// SPDX-License-Identifier: GPL-2.0-or-later
//
// schinit -- PID 1.
//
// This program used to be the whole init system: it parsed manifests, walked
// /etc/schinit, decided respawn policy, and supervised every job. All of that
// now lives in schinitd (PID 2), and what is left here is deliberately, almost
// aggressively small.
//
// The reason is docs/architecture/init.md, and it is worth restating because the
// smallness looks like an accident otherwise:
//
//   **PID 1 is the one program on the machine with no supervisor.** Nothing
//   restarts it, nothing notices it hang, and until proc_reincarnate() lands
//   its death is a panic. A program in that position should not contain a
//   parser, should not walk a filesystem, and should not allocate. So it does
//   none of those things. Its entire job is to be too boring to fail.
//
// What it does:
//
//   - bootstrap the filesystems, because something must and it happens first
//   - fork schinitd and restart it if it dies
//   - fork jobs *on schinitd's behalf*, so every job is a child of PID 1
//   - reap children and say so
//
// That third one is the load-bearing one. schinitd decides and asks; PID 1
// acts. So schinitd crashing orphans nothing, reparents nothing and restarts
// nothing -- it comes back, reads /proc/schinit, and finds the world exactly
// as it left it. A supervisor that forked its own jobs could not have that.

#include "types.h"
#include "user.h"
#include "fcntl.h"
#include "posix.h"
#include "socket.h"
#include "schinit.h"

#define SUPERVISOR  "/bin/schinitd"

// Jobs PID 1 is holding for schinitd. This is a table of pids and labels, not
// a table of policy: PID 1 does not know what a job *is*, only that it forked
// one and what to call it in /proc/schinit.
#define MAXJOBS 32

struct child {
    int  pid;                          // 0 = free slot
    unsigned int cookie;               // schinitd's handle for it
    char label[SCHINIT_MAXLABEL];
    int  started;                      // uptime() ticks
};

static struct child children[MAXJOBS];
static int  sock = -1;                 // to schinitd
static int  supervisor_pid;
static int  last_pong;                 // uptime() when schinitd last answered
static int  last_ping;                 // uptime() when we last asked
static unsigned int generation = 1;    // bumped by proc_reincarnate() later

static int
slen(const char *s)
{
    int n = 0;
    while (s[n]) n++;
    return n;
}

static void
scopy(char *dst, const char *src, int cap)
{
    int i = 0;
    while (src[i] && i < cap - 1) { dst[i] = src[i]; i++; }
    dst[i] = '\0';
}

static int
seq(const char *a, const char *b)
{
    while (*a && *a == *b) { a++; b++; }
    return *a == *b;
}

// Case-insensitive, so a manifest may say KeepAlive, keepalive or KEEPALIVE.
// Keys are for humans and this is not a language.
static int
sieq(const char *a, const char *b)
{
    for (;;) {
        char ca = *a, cb = *b;
        if (ca >= 'A' && ca <= 'Z') ca += 32;
        if (cb >= 'A' && cb <= 'Z') cb += 32;
        if (ca != cb) return 0;
        if (!ca) return 1;
        a++; b++;
    }
}

static int
ends_with(const char *s, const char *suffix)
{
    int ls = slen(s), lf = slen(suffix);
    if (lf > ls) return 0;
    return seq(s + ls - lf, suffix);
}

// How the machine was booted; see /proc/bootmode and kinfo() in the kernel.
static int verbose_boot;

static void
read_bootmode(void)
{
    char b[16];
    int fd = open("/proc/bootmode", O_RDONLY);
    int n;

    if (fd < 0)
        return;
    n = read(fd, b, sizeof(b) - 1);
    close(fd);
    if (n > 0) { b[n] = 0; verbose_boot = (b[0] == 'v'); }
}

static void
out(const char *s)
{
    write(1, (void *)s, slen(s));
}

static void
outn(int v)
{
    char buf[12];
    int i = 11;
    int neg = v < 0;
    if (neg) v = -v;
    buf[i--] = '\0';
    if (v == 0) buf[i--] = '0';
    while (v > 0) { buf[i--] = (char)('0' + v % 10); v /= 10; }
    if (neg) buf[i--] = '-';
    out(&buf[i + 1]);
}

