summaryrefslogtreecommitdiffstats
path: root/src/exec_nopty.c
diff options
context:
space:
mode:
authorDaniel Baumann <daniel.baumann@progress-linux.org>2024-04-15 13:14:46 +0000
committerDaniel Baumann <daniel.baumann@progress-linux.org>2024-04-15 13:14:46 +0000
commit025c439e829e0db9ac511cd9c1b8d5fd53475ead (patch)
treefa6986b4690f991613ffb97cea1f6942427baf5d /src/exec_nopty.c
parentInitial commit. (diff)
downloadsudo-025c439e829e0db9ac511cd9c1b8d5fd53475ead.tar.xz
sudo-025c439e829e0db9ac511cd9c1b8d5fd53475ead.zip
Adding upstream version 1.9.15p5.upstream/1.9.15p5upstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'src/exec_nopty.c')
-rw-r--r--src/exec_nopty.c812
1 files changed, 812 insertions, 0 deletions
diff --git a/src/exec_nopty.c b/src/exec_nopty.c
new file mode 100644
index 0000000..1b1ae6d
--- /dev/null
+++ b/src/exec_nopty.c
@@ -0,0 +1,812 @@
+/*
+ * SPDX-License-Identifier: ISC
+ *
+ * Copyright (c) 2009-2023 Todd C. Miller <Todd.Miller@sudo.ws>
+ *
+ * Permission to use, copy, modify, and distribute this software for any
+ * purpose with or without fee is hereby granted, provided that the above
+ * copyright notice and this permission notice appear in all copies.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+ * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+ * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+ * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+ * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+ * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+ */
+
+/*
+ * This is an open source non-commercial project. Dear PVS-Studio, please check it.
+ * PVS-Studio Static Code Analyzer for C, C++ and C#: http://www.viva64.com
+ */
+
+#include <config.h>
+
+#include <sys/types.h>
+#include <sys/socket.h>
+#include <sys/wait.h>
+#include <sys/ioctl.h>
+
+#if defined(HAVE_STDINT_H)
+# include <stdint.h>
+#elif defined(HAVE_INTTYPES_H)
+# include <inttypes.h>
+#endif
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+#include <errno.h>
+#include <fcntl.h>
+#include <signal.h>
+#include <termios.h> /* for struct winsize on HP-UX */
+
+#include <sudo.h>
+#include <sudo_exec.h>
+#include <sudo_plugin.h>
+#include <sudo_plugin_int.h>
+
+static void handle_sigchld_nopty(struct exec_closure *ec);
+
+/*
+ * Handle window size change events.
+ */
+static void
+handle_sigwinch(struct exec_closure *ec, int fd)
+{
+ struct winsize wsize;
+ debug_decl(handle_sigwinch, SUDO_DEBUG_EXEC);
+
+ if (fd != -1 && ioctl(fd, TIOCGWINSZ, &wsize) == 0) {
+ if (wsize.ws_row != ec->rows || wsize.ws_col != ec->cols) {
+ /* Log window change event. */
+ log_winchange(ec, wsize.ws_row, wsize.ws_col);
+
+ /* Update rows/cols. */
+ ec->rows = wsize.ws_row;
+ ec->cols = wsize.ws_col;
+ }
+ }
+}
+
+/* Note: this is basically the same as mon_errpipe_cb() in exec_monitor.c */
+static void
+errpipe_cb(int fd, int what, void *v)
+{
+ struct exec_closure *ec = v;
+ ssize_t nread;
+ int errval;
+ debug_decl(errpipe_cb, SUDO_DEBUG_EXEC);
+
+ /*
+ * Read errno from child or EOF when command is executed.
+ * Note that the error pipe is *blocking*.
