server: Add support for pending write flag to the serial device. Take 2.

[Wolfgang Walter]

Am Mittwoch, 23. Oktober 2013, 11:12:26 schrieben Sie:
> Wolfgang Walter <wine at stwm.de> wrote:
> > > ../../../tools/runtest -q -P wine -T ../../.. -M kernel32.dll -p
> > > kernel32_test.exe.so comm.c && touch comm.ok comm.c:860: Test failed:
> > > WaitCommEvent failed with a timeout
> > > comm.c:881: Test failed: WaitCommEvent error 997
> > > comm.c:882: Test failed: WaitCommEvent: expected EV_TXEMPTY, got 0
> > > comm.c:887: Test failed: WaitCommEvent used 1003 ms for waiting
> > > comm.c:2226: Tests skipped: interactive tests (set
> > > WINETEST_INTERACTIVE=1)
> > > make: *** [comm.ok] Error 4
> > 
> > This is because the previous test fails: in the previous test wine
> > actually
> > write to the serial device and does not return with
> > ERROR_INVALID_PARAMETER.
> What is this "previous test" you are talking about? Obviously there is no
> a failing test which checks for ERROR_INVALID_PARAMETER in the output posted
> by Alexandre above. And it's worth to mention that these tests pass without
> failures here with and without an external hardware turned on (with the
> patch applied).

With privious test I mean the test which starts at line 836:

    S(U(ovl_write)).Offset = 0;
    S(U(ovl_write)).OffsetHigh = 0;
    ovl_write.hEvent = CreateEventW(NULL, TRUE, FALSE, NULL);
    before = GetTickCount();
    SetLastError(0xdeadbeef);
    res = WriteFile(hcom, tbuf, sizeof(tbuf), &bytes, &ovl_write);
    after = GetTickCount();
todo_wine
    ok(!res && GetLastError() == ERROR_IO_PENDING, "WriteFile returned %d, 
error %d\n", res, GetLastError());
todo_wine
    ok(!bytes, "expected 0, got %u\n", bytes);
    ok(after - before < 30, "WriteFile took %d ms to write %d Bytes at %d 
Baud\n",
       after - before, bytes, baud);
    /* don't wait for WriteFile completion */


This tests sends 17 bytes but does not wait till all have been sent. So the 
next test which fails under wine (comm.c:860: Test failed: WaitCommEvent 
failed with a timeout) does not start with a empty tx queue. There should be a 
sleep(TIMEOUT) after this test.

The test for ERROR_INVALID_PARAMETER before that one (starting at line 831) 
does indeed not contribute any more to

	comm.c:860: Test failed: WaitCommEvent failed with a timeout

any more. When we discussed this it did because then the the code still was:

    SetLastError(0xdeadbeef);
    res = WriteFile(hcom, tbuf, sizeof(tbuf), &bytes, NULL);
todo_wine
    ok(!res, "WriteFile on an overlapped handle without ovl structure should 
fail\n");
todo_wine
    ok(GetLastError() == ERROR_INVALID_PARAMETER, "expected 
ERROR_INVALID_PARAMETER, got %d\n", GetLastError());

So then there were even more bytes pending in the tx queue.


Regards,
-- 
Wolfgang Walter
Studentenwerk München
Anstalt des öffentlichen Rechts