// ---------------------------------------------------------------------------
// Jobs
// ---------------------------------------------------------------------------

// Two formatters, because publish() needs them and pulling in a printf would
// put a format-string parser inside PID 1.
static int
fmt_str(char *dst, const char *s)
{
    int n = 0;
    while (s[n]) { dst[n] = s[n]; n++; }
    return n;
}

static int
fmt_num(char *dst, int v)
{
    char tmp[12];
    int  i = 0, n = 0;
    if (v < 0) { dst[n++] = '-'; v = -v; }
    if (v == 0) { dst[n++] = '0'; return n; }
    while (v > 0) { tmp[i++] = '0' + (v % 10); v /= 10; }
    while (i-- > 0) dst[n++] = tmp[i];
    return n;
}

static void
bootstrap(void)
{
    // /dev before the console, because the console now lives in it.
    //
    // The old order was the other way round -- open a console first so every
    // message below had somewhere to go -- and the fallback below is what keeps
    // that guarantee. If the mount fails there is no /dev/console and no way to
    // say so, so the original path stays: a node called `console` in the root
    // directory, which is where this used to put one.
    //
    // /dev is a filesystem now, not a pile of nodes. Every major and minor it
    // contains is the kernel's -- devsw[] in kernel/dev/cdev.c is what they
    // dispatch on -- and this program used to hold a second copy of that
    // mapping with nothing checking the two agreed.
    //
    // It was also slow in a way that only showed on a fresh image: each mknod
    // was one SchFS transaction, and the twenty of them cost 2600 ms of the
    // first boot before this program printed its first line. On a second boot
    // the nodes already existed and the same code took 0 ms, so the cost was
    // invisible to anyone not building images -- which is to say, invisible to
    // everyone except whoever works on this.
    mkdir("/dev");
    mount("dev", "/dev", "devtmpfs", 0, 0);

    if (open("/dev/console", O_RDWR) < 0) {
        // No devtmpfs. Fall back to a real node on the root filesystem, which
        // is what every boot did before this.
        if (open("console", O_RDWR) < 0) {
            mknod("console", 1, 1);
            open("console", O_RDWR);
        }
    }
    dup(0);
    dup(0);

    // /proc is a real filesystem (kernel/fs/procfs.c), not a handful of nodes:
    // it has to be, because one directory per process cannot be created ahead
    // of time. That is what makes ps and top work.
    //
    // The thirteen mknod() calls that used to follow this mount are gone. They
    // ran *after* it, so every one of them was resolving a path into a
    // read-only synthetic filesystem and failing -- dead weight kept as a
    // fallback for a mount that has not failed in a long time, exactly as the
    // note in procfs.c said it was.
    mkdir("/proc");
    mount("proc", "/proc", "proc", 0, 0);

    // Mount points, and nothing mounts on them yet. They are created by the
    // image builder (see sysroot-install), so these are a fallback for an image
    // that does not have them -- a mkdir that fails because the directory is
    // already there costs a path walk and no writes, where creating one costs a
    // whole SchFS transaction.
    mkdir("/sys");
    mkdir("/run");

    // The hardware, at Linux's paths, which is where lspci, lsusb and
    // hwreport look for it. See kernel/fs/sysfs.c.
    mount("sysfs", "/sys", "sysfs", 0, 0);
}

// ---------------------------------------------------------------------------
// The job table, for /proc/schinit
// ---------------------------------------------------------------------------

// Publish what PID 1 is holding.
//
// docs/architecture/init.md, "State outside the process": it lives outside
// whichever process is
// supervising, so a restarted supervisor can read the world back instead of
// rebuilding it. It is written by PID 1 because PID 1 is the process that
// cannot lose it.
//
// One line per running job: cookie, pid, start tick, label. Written whenever
// the table changes rather than on read, because a procfs generator runs in
// the kernel and the kernel has no business walking this program's memory.
static void
publish(void)
{
    char buf[MAXJOBS * (SCHINIT_MAXLABEL + 40) + 64];
    int  n = 0;

