fix(sendfile): handle EAGAIN on macOS non-blocking sockets (#185)

lib/posix.ml

@@ -49,7 +49,7 @@ let sendfile
      * `Gluten.Server.report_write_result` here given that we put
      * bytes in the file descriptor off-band. `report_write_result`
      * just tracks the internal Faraday buffers. *)
-    Logs.debug (fun m -> m "`sendfile` wrote %d bytes successfully" sent);
+    Logs.debug (fun m -> m "`sendfile` wrote %d bytes" sent);
     Ok ()
   | Error e ->
     Writer.close raw_write_body;

sendfile/sendfile.ml

@@ -1,3 +1,90 @@
+exception Darwin_specific_unix_error of (Unix.error * int)
+
+let () =
+  Callback.register_exception
+    "sendfile_exn_unix_error"
+    (Darwin_specific_unix_error (Unix.E2BIG, 0));
+  Printexc.register_printer (function
+    | Darwin_specific_unix_error (unix_error, len) ->
+      let msg =
+        (* Copied from
+           https://github.com/ocaml/ocaml/blob/trunk/otherlibs/unix/unix_unix.ml *)
+        match unix_error with
+        | E2BIG -> "E2BIG"
+        | EACCES -> "EACCES"
+        | EAGAIN -> "EAGAIN"
+        | EBADF -> "EBADF"
+        | EBUSY -> "EBUSY"
+        | ECHILD -> "ECHILD"
+        | EDEADLK -> "EDEADLK"
+        | EDOM -> "EDOM"
+        | EEXIST -> "EEXIST"
+        | EFAULT -> "EFAULT"
+        | EFBIG -> "EFBIG"
+        | EINTR -> "EINTR"
+        | EINVAL -> "EINVAL"
+        | EIO -> "EIO"
+        | EISDIR -> "EISDIR"
+        | EMFILE -> "EMFILE"
+        | EMLINK -> "EMLINK"
+        | ENAMETOOLONG -> "ENAMETOOLONG"
+        | ENFILE -> "ENFILE"
+        | ENODEV -> "ENODEV"
+        | ENOENT -> "ENOENT"
+        | ENOEXEC -> "ENOEXEC"
+        | ENOLCK -> "ENOLCK"
+        | ENOMEM -> "ENOMEM"
+        | ENOSPC -> "ENOSPC"
+        | ENOSYS -> "ENOSYS"
+        | ENOTDIR -> "ENOTDIR"
+        | ENOTEMPTY -> "ENOTEMPTY"
+        | ENOTTY -> "ENOTTY"
+        | ENXIO -> "ENXIO"
+        | EPERM -> "EPERM"
+        | EPIPE -> "EPIPE"
+        | ERANGE -> "ERANGE"
+        | EROFS -> "EROFS"
+        | ESPIPE -> "ESPIPE"
+        | ESRCH -> "ESRCH"
+        | EXDEV -> "EXDEV"
+        | EWOULDBLOCK -> "EWOULDBLOCK"
+        | EINPROGRESS -> "EINPROGRESS"
+        | EALREADY -> "EALREADY"
+        | ENOTSOCK -> "ENOTSOCK"
+        | EDESTADDRREQ -> "EDESTADDRREQ"
+        | EMSGSIZE -> "EMSGSIZE"
+        | EPROTOTYPE -> "EPROTOTYPE"
+        | ENOPROTOOPT -> "ENOPROTOOPT"
+        | EPROTONOSUPPORT -> "EPROTONOSUPPORT"
+        | ESOCKTNOSUPPORT -> "ESOCKTNOSUPPORT"
+        | EOPNOTSUPP -> "EOPNOTSUPP"
+        | EPFNOSUPPORT -> "EPFNOSUPPORT"
+        | EAFNOSUPPORT -> "EAFNOSUPPORT"
+        | EADDRINUSE -> "EADDRINUSE"
+        | EADDRNOTAVAIL -> "EADDRNOTAVAIL"
+        | ENETDOWN -> "ENETDOWN"
+        | ENETUNREACH -> "ENETUNREACH"
+        | ENETRESET -> "ENETRESET"
+        | ECONNABORTED -> "ECONNABORTED"
+        | ECONNRESET -> "ECONNRESET"
+        | ENOBUFS -> "ENOBUFS"
+        | EISCONN -> "EISCONN"
+        | ENOTCONN -> "ENOTCONN"
+        | ESHUTDOWN -> "ESHUTDOWN"
+        | ETOOMANYREFS -> "ETOOMANYREFS"
+        | ETIMEDOUT -> "ETIMEDOUT"
+        | ECONNREFUSED -> "ECONNREFUSED"
+        | EHOSTDOWN -> "EHOSTDOWN"
+        | EHOSTUNREACH -> "EHOSTUNREACH"
+        | ELOOP -> "ELOOP"
+        | EOVERFLOW -> "EOVERFLOW"
+        | EUNKNOWNERR x -> Printf.sprintf "EUNKNOWNERR %d" x
+      in
+
+      Some
+        (Format.asprintf "Sendfile(Unix_error): %s; remaining len: %d" msg len)
+    | _ -> None)
+
 external sendfile :
    src:Unix.file_descr
   -> dst:Unix.file_descr
@@ -6,38 +93,32 @@ external sendfile :
   -> int
   = "ocaml_sendfile_sendfile_stub"
 
-let sendfile_once_exn ?(off = 0) ?len ~src dst =
-  let len =
-    match len with Some len -> len | None -> (Unix.fstat src).st_size - off
-  in
-  sendfile ~src ~dst ~off ~len
+let sendfile_once_exn ?(off = 0) ~len ~src dst = sendfile ~src ~dst ~off ~len
 
