[PATCH v4 1/5] server: Add hardware_msg_data size for variable size messages.

[Rémi Bernon]

On 5/28/21 3:35 PM, Rémi Bernon wrote:
> On 5/28/21 3:34 PM, Rémi Bernon wrote:
>> On 5/28/21 2:31 PM, Marvin wrote:
>>> Hi,
>>>
>>> While running your changed tests, 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=91396
>>>
>>> Your paranoid android.
>>>
>>>
>>> === debiant2 (32 bit report) ===
>>>
>>> user32:
>>> input: Timeout
>>> menu: Timeout
>>> msg.c:11097: Test failed: MOUSE_LL hook other thread: 0: the msg 
>>> sequence is not complete: expected 0200 - actual 0000
>>> msg.c:11104: Test failed: MOUSE_LL hook same thread: 0: the msg 
>>> sequence is not complete: expected 0200 - actual 0000
>>> msg: Timeout
>>> win.c:9693: Test failed: Timed out waiting for the child process
>>> win.c:9619: Test failed: transparent window didn't get WM_NCHITTEST 
>>> message
>>> win.c:9620: Test failed: button under static window didn't get 
>>> WM_LBUTTONUP
>>> win: Timeout
>>>
>>> === debiant2 (32 bit Chinese:China report) ===
>>>
>>> user32:
>>> edit: Timeout
>>> input: Timeout
>>> menu: Timeout
>>> msg.c:11097: Test failed: MOUSE_LL hook other thread: 0: the msg 
>>> sequence is not complete: expected 0200 - actual 0000
>>> msg.c:11104: Test failed: MOUSE_LL hook same thread: 0: the msg 
>>> sequence is not complete: expected 0200 - actual 0000
>>> msg: Timeout
>>> win.c:9693: Test failed: Timed out waiting for the child process
>>> win.c:9619: Test failed: transparent window didn't get WM_NCHITTEST 
>>> message
>>> win.c:9620: Test failed: button under static window didn't get 
>>> WM_LBUTTONUP
>>> win: Timeout
>>>
>>
>> Looks like this happens just from the additional size field in 
>> hardware_msg_data, did I messed up the alignment or something?
>>
>> It doesn't strike me as incorrect though, everything looks to be 
>> unsigned int / or two consecutive shorts?
>>
>> Or maybe it's the server protocol update which triggers some hidden bug?
> 
> Well, seems to be alignment related somehow, adding another int fixes it:
> 
> https://testbot.winehq.org/JobDetails.pl?Key=91416

What confused me is that I'm not able to reproduce it locally, even with 
a WINEARCH=win32 prefix, but I probably have 64-bit wineserver all the 
time, and that's why.
-- 
Rémi Bernon <rbernon at codeweavers.com>