    // Hand-rolled rather than snprintf, for the same reason as everything else
    // here: no allocation, no format parser, no libc surprise in PID 1.
    n += fmt_str(buf + n, "generation ");
    n += fmt_num(buf + n, (int)generation);
    n += fmt_str(buf + n, "\nsupervisor ");
    n += fmt_num(buf + n, supervisor_pid);
    n += fmt_str(buf + n, "\n");

    for (int i = 0; i < MAXJOBS; i++) {
        if (!children[i].pid)
            continue;
        n += fmt_num(buf + n, (int)children[i].cookie);
        n += fmt_str(buf + n, " ");
        n += fmt_num(buf + n, children[i].pid);
        n += fmt_str(buf + n, " ");
        n += fmt_num(buf + n, children[i].started);
        n += fmt_str(buf + n, " ");
        n += fmt_str(buf + n, children[i].label);
        n += fmt_str(buf + n, "\n");
    }

    int fd = open("/proc/schinit", O_WRONLY);
    if (fd < 0)
        return;                        // no procfs yet, or it is read-only
    write(fd, buf, n);
    close(fd);
}

// Read the table back after a reincarnation.
//
// This is the one parse in PID 1, and it is worth being uneasy about. It is
// bounded: a fixed buffer, four integers and a label per line, no allocation,
// and it reads a format this same program wrote. It happens once, at startup,
// only when the kernel has just re-exec'd us. The alternative is forgetting
// every running job, which is worse than a twenty-line scan of known text.
static void
reclaim(void)
{
    char buf[MAXJOBS * (SCHINIT_MAXLABEL + 40) + 64];
    int fd = open("/proc/schinit", O_RDONLY);
    int n, slot = 0;

    if (fd < 0)
        return;
    n = read(fd, buf, (int)sizeof(buf) - 1);
    close(fd);
    if (n <= 0)
        return;
    buf[n] = 0;

    char *p = buf;
    while (*p && slot < MAXJOBS) {
        char *line = p;
        while (*p && *p != '\n') p++;
        if (*p == '\n') *p++ = 0;

        unsigned int cookie = 0;
        int pid = 0, started = 0, digits = 0;
        char *q = line;

        if (*q < '0' || *q > '9') continue;       // a header line
        while (*q >= '0' && *q <= '9') { cookie = cookie * 10 + (unsigned)(*q++ - '0'); digits++; }
        while (*q == ' ') q++;
        while (*q >= '0' && *q <= '9') pid = pid * 10 + (*q++ - '0');
        while (*q == ' ') q++;
        while (*q >= '0' && *q <= '9') started = started * 10 + (*q++ - '0');
        while (*q == ' ') q++;
        if (!digits || pid <= 0 || !*q) continue;

        children[slot].pid     = pid;
        children[slot].cookie  = cookie;
        children[slot].started = started;
        scopy(children[slot].label, q, SCHINIT_MAXLABEL);
        slot++;
    }
    if (slot) {
        out("schinit: reclaimed ");
        outn(slot);
        out(" running job(s) from /proc/schinit\n");
    }
}

// ---------------------------------------------------------------------------
// Spawning, on request
// ---------------------------------------------------------------------------

static void
redirect(const char *path, int fd, int flags)
{
    if (!path || !path[0])
        return;
    close(fd);
    if (open((char *)path, flags) != fd) {
        // Could not put it on the right descriptor. Say nothing -- this runs
        // in the child, between fork and exec, and the console may be exactly
        // the thing that failed to open.
    }
}

// Fork and exec one job. Returns the pid, or a negative number.
//
// Everything this needs arrives in the message. It does not read a file, does
// not look at a manifest, and does not consult any policy: schinitd worked all
// of that out and sent the answer.
static int
spawn(struct schinit_spawn *s)
{
    // Rebuild argv and envp from the flat buffer. Pointers cannot cross a
    // process boundary, so the message carries offsets and they are turned
    // back into pointers here, on this stack, where they are valid.
    char *argv[SCHINIT_MAXARGS + 1];
    char *envp[SCHINIT_MAXENVV + 1];
    unsigned int na = s->nargs, ne = s->nenvs;

    if (na > SCHINIT_MAXARGS) na = SCHINIT_MAXARGS;
    if (ne > SCHINIT_MAXENVV) ne = SCHINIT_MAXENVV;

