run p
calls the Lwt scheduler repeatedly until p
resolves, and returns the value of p
if it is fulfilled. If p
is rejected with an exception, that exception is raised.
Every native or bytecode program that uses Lwt should always use this function for evaluating a promise at the top level (such as its main function or main loop), otherwise promises that depend on I/O operations will not be resolved.
Example:
let main () = Lwt_io.write_line Lwt_io.stdout "hello world"
let () = Lwt_main.run @@ main ()
When targeting JavaScript, Lwt_main.run
is not available, but neither it's necessary since the JS environment automatically takes care of the I/O considerations.
On Unix, calling Lwt_main.run
installs a SIGCHLD
handler, which is needed for the implementations of Lwt_unix.waitpid
and Lwt_unix.wait4
. As a result, programs that call Lwt_main.run
and also use non-Lwt system calls need to handle those system calls failing with EINTR
.
Note that you should avoid using run
inside threads
- The calling threads will not resume before
run
returns. - Successive invocations of
run
are serialized: an invocation of run
will not terminate before all subsequent invocations are terminated.
Note also that it is not safe to call run
in a function registered with Pervasives.at_exit
, use the at_exit
function of this module instead.