+ */
+ nread = read(fd, &errval, sizeof(errval));
+ switch (nread) {
+ case -1:
+ if (errno != EAGAIN && errno != EINTR) {
+ if (ec->cstat->val == CMD_INVALID) {
+ /* XXX - need a way to distinguish non-exec error. */
+ ec->cstat->type = CMD_ERRNO;
+ ec->cstat->val = errno;
+ }
+ sudo_debug_printf(SUDO_DEBUG_ERROR|SUDO_DEBUG_ERRNO,
+ "%s: failed to read error pipe", __func__);
+ sudo_ev_loopbreak(ec->evbase);
+ }
+ break;
+ default:
+ if (nread == 0) {
+ /* The error pipe closes when the command is executed. */
+ sudo_debug_printf(SUDO_DEBUG_INFO, "EOF on error pipe");
+ } else {
+ /* Errno value when child is unable to execute command. */
+ sudo_debug_printf(SUDO_DEBUG_INFO, "errno from child: %s",
+ strerror(errval));
+ ec->cstat->type = CMD_ERRNO;
+ ec->cstat->val = errval;
+ }
+ sudo_ev_del(ec->evbase, ec->backchannel_event);
+ close(fd);
+ break;
+ }
+ debug_return;
+}
+
+/* Signal callback */
+static void
+signal_cb_nopty(int signo, int what, void *v)
+{
+ struct sudo_ev_siginfo_container *sc = v;
+ struct exec_closure *ec = sc->closure;
+ char signame[SIG2STR_MAX];
+ pid_t si_pgrp;
+ debug_decl(signal_cb_nopty, SUDO_DEBUG_EXEC);
+
+ if (ec->cmnd_pid == -1)
+ debug_return;
+
+ if (sig2str(signo, signame) == -1)
+ (void)snprintf(signame, sizeof(signame), "%d", signo);
+ sudo_debug_printf(SUDO_DEBUG_DIAG,
+ "%s: evbase %p, command: %d, signo %s(%d), cstat %p",
+ __func__, ec->evbase, (int)ec->cmnd_pid, signame, signo, ec->cstat);
+
+ switch (signo) {
+ case SIGCHLD:
+ handle_sigchld_nopty(ec);
+ if (ec->cmnd_pid == -1) {
+ /* Command exited or was killed, exit event loop. */
+ sudo_ev_loopexit(ec->evbase);
+ }
+ debug_return;
+ case SIGWINCH:
+ handle_sigwinch(ec, io_fds[SFD_USERTTY]);
+ FALLTHROUGH;
+#ifdef SIGINFO
+ case SIGINFO:
+#endif
+ case SIGINT:
+ case SIGQUIT:
+ case SIGTSTP:
+ /*
+ * Only forward user-generated signals not sent by a process other than
+ * the command itself or a member of the command's process group (but
+ * only when either sudo or the command is the process group leader).
+ * Signals sent by the kernel may include SIGTSTP when the user presses
+ * ^Z. Curses programs often trap ^Z and send SIGTSTP to their own
+ * process group, so we don't want to send an extra SIGTSTP.
+ */
+ if (!USER_SIGNALED(sc->siginfo))
+ debug_return;
+ if (sc->siginfo->si_pid != 0) {
+ if (sc->siginfo->si_pid == ec->cmnd_pid)
+ debug_return;
+ si_pgrp = getpgid(sc->siginfo->si_pid);
+ if (si_pgrp != -1) {
+ if (si_pgrp == ec->cmnd_pid || si_pgrp == ec->sudo_pid)
+ debug_return;
+ }
+ }
+ break;
+ default:
+ /*
+ * Do not forward signals sent by the command itself or a member of the
+ * command's process group (but only when either sudo or the command is
+ * the process group leader). We don't want the command to indirectly
+ * kill itself. For example, this can happen with some versions of
+ * reboot that call kill(-1, SIGTERM) to kill all other processes.
+ */
+ if (USER_SIGNALED(sc->siginfo) && sc->siginfo->si_pid != 0) {
+ if (sc->siginfo->si_pid == ec->cmnd_pid)
+ debug_return;
+ si_pgrp = getpgid(sc->siginfo->si_pid);
+ if (si_pgrp != -1) {
+ if (si_pgrp == ec->cmnd_pid || si_pgrp == ec->sudo_pid)
+ debug_return;
+ }
+ }
+ break;
+ }
+
+ /* Send signal to command. */
+ if (signo == SIGALRM) {
+ terminate_command(ec->cmnd_pid, false);
+ } else if (kill(ec->cmnd_pid, signo) != 0) {
+ sudo_warn("kill(%d, SIG%s)", (int)ec->cmnd_pid, signame);
+ }
+
+ debug_return;
+}
+
+
+/*
+ * Fill in the exec closure and setup initial exec events.