    // A hostile or broken offset must not walk off the buffer. The strings are
    // NUL-terminated inside strbuf and the last byte is forced to NUL below,
    // so a bad offset yields a short string rather than a read past the end.
    s->strbuf[SCHINIT_STRBUF - 1] = 0;
    s->program[SCHINIT_MAXPATH - 1] = 0;
    s->label[SCHINIT_MAXLABEL - 1] = 0;

    for (unsigned int i = 0; i < na; i++) {
        unsigned short off = s->args[i];
        argv[i] = (off < SCHINIT_STRBUF) ? &s->strbuf[off] : (char *)"";
    }
    argv[na] = 0;
    for (unsigned int i = 0; i < ne; i++) {
        unsigned short off = s->envs[i];
        envp[i] = (off < SCHINIT_STRBUF) ? &s->strbuf[off] : (char *)"";
    }
    envp[ne] = 0;

    if (!argv[0])
        argv[0] = s->program;

    int pid = fork();
    if (pid < 0)
        return -1;

    if (pid == 0) {
        // A session of its own, so terminal input reaches it. A session with
        // no controlling terminal is not delivered terminal input at all --
        // without this the shell echoes and executes nothing.
        setsid();

        redirect(s->stdin_path,  0, O_RDWR);
        redirect(s->stdout_path, 1, O_RDWR);
        redirect(s->stderr_path, 2, O_RDWR);

        if (s->own_tty)
            ioctl(0, TIOCSCTTY, 0);

        // The socket to schinitd is PID 1's and nobody else's.
        close(sock);

        execve(s->program, argv, ne ? envp : 0);

        out("schinit: cannot exec ");
        out(s->program);
        out("\n");
        exit();
    }
    return pid;
}

// ---------------------------------------------------------------------------
// The supervisor
// ---------------------------------------------------------------------------

static int
start_supervisor(void)
{
    int sv[2];

    if (socketpair(AF_UNIX, SOCK_STREAM, 0, sv) < 0) {
        out("schinit: socketpair failed\n");
        return -1;
    }

    int pid = fork();
    if (pid < 0) {
        close(sv[0]);
        close(sv[1]);
        return -1;
    }

    if (pid == 0) {
        close(sv[0]);
        // A fixed descriptor rather than an argument: schinitd is exec'd, so
        // there is nowhere to pass a number, and a well-known one is one less
        // thing for the two of them to disagree about.
        if (sv[1] != SCHINIT_SOCK_FD) {
            close(SCHINIT_SOCK_FD);
            dup2(sv[1], SCHINIT_SOCK_FD);
            close(sv[1]);
        }
        char *argv[] = { (char *)"schinitd", 0 };
        execve((char *)SUPERVISOR, argv, 0);
        out("schinit: cannot exec " SUPERVISOR "\n");
        exit();
    }

    close(sv[1]);
    sock = sv[0];
    supervisor_pid = pid;
    last_pong = last_ping = uptime();   // it has not missed anything yet
    return pid;
}

static void
send_reply(unsigned int op, unsigned int cookie, int pid, int status)
{
    struct schinit_reply r;
    r.op     = op;
    r.cookie = cookie;
    r.pid    = pid;
    r.status = status;
    // A supervisor that has stopped reading must not wedge PID 1. The socket
    // buffer is far larger than anything this sends in a burst, and if the
    // write fails the supervisor is already lost -- it will be restarted, and
    // it will read /proc/schinit rather than this message.
    write(sock, &r, sizeof(r));
}

