kernel: HTTPS through wget, half-close, and /proc/<pid>/fd

busybox wget hands TLS to ssl_client over a socketpair and then does
shutdown(SHUT_WR). That cut the AF_UNIX link both ways, so ssl_client's
reply was a SIGPIPE and every https:// download hung. SHUT_WR is now a
half-close: the peer reads EOF and can still answer. TCP sockets get
shutdown() at all (tcp_shutdown for the FIN); it used to do nothing.

- exec keeps ignored signals and the signal mask, as execve(2) says.
- /proc/<pid>/fd: a link per descriptor -- the path it was opened by,
  socket:[n], pipe:[n], anon_inode:[eventpoll]. /proc/self too. The
  global inode bit moves above pid << 8; at 0x10000 it was /proc/256.
- fstat on a socket or pipe answers S_IFSOCK / S_IFIFO instead of EINVAL.
- poll, select and epoll wake as soon as a pipe, socket or input device
  changes, not on the next tick.
- Pipes are 64 KB of lazily allocated pages, copied in blocks, with
  PIPE_BUF atomicity: 300 MB through `yes | head | wc` in 7 s.

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

include/core/defs.h

@@ -534,6 +534,11 @@ int             inet_sendto(void *h, char *buf, int n,
                             uint addr, ushort port);
 int             inet_readable(void *h);
 int             inet_hup(void *h);
+int             inet_shutdown(void *h, int how);
+int             file_describe(struct file *f, char *buf, int cap);
+int             proc_fd_mask(int pid, uint64 mask[2]);
+int             proc_fd_link(int pid, int fd, char *buf, int cap);
+const char     *ev_kind_name(struct evobj *o);
 void            inet_close(void *h);
 
 // proc.c
@@ -552,6 +557,8 @@ int             sigsend(int pid, int sig);
 void            alarm_arm(struct proc *p, uint after, uint every);
 uint            alarm_left(struct proc *p);
 void            alarm_thread_start(void);
+void            poll_kick(void);
+int             poll_wait_tick(void);
 int             sig_default_kills(int sig);
 int             proc_exists(int);
 int             proc_num(void);

include/fs/file.h

@@ -25,6 +25,11 @@ struct file {
   int memfd_seg;
   struct evobj *ev;
   uint64 off;
+  // What /proc/<pid>/fd shows. `ino` names the open thing (one per socket,
+  // one per pipe); `path` is the absolute path an inode was opened by, as
+  // written -- symlinks along it are not resolved. kmalloc'd, or 0.
+  uint64 ino;
+  char *path;
   struct file *fnext;   // file.c's free list, while ref is 0
 };
 

include/net/socket.h

@@ -330,6 +330,9 @@ struct socket {
   // bidirectional connection.
   struct socket *peer;
   int   peer_closed;
+  // shutdown(SHUT_WR): this end has said all it will. The peer reads EOF but
+  // can still answer, which is what a request/response helper depends on.
+  int   wr_shut;
 
   char  data[SOCKBUF];
   uint  nread, nwrite;