+ * Allocates events for the signal pipe and error pipe.
+ */
+static void
+fill_exec_closure(struct exec_closure *ec, struct command_status *cstat,
+ struct command_details *details, const struct user_details *user_details,
+ struct sudo_event_base *evbase, int errfd)
+{
+ debug_decl(fill_exec_closure, SUDO_DEBUG_EXEC);
+
+ /* Fill in the non-event part of the closure. */
+ ec->sudo_pid = getpid();
+ ec->ppgrp = getpgrp();
+ ec->cstat = cstat;
+ ec->details = details;
+ ec->rows = user_details->ts_rows;
+ ec->cols = user_details->ts_cols;
+
+ /* Setup event base and events. */
+ ec->evbase = evbase;
+
+ /* Event for command status via errfd. */
+ ec->backchannel_event = sudo_ev_alloc(errfd,
+ SUDO_EV_READ|SUDO_EV_PERSIST, errpipe_cb, ec);
+ if (ec->backchannel_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->backchannel_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+ sudo_debug_printf(SUDO_DEBUG_INFO, "error pipe fd %d\n", errfd);
+
+ /* Events for local signals. */
+ ec->sigint_event = sudo_ev_alloc(SIGINT,
+ SUDO_EV_SIGINFO, signal_cb_nopty, ec);
+ if (ec->sigint_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->sigint_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+
+ ec->sigquit_event = sudo_ev_alloc(SIGQUIT,
+ SUDO_EV_SIGINFO, signal_cb_nopty, ec);
+ if (ec->sigquit_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->sigquit_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+
+ ec->sigtstp_event = sudo_ev_alloc(SIGTSTP,
+ SUDO_EV_SIGINFO, signal_cb_nopty, ec);
+ if (ec->sigtstp_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->sigtstp_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+
+ ec->sigterm_event = sudo_ev_alloc(SIGTERM,
+ SUDO_EV_SIGINFO, signal_cb_nopty, ec);
+ if (ec->sigterm_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->sigterm_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+
+ ec->sighup_event = sudo_ev_alloc(SIGHUP,
+ SUDO_EV_SIGINFO, signal_cb_nopty, ec);
+ if (ec->sighup_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->sighup_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+
+ ec->sigalrm_event = sudo_ev_alloc(SIGALRM,
+ SUDO_EV_SIGINFO, signal_cb_nopty, ec);
+ if (ec->sigalrm_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->sigalrm_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+
+ ec->sigpipe_event = sudo_ev_alloc(SIGPIPE,
+ SUDO_EV_SIGINFO, signal_cb_nopty, ec);
+ if (ec->sigpipe_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->sigpipe_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+
+ ec->sigusr1_event = sudo_ev_alloc(SIGUSR1,
+ SUDO_EV_SIGINFO, signal_cb_nopty, ec);
+ if (ec->sigusr1_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->sigusr1_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+
+ ec->sigusr2_event = sudo_ev_alloc(SIGUSR2,
+ SUDO_EV_SIGINFO, signal_cb_nopty, ec);
+ if (ec->sigusr2_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->sigusr2_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+
+ ec->sigchld_event = sudo_ev_alloc(SIGCHLD,
+ SUDO_EV_SIGINFO, signal_cb_nopty, ec);
+ if (ec->sigchld_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->sigchld_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+
+ ec->sigcont_event = sudo_ev_alloc(SIGCONT,
+ SUDO_EV_SIGINFO, signal_cb_nopty, ec);
+ if (ec->sigcont_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->sigcont_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+
+#ifdef SIGINFO
+ ec->siginfo_event = sudo_ev_alloc(SIGINFO,
+ SUDO_EV_SIGINFO, signal_cb_nopty, ec);
+ if (ec->siginfo_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->siginfo_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+#endif
+
+ ec->sigwinch_event = sudo_ev_alloc(SIGWINCH,
+ SUDO_EV_SIGINFO, signal_cb_nopty, ec);
+ if (ec->sigwinch_event == NULL)
+ sudo_fatalx(U_("%s: %s"), __func__, U_("unable to allocate memory"));
+ if (sudo_ev_add(ec->evbase, ec->sigwinch_event, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+
+ /* Set the default event base. */
+ sudo_ev_base_setdef(ec->evbase);
+
+ debug_return;
+}
+
+/*
+ * Read an iobuf that is ready.