// One message from schinitd. Returns 0 if the socket has closed.
static int
serve(void)
{
    union schinit_msg m;
    int n = read(sock, &m, sizeof(m));

    if (n <= 0)
        return 0;                      // supervisor gone

    switch (m.op) {
    case SCHINIT_HELLO: {
        struct schinit_hello h;
        h.op         = SCHINIT_WELCOME;
        h.proto      = SCHINIT_PROTO;
        h.generation = generation;
        h.reserved   = 0;
        write(sock, &h, sizeof(h));
        break;
    }

    case SCHINIT_PONG:
        last_pong = uptime();
        break;

    case SCHINIT_SPAWN: {
        int pid = spawn(&m.spawn);
        if (pid > 0) {
            for (int i = 0; i < MAXJOBS; i++) {
                if (children[i].pid)
                    continue;
                children[i].pid     = pid;
                children[i].cookie  = m.spawn.cookie;
                children[i].started = uptime();
                scopy(children[i].label, m.spawn.label, SCHINIT_MAXLABEL);
                break;
            }
            publish();
        }
        send_reply(SCHINIT_SPAWNED, m.spawn.cookie, pid, 0);
        break;
    }

    default:
        // An op PID 1 does not know. Ignore it rather than guess: this is the
        // only process that cannot afford to act on a message it misread.
        break;
    }
    return 1;
}

// ---------------------------------------------------------------------------

int
main(int argc, char *argv[])
{
    // argv[1] is the generation, and it is only there when the kernel put it
    // there -- proc_reincarnate() re-execs this program in place after PID 1
    // has died. A normal boot has no argument and is generation 1.
    if (argc > 1) {
        unsigned int g = 0;
        for (char *q = argv[1]; *q >= '0' && *q <= '9'; q++)
            g = g * 10 + (unsigned int)(*q - '0');
        if (g)
            generation = g;
    }

    bootstrap();

    // Only now: /proc is what bootstrap() mounts, and this reads a file in it.
    // Before the mount it silently returned "quiet" every time, which made the
    // verbose boot quiet in exactly the one process that runs first.
    read_bootmode();

    // A reincarnated PID 1 has the children but not the table.
    //
    // exec() replaced this program's memory and kept everything else -- the
    // pid, the open files, and crucially the *children*, which are still
    // parented here because nothing killed them. What did not survive is the
    // array describing them, so it is read back from where it was published
    // precisely so that it could be.
    //
    // Without this the next publish() overwrites the record with an empty
    // table, the supervisor reads that, believes nothing is running and starts
    // everything a second time. Same split brain as before, one level up.
    if (generation > 1) {
        // exec() keeps open files, which is what lets this program still have
        // a console to complain to -- and also means the previous life's
        // socket to the old supervisor is still open here. Leaving it open
        // holds the old supervisor's channel up, so it never sees a hangup,
        // never exits, and sits there reporting that PID 1 has stopped
        // pinging it. Which is true, and no longer its business.
        for (int fd = 3; fd < 32; fd++)
            close(fd);
        reclaim();
    }

    // Quiet unless asked, or unless this is not the first pid 1.
    //
    // "SchOS process 1" is a line about the system working exactly as designed,
    // which is the definition of what a quiet boot leaves out. A *reincarnated*
    // pid 1 is the opposite -- it means the last one died and was replaced --
    // so that one prints either way.
    if (verbose_boot || generation > 1) {
        out("schinit: SchOS process 1");
        if (generation > 1) {
            out(", generation ");
            outn((int)generation);
            out(" (reincarnated)");
        }
        out("\n");
    }

    // Generation 4 is the safety net: a manifest that kills init must not kill
    // it for ever, so this generation does not read /etc/schinit at all. One
    // shell, on the console, and nothing else -- enough to log in and fix
    // whatever has been taking the machine down.
    if (generation >= 4) {
        out("schinit: safe mode -- console shell only, /etc/schinit ignored\n");
        if (fork() == 0) {
            setsid();
            ioctl(0, TIOCSCTTY, 0);
            char *sh[] = { (char *)"-zsh", 0 };
            execve((char *)"/bin/zsh", sh, 0);
            char *fb[] = { (char *)"sh", 0 };
            execve((char *)"/bin/sh", fb, 0);
            exit();
        }
        publish();
        for (;;) {
            int st, pid;
            while ((pid = waitpid(-1, &st, WNOHANG)) > 0)
                ;
            sleep(20);
        }
    }

