[PATCH] ws2_32: Ignore small SNDBUF values

[Alistair Leslie-Hughes]

From: Sebastian Lackner

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=40264
Signed-off-by: Alistair Leslie-Hughes 
---
 dlls/ws2_32/socket.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/dlls/ws2_32/socket.c b/dlls/ws2_32/socket.c
index df068fe..ee265d6 100644
--- a/dlls/ws2_32/socket.c
+++ b/dlls/ws2_32/socket.c
@@ -5809,9 +5809,10 @@ int WINAPI WS_setsockopt(SOCKET s, int level, int optname,
             break;

case WS_SO_RCVBUF:
+        case WS_SO_SNDBUF:
             if (*(const int*)optval < 2048)
             {
-                WARN("SO_RCVBF for %d bytes is too small: ignored\n", *(const int*)optval );
+                WARN("SO_RCVBF/WS_SO_SNDBUF for %d bytes is too small: ignored\n", *(const int*)optval );
                 return 0;
             }
             /* Fall through */
@@ -5828,7 +5829,6 @@ int WINAPI WS_setsockopt(SOCKET s, int level, int optname,
          * however, using it the BSD way fixes bug 8513 and seems to be what
          * most programmers assume, anyway */
         case WS_SO_REUSEADDR:
-        case WS_SO_SNDBUF:
         case WS_SO_TYPE:
             convert_sockopt(&level, &optname);
             break;

[Marvin]

Hi,

While running your changed tests on Windows, I think I found new failures.
Being a bot and all I'm not very good at pattern recognition, so I might be
wrong, but could you please double-check?

Full results can be found at:
https://testbot.winehq.org/JobDetails.pl?Key=44465

Your paranoid android.

=== debian9 (32 bit Wine report) ===

ws2_32:
sock.c:5440: Test failed: Failed to start overlapped send 0 - 0 - 1048576/1048576
sock.c:5480: Test failed: Failed to start overlapped send 0 - 0

=== debian9 (64 bit Wow Wine report) ===

ws2_32:
sock.c:5440: Test failed: Failed to start overlapped send 0 - 0 - 1048576/1048576
sock.c:5480: Test failed: Failed to start overlapped send 0 - 0

[Hans Leidekker]

On Fri, 2018-11-16 at 03:40 +0000, Alistair Leslie-Hughes wrote:
> From: Sebastian Lackner
>
> Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=40264
> Signed-off-by: Alistair Leslie-Hughes
>
> dlls/ws2_32/socket.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/dlls/ws2_32/socket.c b/dlls/ws2_32/socket.c
> index df068fe..ee265d6 100644
> --- a/dlls/ws2_32/socket.c
> +++ b/dlls/ws2_32/socket.c
> @@ -5809,9 +5809,10 @@ int WINAPI WS_setsockopt(SOCKET s, int level, int optname,
>              break;
>      case WS_SO_RCVBUF:
> -
>    case WS_SO_SNDBUF:
>        if (*(const int*)optval < 2048)
>        {
> -
>            WARN("SO_RCVBF for %d bytes is too small: ignored\n", *(const int*)optval );
> -
>            WARN("SO_RCVBF/WS_SO_SNDBUF for %d bytes is too small: ignored\n", *(const int*)optval );
>            return 0;

That bug is about macOS rejecting setsockopt(SO_SNDBUF) with 0 buffer
size. This is supposed to disable send buffering on Windows but macOS
doesn't support that (nor does Linux, it sets a default value in this
case).

I think it would be better to print a fixme for this specific value and
then return early only on macOS, to avoid the error from setsockopt.