+ */
+static void
+read_callback(int fd, int what, void *v)
+{
+ struct io_buffer *iob = v;
+ struct sudo_event_base *evbase = sudo_ev_get_base(iob->revent);
+ ssize_t n;
+ debug_decl(read_callback, SUDO_DEBUG_EXEC);
+
+ n = read(fd, iob->buf + iob->len, sizeof(iob->buf) - iob->len);
+ switch (n) {
+ case -1:
+ if (errno == EAGAIN || errno == EINTR) {
+ /* Not an error, retry later. */
+ break;
+ }
+ /* Treat read error as fatal and close the fd. */
+ sudo_debug_printf(SUDO_DEBUG_ERROR,
+ "error reading fd %d: %s", fd, strerror(errno));
+ FALLTHROUGH;
+ case 0:
+ /* got EOF */
+ if (n == 0) {
+ sudo_debug_printf(SUDO_DEBUG_INFO,
+ "read EOF from fd %d", fd);
+ }
+ safe_close(fd);
+ ev_free_by_fd(evbase, fd);
+ /* If writer already consumed the buffer, close it too. */
+ if (iob->wevent != NULL && iob->off == iob->len) {
+ safe_close(sudo_ev_get_fd(iob->wevent));
+ ev_free_by_fd(evbase, sudo_ev_get_fd(iob->wevent));
+ iob->off = iob->len = 0;
+ }
+ break;
+ default:
+ sudo_debug_printf(SUDO_DEBUG_INFO,
+ "read %zd bytes from fd %d", n, fd);
+ if (!iob->action(iob->buf + iob->len, (unsigned int)n, iob)) {
+ terminate_command(iob->ec->cmnd_pid, false);
+ iob->ec->cmnd_pid = -1;
+ }
+ iob->len += (unsigned int)n;
+ /* Disable reader if buffer is full. */
+ if (iob->len == sizeof(iob->buf))
+ sudo_ev_del(evbase, iob->revent);
+ /* Enable writer now that there is new data in the buffer. */
+ if (iob->wevent != NULL) {
+ if (sudo_ev_add(evbase, iob->wevent, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+ }
+ break;
+ }
+
+ debug_return;
+}
+
+/*
+ * Write an iobuf that is ready.
+ */
+static void
+write_callback(int fd, int what, void *v)
+{
+ struct io_buffer *iob = v;
+ struct sudo_event_base *evbase = sudo_ev_get_base(iob->wevent);
+ ssize_t n;
+ debug_decl(write_callback, SUDO_DEBUG_EXEC);
+
+ n = write(fd, iob->buf + iob->off, iob->len - iob->off);
+ if (n == -1) {
+ switch (errno) {
+ case EPIPE:
+ case EBADF:
+ /* other end of pipe closed */
+ sudo_debug_printf(SUDO_DEBUG_INFO,
+ "unable to write %u bytes to fd %d",
+ iob->len - iob->off, fd);
+ /* Close reader if there is one. */
+ if (iob->revent != NULL) {
+ safe_close(sudo_ev_get_fd(iob->revent));
+ ev_free_by_fd(evbase, sudo_ev_get_fd(iob->revent));
+ }
+ safe_close(fd);
+ ev_free_by_fd(evbase, fd);
+ break;
+ case EINTR:
+ case EAGAIN:
+ /* Not an error, retry later. */
+ break;
+ default:
+ /* XXX - need a way to distinguish non-exec error. */
+ iob->ec->cstat->type = CMD_ERRNO;
+ iob->ec->cstat->val = errno;
+ sudo_debug_printf(SUDO_DEBUG_ERROR,
+ "error writing fd %d: %s", fd, strerror(errno));
+ sudo_ev_loopbreak(evbase);
+ break;
+ }
+ } else {
+ sudo_debug_printf(SUDO_DEBUG_INFO,
+ "wrote %zd bytes to fd %d", n, fd);
+ iob->off += (unsigned int)n;
+ /* Disable writer and reset the buffer if fully consumed. */
+ if (iob->off == iob->len) {
+ iob->off = iob->len = 0;
+ sudo_ev_del(evbase, iob->wevent);
+ /* Forward the EOF from reader to writer. */
+ if (iob->revent == NULL) {
+ safe_close(fd);
+ ev_free_by_fd(evbase, fd);
+ }
+ }
+ /*
+ * Enable reader if buffer is not full but avoid reading
+ * /dev/tty if the command is no longer running.
