Pseudo-terminals, and what tmux and htop needed around them

- /dev/ptmx and /dev/pts/N: Unix98 pairs with a full line discipline,
  window sizes, job control and hangup (kernel/dev/pty.c).
- The console speaks scroll regions, insert/delete line, SU/SD, RI/IND/NEL,
  DECSC/DECRC; swallows charset designations and private CSI sequences;
  and LF only returns the carriage under ONLCR.
- setsid() drops the controlling terminal, so a pane's pty becomes its ctty.
- pipe2() honours O_NONBLOCK and O_CLOEXEC (libevent's signal pipe).
- File modes are stored and enforced: chmod, fchmod(at), umask, creation
  modes; /tmp is 1777.
- /proc/stat has real user/system/idle; /proc/<pid>/task; kthread flag.
- SO_PEERCRED, IOV_MAX 1024.
- /proc/strace: name a program and its syscalls go to the kernel log.
- World gains htop, btop, tmux, mc, ncdu, fastfetch, games, X toys, NetSurf.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01S1giNvrHTrPMPw7V6x1kQ4

kernel/dev/procfs.c

@@ -194,16 +194,36 @@ static void
 gen_stat(struct sbuf *sb)
 {
   // USER_HZ, like times() -- everything that parses /proc/stat divides by 100.
-  uint t = (uint)((uint64)uptime() * USER_HZ / HZ);
+  //
+  // Busy is what processes were charged; idle is the rest of every online
+  // processor's time. Calling it all user time, as this did, is a machine at
+  // 100% forever in htop and top. The charge is not kept per processor, so
+  // each cpuN line is an equal share of the total -- right on average, which
+  // is all a meter can show anyway.
+  uint64 n = ncpu_online > 0 ? ncpu_online : 1;
+  uint64 all = (uint64)uptime() * USER_HZ / HZ * n;
+  uint64 user, sys;
+  proc_cputime(&user, &sys);
+  user = user * USER_HZ / HZ;
+  sys  = sys * USER_HZ / HZ;
+  if (user + sys > all)
+    all = user + sys;
+  uint64 idle = all - user - sys;
+
+  // user nice system idle iowait irq softirq steal guest guest_nice
   sb_puts(sb, "cpu  ");
-  sb_putu(sb, t); sb_puts(sb, " 0 0 0 0 0 0 0 0 0\n");
+  sb_putu(sb, user); sb_puts(sb, " 0 ");
+  sb_putu(sb, sys);  sb_putc(sb, ' ');
+  sb_putu(sb, idle); sb_puts(sb, " 0 0 0 0 0 0\n");
   // Online, for the same reason as cpuinfo: top and friends count the cpuN
   // lines here to decide how many cores to draw.
   for (int i = 0; i < ncpu_online; i++) {
     sb_puts(sb, "cpu");
     sb_putu(sb, i);
     sb_putc(sb, ' ');
-    sb_putu(sb, t); sb_puts(sb, " 0 0 0 0 0 0 0 0 0\n");
+    sb_putu(sb, user / n); sb_puts(sb, " 0 ");
+    sb_putu(sb, sys / n);  sb_putc(sb, ' ');
+    sb_putu(sb, idle / n); sb_puts(sb, " 0 0 0 0 0 0\n");
   }
   sb_kv(sb, "processes ", (uint64)proc_num(), 0);
   sb_kv(sb, "procs_running ", 1, 0);
@@ -388,12 +408,42 @@ gen_filesystems(struct sbuf *sb)
 #define PROC_BOOTDEBUG    30
 #define PROC_SLAB         20
 #define PROC_PARTITIONS   21
+#define PROC_STRACE       31
 
 // Render global file `id` (a PROC_* constant) into `buf`. Returns its length.
 //
 // Split out of procread() so the /proc *filesystem* (kernel/fs/procfs.c) can
 // use the same generators: that file owns the namespace, this one owns the
 // content, and they have no other business with each other.
+// /proc/strace: the name of the program to trace, or empty. Read by exec, so
+// a change takes effect at the next program started, not in running ones.
+static char strace_name[16];
+
+int
+strace_wanted(const char *name)
+{
+  return strace_name[0] && strncmp(name, strace_name, sizeof(strace_name)) == 0;
+}
+
+int
+proc_write_strace(int user_src, uint64 src, uint n)
+{
+  char tmp[16];
+  uint len = n < sizeof(tmp) - 1 ? n : sizeof(tmp) - 1;
+
+  if (user_src) {
+    if (copy_from_user(tmp, src, len) < 0)
+      return -EFAULT;
+  } else {
+    memmove(tmp, (char *)src, len);
+  }
+  while (len > 0 && (tmp[len - 1] == '\n' || tmp[len - 1] == ' '))
+    len--;
+  tmp[len] = 0;
+  memmove(strace_name, tmp, sizeof(strace_name));
+  return (int)n;
+}
+
 int
 proc_gen_global(int id, char *buf, int cap)
 {
@@ -442,6 +492,10 @@ proc_gen_global(int id, char *buf, int cap)
     sb_puts(&sb, boot_debug ? "on\n" : "off\n");
     break;
   case PROC_DEVICES:     gen_devices(&sb);     break;
+  case PROC_STRACE:
+    sb_puts(&sb, strace_name);
+    sb_putc(&sb, '\n');
+    break;
   // The block layer renders its own table, like PCI and SchFS above.
   case PROC_PARTITIONS:  sb.len = blkdev_info(buf, cap); break;
   default:               return 0;