[Bug 51873] Joy.cpl crashes on the FFB tab when the FFB device is selected in the drop down list

[WineHQ Bugzilla]

https://bugs.winehq.org/show_bug.cgi?id=51873

Ivo Ivanov <logos128 at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|---                         |FIXED

--- Comment #28 from Ivo Ivanov <logos128 at gmail.com> ---
(In reply to Rémi Bernon from comment #27)
> I believe this specific issue has been fixed since
> c2235c44c5f8fedaad775c69b9a475725874bbc7, could you confirm?
> 
> Regarding effect state report and the way IOCTL_HID_GET_INPUT_REPORT is
> supposed to work I suggest moving the discussion to
> https://bugs.winehq.org/show_bug.cgi?id=52062.

I can confirm it is fixed. I was going to change the bug status, just wanted to
write my report from the weekend :)

I am happy to say that on the UDEV hidraw backend everything works as expected.
With the games I tested the FFB is pretty much equal to Windows. I compared
some captured usb traffic and the sent commands are similar in frequency and
sequence. There are some small differences mostly around the RESET command, but
I'll have to confirm that further with some conformance testing. The direction
used by WheelCheck and the most games is 0x2d on both Windows and Wine.

I tested also the other backends. The SDL backend works very well now. The FFB
is functional, although seems somewhat smoother with less details in comparison
to Windows / Wine (hidraw). The characteristic is the same, just with less
'punch'. My first thought was that it could be higher frequency of the FFB
rate, which in theory could provide a smoother force curve. As these direct
drive devices are capable of playing very small changes of the forces on almost
ms level.
The frequencies were similar, though, ~560Hz on Wine (hidraw / SDL) and ~400Hz
on Windows (Automobilista 2). So I compared some usb traffic and the first
noticeable thing is a difference in the direction value, 0x2d on Windows / Wine
(hidraw) and 0x5a on SDL. IMO this seems like a candidate for such slight
difference in the FFB behavior. I plan to compare it with proton's joystick_sdl
behavior.

The UDEV lnxev backend doesn't produce FFB, though. It manages to turn off the
bump stops of the wheel, which can be resolved normally with a power-cycle of
the device. There is a Reset FFB command in the TrueDrive software (Simucube
2's controlling app), which also resolves the normal behavior. That leads me to
think it just turns off the actuators with an FFB command. Probably this will
need another bug report.

I have also a strange crash of the winebus.sys process when using the UDEV bus.
It occurs only when using an administrator account (with superuser rights). I
will post a bug report with a patch, since finally had some time and found the
issue.

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.