+ */
+ if (iob->revent != NULL && iob->len != sizeof(iob->buf)) {
+ if (!USERTTY_EVENT(iob->revent) || iob->ec->cmnd_pid != -1) {
+ if (sudo_ev_add(evbase, iob->revent, NULL, false) == -1)
+ sudo_fatal("%s", U_("unable to add event to queue"));
+ }
+ }
+ }
+
+ debug_return;
+}
+
+/*
+ * If std{in,out,err} are not connected to a terminal, interpose
+ * ourselves using a pipe. Fills in io_pipe[][].
+ */
+static void
+interpose_pipes(struct exec_closure *ec, const char *tty, int io_pipe[3][2])
+{
+ bool interpose[3] = { false, false, false };
+ struct stat sb, tty_sbuf, *tty_sb = NULL;
+ struct plugin_container *plugin;
+ bool want_winch = false;
+ debug_decl(interpose_pipes, SUDO_DEBUG_EXEC);
+
+ /*
+ * Determine whether any of std{in,out,err} or window size changes
+ * should be logged.
+ */
+ TAILQ_FOREACH(plugin, &io_plugins, entries) {
+ if (plugin->u.io->log_stdin)
+ interpose[STDIN_FILENO] = true;
+ if (plugin->u.io->log_stdout)
+ interpose[STDOUT_FILENO] = true;
+ if (plugin->u.io->log_stderr)
+ interpose[STDERR_FILENO] = true;
+ if (plugin->u.io->version >= SUDO_API_MKVERSION(1, 12)) {
+ if (plugin->u.io->change_winsize)
+ want_winch = true;
+ }
+ }
+
+ /*
+ * If stdin, stdout or stderr is not the user's tty and logging is
+ * enabled, use a pipe to interpose ourselves.
+ */
+ if (tty != NULL && stat(tty, &tty_sbuf) != -1)
+ tty_sb = &tty_sbuf;
+
+ if (interpose[STDIN_FILENO]) {
+ if (!fd_matches_tty(STDIN_FILENO, tty_sb, &sb)) {
+ sudo_debug_printf(SUDO_DEBUG_INFO,
+ "stdin not user's tty, creating a pipe");
+ if (pipe2(io_pipe[STDIN_FILENO], O_CLOEXEC) != 0)
+ sudo_fatal("%s", U_("unable to create pipe"));
+ io_buf_new(STDIN_FILENO, io_pipe[STDIN_FILENO][1],
+ log_stdin, read_callback, write_callback, ec);
+ }
+ }
+ if (interpose[STDOUT_FILENO]) {
+ if (!fd_matches_tty(STDOUT_FILENO, tty_sb, &sb)) {
+ sudo_debug_printf(SUDO_DEBUG_INFO,
+ "stdout not user's tty, creating a pipe");
+ if (pipe2(io_pipe[STDOUT_FILENO], O_CLOEXEC) != 0)
+ sudo_fatal("%s", U_("unable to create pipe"));
+ io_buf_new(io_pipe[STDOUT_FILENO][0], STDOUT_FILENO,
+ log_stdout, read_callback, write_callback, ec);
+ }
+ }
+ if (interpose[STDERR_FILENO]) {
+ if (!fd_matches_tty(STDERR_FILENO, tty_sb, &sb)) {
+ sudo_debug_printf(SUDO_DEBUG_INFO,
+ "stderr not user's tty, creating a pipe");
+ if (pipe2(io_pipe[STDERR_FILENO], O_CLOEXEC) != 0)
+ sudo_fatal("%s", U_("unable to create pipe"));
+ io_buf_new(io_pipe[STDERR_FILENO][0], STDERR_FILENO,
+ log_stderr, read_callback, write_callback, ec);
+ }
+ }
+ if (want_winch) {
+ /* Need /dev/tty for SIGWINCH handling. */
+ io_fds[SFD_USERTTY] = open(_PATH_TTY, O_RDWR);
+ }
+}
+
+/*
+ * Execute a command and wait for it to finish.