-let sendfile_once ?(off = 0) ?len ~src dst =
-  try
-    let ret = sendfile_once_exn ~src ~off ?len dst in
-    Ok ret
-  with
+let sendfile_once ?(off = 0) ~len ~src dst =
+  try Ok (sendfile_once_exn ~src ~off ~len dst) with
   | Unix.Unix_error (unix_err, _, _msg) -> Error unix_err
 
-let rec sendfile_exn ?(off = 0) ?len ~src dst =
+let sendfile_exn ?(off = 0) ?len ~src dst =
+  let rec sendfile_exn ~off ~len ~src dst =
+    match sendfile_once_exn ~src ~off ~len dst with
+    | c when c = len -> len
+    | c -> sendfile_exn ~src ~off:(off + c) ~len:(len - c) dst
+    | exception Unix.Unix_error ((EINTR | EAGAIN), _, _) ->
+      sendfile_exn ~src ~off ~len dst
+    | exception Darwin_specific_unix_error ((EINTR | EAGAIN), sent) ->
+      (* Darwin systems signal the number of bytes partially sent on EINTR /
+         EAGAIN. *)
+      sendfile_exn ~src ~off:(off + sent) ~len:(len - sent) dst
+  in
   let len =
     match len with Some len -> len | None -> (Unix.fstat src).st_size - off
   in
-  match sendfile_once_exn ~src ~off ~len dst with
-  | c when c = len -> len
-  | c -> sendfile_exn ~src ~off:(off + c) ~len:(len - c) dst
-  | exception Unix.Unix_error (Unix.EINTR, _, _) ->
-    (* From:
-       https://developer.apple.com/library/archive/documentation/System/Conceptual/ManPages_iPhoneOS/man2/sendfile.2.html
-
-       A signal interrupted sendfile() before it could be completed. If
-       specified, the number of bytes success-fully successfully fully sent will
-       be returned in *len. *)
-    sendfile_exn ~src ~off ~len dst
+  let _sent = sendfile_exn ~off ~len ~src dst in
+  (* If we're here, we sent the whole file correctly, so we return the total
+     length sent. *)
+  len
 
 let sendfile ?off ?len ~src dst =
-  try
-    let ret = sendfile_exn ~src ?off ?len dst in
-    Ok ret
-  with
+  try Ok (sendfile_exn ~src ?off ?len dst) with
   | Unix.Unix_error (unix_err, _, _msg) -> Error unix_err

sendfile/sendfile.mli

@@ -1,7 +1,6 @@
 val sendfile_once_exn :
    ?off:int (** Defaults to 0. *)
-  -> ?len:int
-       (** Defaults to length of data (file) associated with descriptor [fd]. *)
+  -> len:int
   -> src:Unix.file_descr
   -> Unix.file_descr
   -> int
@@ -11,8 +10,7 @@ val sendfile_once_exn :
 
 val sendfile_once :
    ?off:int (** Defaults to 0. *)
-  -> ?len:int
-       (** Defaults to length of data (file) associated with descriptor [fd]. *)
+  -> len:int
   -> src:Unix.file_descr
   -> Unix.file_descr
   -> (int, Unix.error) result
@@ -27,8 +25,8 @@ val sendfile_exn :
   -> src:Unix.file_descr
   -> Unix.file_descr
   -> int
-(** Calls `sendfile(2)`, possibly repeatedly (if EINTR is returned) until the
-    entire file has been sent.
+(** Calls `sendfile(2)`, possibly repeatedly (if EINTR / EAGAIN is returned)
+    until the entire file has been sent.
 
     @raise Sys_error if `sendfile` is not supported by the platform. *)
 

sendfile/sendfile_stubs.c

@@ -19,13 +19,28 @@
 #include <sys/sendfile.h>
 #endif
 
+#include <errno.h>
+
 #define CAML_NAME_SPACE
+#include <caml/alloc.h>
+#include <caml/callback.h>
+#include <caml/custom.h>
 #include <caml/fail.h>
 #include <caml/memory.h>
 #include <caml/mlvalues.h>
 #include <caml/threads.h>
 #include <caml/unixsupport.h>
 
+value sendfile_unix_error(int errcode, off_t len) {
+  CAMLparam0();
+  value err, res;
+  err = caml_unix_error_of_code(errcode);
+  res = caml_alloc_small(2, 0);
+  Field(res, 0) = err;
+  Field(res, 1) = Val_long(len);
+  CAMLreturn(res);
+}
+
 /*
  * Sendfile has different signatures on macOS and Linux.
  */
@@ -42,8 +57,26 @@ CAMLprim value ocaml_sendfile_sendfile_stub(value v_fd, value v_sock,
   ret = sendfile(Int_val(v_fd), Int_val(v_sock), offset, &len, NULL, 0);
   caml_acquire_runtime_system();
 
-  if (ret == -1)
-    uerror("sendfile", Nothing);
+  /*
+   * For the macOS case we raise a custom exception with the remaining length
+   * to be sent, due to `sendfile` differences in the Apple platform.
+   *
+   * From
+   * https://developer.apple.com/library/archive/documentation/System/Conceptual/ManPages_iPhoneOS/man2/sendfile.2.html:
+   *
+   * EINTR: A signal interrupted sendfile() before it could be completed.  If
+   * specified, the number of bytes successfully fully sent will
+   * be returned in *len.
+   *
+   * Similarly for EAGAIN:
+   * The socket is marked for non-blocking I/O and not all data was sent due to
+   * the socket buffer being full.  If specified, the number of bytes
+   * successfully sent will be returned in *len.
+   */
+  if (ret == -1) {
+    caml_raise_with_arg(*caml_named_value("sendfile_exn_unix_error"),
+                        sendfile_unix_error(errno, len));
+  }
 
   CAMLreturn(Val_long(len));
 }