Fixes a testbot failure.
From: Alex Henrie alexhenrie24@gmail.com
Fixes a testbot failure. --- dlls/rpcrt4/tests/rpc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/dlls/rpcrt4/tests/rpc.c b/dlls/rpcrt4/tests/rpc.c index a4653fd603f..3f6dd8745b9 100644 --- a/dlls/rpcrt4/tests/rpc.c +++ b/dlls/rpcrt4/tests/rpc.c @@ -188,7 +188,8 @@ static void test_DceErrorInqTextA(void) * 3814 is generally quite a long message */ ok ((DceErrorInqTextA (3814, (unsigned char*)buffer) == RPC_S_OK && - strcmp (buffer, bufferInvalid) == 0), + (strcmp (buffer, bufferInvalid) == 0 || + broken (buffer[0] == '?') /* win10 21H1 locale mix */)), "DceErrorInqTextA(deviation...)\n"); } else
I guess, although honestly, the test seems not to really be checking A/W conversion, so maybe it'd make more sense to sidestep it by just changing the test to use W functions instead?
[Also, the test is quite awful in a lot of ways, like checking multiple things at once in an ok() message, not printing the expected or received values, using ok (0) inside a conditional, using magic numbers, and perhaps most confusingly, I can't even tell what it's trying to test. That DceErrorInqText() matches FormatMessage()? If so, why does it never test that real messages match...]
It looks like the string returned for '3814' is just different in this case, see: https://testbot.winehq.org/JobDetails.pl?Key=128382#k113
What's even stranger is that the unicode version sometimes doesn't match with ansi version, see: https://testbot.winehq.org/JobDetails.pl?Key=128384#k115
I think it's probably best to just remove the test for this code. Bonus points for cleaning things up a bit too.
I guess I don't really care about this enough to fix it properly. Feel free to take over.
This merge request was closed by Alex Henrie.