+ */
+void
+exec_nopty(struct command_details *details,
+ const struct user_details *user_details,
+ struct sudo_event_base *evbase, struct command_status *cstat)
+{
+ int io_pipe[3][2] = { { -1, -1 }, { -1, -1 }, { -1, -1 } };
+ int errpipe[2], intercept_sv[2] = { -1, -1 };
+ struct exec_closure ec = { 0 };
+ sigset_t set, oset;
+ debug_decl(exec_nopty, SUDO_DEBUG_EXEC);
+
+ /*
+ * The policy plugin's session init must be run before we fork
+ * or certain pam modules won't be able to track their state.
+ */
+ if (policy_init_session(details) != true)
+ sudo_fatalx("%s", U_("policy plugin failed session initialization"));
+
+ /*
+ * We use a pipe to get errno if execve(2) fails in the child.
+ */
+ if (pipe2(errpipe, O_CLOEXEC) != 0)
+ sudo_fatal("%s", U_("unable to create pipe"));
+
+ if (ISSET(details->flags, CD_INTERCEPT|CD_LOG_SUBCMDS)) {
+ if (!ISSET(details->flags, CD_USE_PTRACE)) {
+ /*
+ * Allocate a socketpair for communicating with sudo_intercept.so.
+ * This must be inherited across exec, hence no FD_CLOEXEC.
+ */
+ if (socketpair(PF_UNIX, SOCK_STREAM, 0, intercept_sv) == -1)
+ sudo_fatal("%s", U_("unable to create sockets"));
+ }
+ }
+
+ /* Interpose std{in,out,err} with pipes if logging I/O. */
+ interpose_pipes(&ec, user_details->tty, io_pipe);
+
+ /*
+ * Block signals until we have our handlers setup in the parent so
+ * we don't miss SIGCHLD if the command exits immediately.
+ */
+ sigfillset(&set);
+ sigprocmask(SIG_BLOCK, &set, &oset);
+
+ /* Check for early termination or suspend signals before we fork. */
+ if (sudo_terminated(cstat)) {
+ sigprocmask(SIG_SETMASK, &oset, NULL);
+ debug_return;
+ }
+
+#ifdef HAVE_SELINUX
+ if (ISSET(details->flags, CD_RBAC_ENABLED)) {
+ if (selinux_relabel_tty(details->tty, -1) == -1) {
+ cstat->type = CMD_ERRNO;
+ cstat->val = errno;
+ debug_return;
+ }
+ selinux_audit_role_change();
+ }
+#endif
+
+ ec.cmnd_pid = sudo_debug_fork();
+ switch (ec.cmnd_pid) {
+ case -1:
+ sudo_fatal("%s", U_("unable to fork"));
+ break;
+ case 0:
+ /* child */
+ close(errpipe[0]);
+ if (intercept_sv[0] != -1)
+ close(intercept_sv[0]);
+ /* Replace stdin/stdout/stderr with pipes as needed and exec. */
+ if (io_pipe[STDIN_FILENO][0] != -1) {
+ if (dup3(io_pipe[STDIN_FILENO][0], STDIN_FILENO, 0) == -1)
+ sudo_fatal("dup3");
+ close(io_pipe[STDIN_FILENO][0]);
+ close(io_pipe[STDIN_FILENO][1]);
+ }
