Hi,
On Mon, Jan 17, 2022 at 4:51 AM Lev Stipakov wrote:
> From: Lev Stipakov
>
> tun_finalize() is essentially subset of socket_finalize() apart from:
>
> - using WSAFoo() functions instead of Foo()
>
> - "from" address is not returned
>
> There is no clear official statement that one can use non-WSA
> API on handles, so let's be on a safe side and use both.
>
> Introduce sockethandle_t abstraction, which represents
> socket and handle. Add SocketHandle* routines which call
> proper API depends on underlying type in abstraction.
>
> Rename socket_finalize() to sockethandle_finalize(), take
> sockethandle_t and new routines into use and kick tun_finalize().
>
> Signed-off-by: Lev Stipakov
> ---
> v3: replace macros with inline functions
>
> v2: explicitly initialize .is_handle to false for readablity
>
>
> src/openvpn/forward.c | 3 +-
> src/openvpn/socket.c | 37 -
> src/openvpn/socket.h | 49 ++
> src/openvpn/tun.c | 94 +--
> src/openvpn/tun.h | 34 +---
> 5 files changed, 80 insertions(+), 137 deletions(-)
>
> diff --git a/src/openvpn/forward.c b/src/openvpn/forward.c
> index f82386a1..a905f993 100644
> --- a/src/openvpn/forward.c
> +++ b/src/openvpn/forward.c
> @@ -1115,7 +1115,8 @@ read_incoming_tun(struct context *c)
> }
> else
> {
> -read_tun_buffered(c->c1.tuntap, >c2.buf);
> +sockethandle_t sh = { .is_handle = true, .h = c->c1.tuntap->hand
> };
> +sockethandle_finalize(sh, >c1.tuntap->reads, >c2.buf, NULL);
> }
> #else /* ifdef _WIN32 */
> ASSERT(buf_init(>c2.buf, FRAME_HEADROOM(>c2.frame)));
> diff --git a/src/openvpn/socket.c b/src/openvpn/socket.c
> index df736746..780c5cb3 100644
> --- a/src/openvpn/socket.c
> +++ b/src/openvpn/socket.c
> @@ -3198,7 +3198,8 @@ link_socket_read_tcp(struct link_socket *sock,
> if (!sock->stream_buf.residual_fully_formed)
> {
> #ifdef _WIN32
> -len = socket_finalize(sock->sd, >reads, buf, NULL);
> +sockethandle_t sh = { .s = sock->sd };
> +len = sockethandle_finalize(sh, >reads, buf, NULL);
> #else
> struct buffer frag;
> stream_buf_get_next(>stream_buf, );
> @@ -3664,10 +3665,10 @@ socket_send_queue(struct link_socket *sock, struct
> buffer *buf, const struct lin
> }
>
> int
> -socket_finalize(SOCKET s,
> -struct overlapped_io *io,
> -struct buffer *buf,
> -struct link_socket_actual *from)
> +sockethandle_finalize(sockethandle_t sh,
> + struct overlapped_io *io,
> + struct buffer *buf,
> + struct link_socket_actual *from)
> {
> int ret = -1;
> BOOL status;
> @@ -3675,13 +3676,7 @@ socket_finalize(SOCKET s,
> switch (io->iostate)
> {
> case IOSTATE_QUEUED:
> -status = WSAGetOverlappedResult(
> -s,
> ->overlapped,
> ->size,
> -FALSE,
> ->flags
> -);
> +status = SocketHandleGetOverlappedResult(sh, io);
> if (status)
> {
> /* successful return for a queued operation */
> @@ -3693,18 +3688,18 @@ socket_finalize(SOCKET s,
> io->iostate = IOSTATE_INITIAL;
> ASSERT(ResetEvent(io->overlapped.hEvent));
>
> -dmsg(D_WIN32_IO, "WIN32 I/O: Socket Completion success
> [%d]", ret);
> +dmsg(D_WIN32_IO, "WIN32 I/O: Completion success [%d]",
> ret);
> }
> else
> {
> /* error during a queued operation */
> ret = -1;
> -if (WSAGetLastError() != WSA_IO_INCOMPLETE)
> +if (SocketHandleGetLastError(sh) != ERROR_IO_INCOMPLETE)
>
WSA_IO_INCOMPLETE is the same as ERROR_IO_INCOMPLETE, so using them
interchangeably looks ok.
{
> /* if no error (i.e. just not finished yet), then
> DON'T execute this code */
> io->iostate = IOSTATE_INITIAL;
> ASSERT(ResetEvent(io->overlapped.hEvent));
> -msg(D_WIN32_IO | M_ERRNO, "WIN32 I/O: Socket
> Completion error");
> +msg(D_WIN32_IO | M_ERRNO, "WIN32 I/O: Completion
> error");
> }
> }
> break;
> @@ -3715,9 +3710,9 @@ socket_finalize(SOCKET s,
> if (io->status)
> {
> /* error return for a non-queued operation */
> -WSASetLastError(io->status);
> +SocketHandleSetLastError(sh, io->status);
> ret = -1;
> -msg(D_WIN32_IO | M_ERRNO, "WIN32 I/O: Socket Completion
> non-queued error");
> +msg(D_WIN32_IO | M_ERRNO, "WIN32 I/O: Completion
> non-queued error");
> }
> else
>