    if (start_supervisor() < 0) {
        // No supervisor. Rather than idle usefully-looking, fall back to a
        // shell so the machine is still usable and the failure is visible.
        out("schinit: no supervisor; falling back to a shell\n");
        if (fork() == 0) {
            setsid();
            char *argv[] = { (char *)"sh", 0 };
            execve((char *)"/bin/sh", argv, 0);
            exit();
        }
    }

    publish();
    if (verbose_boot)
        out("schinit: ready (pid 1)\n");

    for (;;) {
        // Two things can happen -- a request arrives, or a child exits -- and
        // both of them block. poll() with a timeout plus WNOHANG turns that
        // into one loop with no signal handler, which is the simplest thing
        // that works and therefore the right thing for this program.
        struct pollfd pf;
        pf.fd      = sock;
        pf.events  = POLLIN;
        pf.revents = 0;

        if (sock >= 0 && poll(&pf, 1, 200) > 0 && (pf.revents & POLLIN)) {
            if (!serve()) {
                close(sock);
                sock = -1;
            }
        } else if (sock < 0) {
            sleep(10);
        }

        // The heartbeat. A supervisor that has stopped answering is stuck,
        // and a stuck supervisor is indistinguishable from a working one by
        // any means except asking it something -- which is the whole reason
        // this exists. See docs/architecture/init.md, "The heartbeat".
        if (sock >= 0) {
            int now = uptime();
            if (now - last_ping >= SCHINIT_PING_SECS * 100) {
                struct schinit_hello p;
                p.op = SCHINIT_PING;
                p.proto = SCHINIT_PROTO;
                p.generation = generation;
                p.reserved = 0;
                write(sock, &p, sizeof(p));
                last_ping = now;
            }
            if (now - last_pong >= SCHINIT_DEADLINE_SECS * 100) {
                out("schinit: supervisor stopped answering; replacing it\n");

                // Ask it to die, then replace it *without waiting*.
                //
                // kill() sets a flag the target checks the next time it runs,
                // and the whole premise here is that it is not running. A
                // supervisor stuck in a loop will notice at its next trap; one
                // blocked forever never will. Waiting for a corpse that may
                // never be produced is how a recovery path becomes a second
                // hang -- which is exactly what the first version of this did.
                //
                // Two supervisors briefly existing is safe, and safe in the
                // way that matters: the old one's socket is closed here, so
                // the moment it runs again its first write fails and it exits.
                // It cannot spawn anything, because it cannot ask.
                kill(supervisor_pid);
                close(sock);
                sock = -1;
                int old = supervisor_pid;
                if (start_supervisor() > 0)
                    out("schinit: new supervisor started\n");
                publish();             // the table names the live supervisor
                (void)old;
            }
        }

        int status;
        int pid;
        while ((pid = waitpid(-1, &status, WNOHANG)) > 0) {
            if (pid == supervisor_pid) {
                out("schinit: supervisor died; restarting\n");
                if (sock >= 0) { close(sock); sock = -1; }
                supervisor_pid = 0;
                // The jobs are untouched. They are children of this process,
                // not of the supervisor, so nothing was orphaned and nothing
                // needs restarting -- which is the entire reason jobs are
                // forked here. The new supervisor reads /proc/schinit and
                // picks up where the old one stopped.
                sleep(100);            // one second, so a crash loop is slow
                start_supervisor();
                publish();
                continue;
            }

            for (int i = 0; i < MAXJOBS; i++) {
                if (children[i].pid != pid)
                    continue;
                unsigned int cookie = children[i].cookie;
                children[i].pid = 0;
                children[i].label[0] = 0;
                publish();
                if (sock >= 0)
                    send_reply(SCHINIT_EXITED, cookie, pid, status);
                break;
            }
            // A pid in no slot is an orphan the kernel reparented here. It has
            // been reaped, which was the whole obligation.
        }
    }
}