+ if (io_pipe[STDOUT_FILENO][0] != -1) {
+ if (dup3(io_pipe[STDOUT_FILENO][1], STDOUT_FILENO, 0) == -1)
+ sudo_fatal("dup3");
+ close(io_pipe[STDOUT_FILENO][0]);
+ close(io_pipe[STDOUT_FILENO][1]);
+ }
+ if (io_pipe[STDERR_FILENO][0] != -1) {
+ if (dup3(io_pipe[STDERR_FILENO][1], STDERR_FILENO, 0) == -1)
+ sudo_fatal("dup3");
+ close(io_pipe[STDERR_FILENO][0]);
+ close(io_pipe[STDERR_FILENO][1]);
+ }
+ exec_cmnd(details, &oset, intercept_sv[1], errpipe[1]);
+ while (write(errpipe[1], &errno, sizeof(int)) == -1) {
+ if (errno != EINTR)
+ break;
+ }
+ sudo_debug_exit_int(__func__, __FILE__, __LINE__, sudo_debug_subsys, 1);
+ _exit(EXIT_FAILURE);
+ /* NOTREACHED */
+ }
+ sudo_debug_printf(SUDO_DEBUG_INFO, "executed %s, pid %d", details->command,
+ (int)ec.cmnd_pid);
+ /* Close the other end of the pipes and socketpairs. */
+ if (io_pipe[STDIN_FILENO][0] != -1)
+ close(io_pipe[STDIN_FILENO][0]);
+ if (io_pipe[STDOUT_FILENO][1] != -1)
+ close(io_pipe[STDOUT_FILENO][1]);
+ if (io_pipe[STDERR_FILENO][1] != -1)
+ close(io_pipe[STDERR_FILENO][1]);
+ close(errpipe[1]);
+ if (intercept_sv[1] != -1)
+ close(intercept_sv[1]);
+
+ /* No longer need execfd. */
+ if (details->execfd != -1) {
+ close(details->execfd);
+ details->execfd = -1;
+ }
+
+ /* Set command timeout if specified. */
+ if (ISSET(details->flags, CD_SET_TIMEOUT))
+ alarm(details->timeout);
+
+ /*
+ * Fill in exec closure, allocate event base, signal events and
+ * the error pipe event.
+ */
+ fill_exec_closure(&ec, cstat, details, user_details, evbase, errpipe[0]);
+
+ if (ISSET(details->flags, CD_INTERCEPT|CD_LOG_SUBCMDS)) {
+ int rc = 1;
+
+ /* Create event and closure for intercept mode. */
+ ec.intercept = intercept_setup(intercept_sv[0], ec.evbase, details);
+ if (ec.intercept == NULL) {
+ rc = -1;
+ } else if (ISSET(details->flags, CD_USE_PTRACE)) {
+ /* Try to seize control of the command using ptrace(2). */
+ rc = exec_ptrace_seize(ec.cmnd_pid);
+ if (rc == 0) {
+ /* There is another tracer present. */
+ CLR(details->flags, CD_INTERCEPT|CD_LOG_SUBCMDS|CD_USE_PTRACE);
+ }
+ }
+ if (rc == -1)
+ terminate_command(ec.cmnd_pid, false);
+ }
+
+ /* Enable any I/O log events. */
+ add_io_events(&ec);
+
+ /* Restore signal mask now that signal handlers are setup. */
+ sigprocmask(SIG_SETMASK, &oset, NULL);
+
+ /*
+ * Non-pty event loop.
+ * Wait for command to exit, handles signals and the error pipe.
