On Mon, Aug 16, 2021 at 10:19 AM Zebediah Figura (she/her) <zfigura(a)codeweavers.com> wrote:
On 8/13/21 12:56 AM, Alex Henrie wrote:
@@ -1453,12 +1451,10 @@ todo_wine k = 99; SetLastError(0xdeadbeef); err = getsockopt(s, IPPROTO_IP, IP_HDRINCL, (char *) &k, &size); + ok(err == -1, "Expected -1, got %d\n", err); todo_wine - { - ok(err == -1, "Expected -1, got %d\n", err); ok(GetLastError() == WSAEINVAL, "Expected 10022, got %d\n", GetLastError()); - ok(k == 99, "Expected 99, got %d\n", k); - } + ok(k == 99, "Expected 99, got %d\n", k); } else /* <= 2003 the tests differ between TCP and UDP, UDP silently accepts */ {
If this is supposed to return WSAEINVAL, any reason not to do so in this patch?
There are a lot of options that should cause WSAEINVAL (usually due to using an IPv4 option on an IPv6 socket or vice-versa). I figured I would send another patch later to add a block of case statements for all of them, but it would also be fine to start with just IP_HDRINCL. I can send a revised patch tonight. -Alex