINITY,
+ .rlim_max = RLIM_INFINITY,
+ };
/* Enable default signal handler for core dump */
- zero(sa);
- sa.sa_handler = SIG_DFL;
- sigaction(sig, &sa, NULL);
+ sa = (struct sigaction) {
+ .sa_handler = SIG_DFL,
+ };
+ (void) sigaction(sig, &sa, NULL);
/* Don't limit the core dump size */
- rl.rlim_cur = RLIM_INFINITY;
- rl.rlim_max = RLIM_INFINITY;
- setrlimit(RLIMIT_CORE, &rl);
+ (void) setrlimit(RLIMIT_CORE, &rl);
/* Just to be sure... */
(void) chdir("/");
/* Raise the signal again */
pid = raw_getpid();
- kill(pid, sig); /* raise() would kill the parent */
+ (void) kill(pid, sig); /* raise() would kill the parent */
assert_not_reached("We shouldn't be here...");
- _exit(1);
+ _exit(EXIT_FAILURE);
} else {
siginfo_t status;
int r;
@@ -190,7 +191,7 @@ noreturn static void crash(int sig) {
}
if (arg_crash_chvt)
- chvt(arg_crash_chvt);
+ (void) chvt(arg_crash_chvt);
if (arg_crash_shell) {
struct sigaction sa = {
@@ -200,20 +201,20 @@ noreturn static void crash(int sig) {
pid_t pid;
log_info("Executing crash shell in 10s...");
- sleep(10);
+ (void) sleep(10);
/* Let the kernel reap children for us */
- assert_se(sigaction(SIGCHLD, &sa, NULL) == 0);
+ (void) sigaction(SIGCHLD, &sa, NULL);
pid = raw_clone(SIGCHLD, NULL);
if (pid < 0)
log_emergency_errno(errno, "Failed to fork off crash shell: %m");
else if (pid == 0) {
- make_console_stdio();
- execle("/bin/sh", "/bin/sh", NULL, environ);
+ (void) make_console_stdio();
+ (void) execle("/bin/sh", "/bin/sh", NULL, environ);
log_emergency_errno(errno, "execle() failed: %m");
- _exit(1);
+ _exit(EXIT_FAILURE);
} else
log_info("Successfully spawned crash shell as PID "PID_FMT".", pid);
}
@@ -1824,7 +1825,7 @@ finish:
* that the new systemd can pass the kernel default to
* its child processes */
if (saved_rlimit_nofile.rlim_cur > 0)
- setrlimit(RLIMIT_NOFILE, &saved_rlimit_nofile);
+ (void) setrlimit(RLIMIT_NOFILE, &saved_rlimit_nofile);
if (switch_root_dir) {
/* Kill all remaining processes from the
@@ -1866,10 +1867,10 @@ finish:
/* do not pass along the environment we inherit from the kernel or initrd */
if (switch_root_dir)
- clearenv();
+ (void) clearenv();
assert(i <= args_size);
- execv(args[0], (char* const*) args);
+ (void) execv(args[0], (char* const*) args);
}
/* Try the fallback, if there is any, without any
@@ -1882,7 +1883,7 @@ finish:
fds = fdset_free(fds);
/* Reopen the console */
- make_console_stdio();
+ (void) make_console_stdio();
for (j = 1, i = 1; j < (unsigned) argc; j++)
args[i++] = argv[j];
@@ -1896,19 +1897,19 @@ finish:
if (switch_root_init) {
args[0] = switch_root_init;
- execv(args[0], (char* const*) args);
+ (void) execv(args[0], (char* const*) args);
log_warning_errno(errno, "Failed to execute configured init, trying fallback: %m");
}
args[0] = "/sbin/init";
- execv(args[0], (char* const*) args);
+ (void) execv(args[0], (char* const*) args);
if (errno == ENOENT) {
log_warning("No /sbin/init, trying fallback");
args[0] = "/bin/sh";
args[1] = NULL;
- execv(args[0], (char* const*) args);
+ (void) execv(args[0], (char* const*) args);
log_error_errno(errno, "Failed to execute /bin/sh, giving up: %m");
} else
log_warning_errno(errno, "Failed to execute /sbin/init, giving up: %m");
@@ -1944,6 +1945,7 @@ finish:
xsprintf(log_level, "%d", log_get_max_level());
switch (log_get_target()) {
+
case LOG_TARGET_KMSG:
case LOG_TARGET_JOURNAL_OR_KMSG:
case LOG_TARGET_SYSLOG_OR_KMSG:
@@ -1985,7 +1987,7 @@ finish:
/* Tell the binary how often to ping, ignore failure */
if (asprintf(&e, "WATCHDOG_USEC="USEC_FMT, arg_shutdown_watchdog) > 0)
- strv_push(&env_block, e);
+ (void) strv_push(&env_block, e);
} else
watchdog_close(true);