+ */
+ if (sudo_ev_dispatch(ec.evbase) == -1)
+ sudo_warn("%s", U_("error in event loop"));
+ if (sudo_ev_got_break(ec.evbase)) {
+ /* error from callback */
+ sudo_debug_printf(SUDO_DEBUG_ERROR, "event loop exited prematurely");
+ /* kill command */
+ terminate_command(ec.cmnd_pid, false);
+ ec.cmnd_pid = -1;
+ }
+
+#ifdef HAVE_SELINUX
+ if (ISSET(details->flags, CD_RBAC_ENABLED)) {
+ if (selinux_restore_tty() != 0)
+ sudo_warnx("%s", U_("unable to restore tty label"));
+ }
+#endif
+
+ /* Flush any remaining output. */
+ del_io_events(true);
+
+ /* Free things up. */
+ free_io_bufs();
+ free_exec_closure(&ec);
+
+ debug_return;
+}
+
+/*
+ * Wait for command status after receiving SIGCHLD.
+ * If the command exits, fill in cstat and stop the event loop.
+ * If the command stops, save the tty pgrp, suspend sudo, then restore
+ * the tty pgrp when sudo resumes.
+ */
+static void
+handle_sigchld_nopty(struct exec_closure *ec)
+{
+ pid_t pid;
+ int status;
+ char signame[SIG2STR_MAX];
+ debug_decl(handle_sigchld_nopty, SUDO_DEBUG_EXEC);
+
+ /* There may be multiple children in intercept mode. */
+ for (;;) {
+ do {
+ pid = waitpid(-1, &status, __WALL|WUNTRACED|WNOHANG);
+ } while (pid == -1 && errno == EINTR);
+ switch (pid) {
+ case -1:
+ if (errno != ECHILD) {
+ sudo_warn(U_("%s: %s"), __func__, "waitpid");
+ debug_return;
+ }
+ FALLTHROUGH;
+ case 0:
+ /* Nothing left to wait for. */
+ debug_return;
+ }
+
+ if (WIFSTOPPED(status)) {
+ const int signo = WSTOPSIG(status);
+
+ if (sig2str(signo, signame) == -1)
+ (void)snprintf(signame, sizeof(signame), "%d", signo);
+ sudo_debug_printf(SUDO_DEBUG_INFO,
+ "%s: process %d stopped, SIG%s", __func__, (int)pid, signame);
+
+ if (ISSET(ec->details->flags, CD_USE_PTRACE)) {
+ /* If not a group-stop signal, just continue. */
+ if (!exec_ptrace_stopped(pid, status, ec->intercept))
+ continue;
+ }
+
+ /* If the main command is suspended, suspend sudo too. */
+ if (pid == ec->cmnd_pid) {
+ sudo_suspend_parent(signo, ec->sudo_pid, ec->ppgrp,
+ ec->cmnd_pid, ec, log_suspend);
+ }
+ } else {
+ if (WIFSIGNALED(status)) {
+ if (sig2str(WTERMSIG(status), signame) == -1) {
+ (void)snprintf(signame, sizeof(signame), "%d",
+ WTERMSIG(status));
+ }
+ sudo_debug_printf(SUDO_DEBUG_INFO,
+ "%s: process %d killed, SIG%s", __func__,
+ (int)pid, signame);
+ } else if (WIFEXITED(status)) {
+ sudo_debug_printf(SUDO_DEBUG_INFO,
+ "%s: process %d exited: %d", __func__,
+ (int)pid, WEXITSTATUS(status));
+ } else {
+ sudo_debug_printf(SUDO_DEBUG_WARN,
+ "%s: unexpected wait status 0x%x for process %d",
+ __func__, status, (int)pid);
+ }
+
+ /* Only store exit status of the main command. */
+ if (pid != ec->cmnd_pid)
+ continue;
+
+ /* Don't overwrite execve() failure with command exit status. */
+ if (ec->cstat->type == CMD_INVALID) {
+ ec->cstat->type = CMD_WSTATUS;
+ ec->cstat->val = status;
+ } else {
+ sudo_debug_printf(SUDO_DEBUG_WARN,
+ "%s: not overwriting command status %d,%d with %d,%d",
+ __func__, ec->cstat->type, ec->cstat->val,
+ CMD_WSTATUS, status);
+ }
+ ec->cmnd_pid = -1;
+ }
+ }
+}