Asynchronus serial port

[Fenixk19]

Hello! I've already post the
bug(http://bugs.winehq.org/show_bug.cgi?id=19713) on this subject, but i
need more help. So I've decided to write here.
There is a problem in wine. When I use asynchronous serial port read, 
data never comes. Event, caused by select comes. But operation status
stays pending, and i can't do anything to this serial port anymore. In
windows it never get pending, and port can be accessed just after data
arrival. Seems to be wineserver problem, but i don't know, where to look
at. What code respond for asynchronous serial port in wineserver?
Alexander.
P.S. Test program attached.

[Mike Kaplinskiy]

On Wed, Sep 2, 2009 at 7:31 PM, Fenixk19 wrote:
> Hello! I've already post the
> bug(http://bugs.winehq.org/show_bug.cgi?id=19713) on this subject, but i
> need more help. So I've decided to write here.
> There is a problem in wine. When I use asynchronous serial port read,
> data never comes. Event, caused by select comes. But operation status
> stays pending, and i can't do anything to this serial port anymore. In
> windows it never get pending, and port can be accessed just after data
> arrival. Seems to be wineserver problem, but i don't know, where to look
> at. What code respond for asynchronous serial port in wineserver?
> Alexander.
> P.S. Test program attached.

Hi,

Alexandre would be the guy to talk to about wineserver-related things.
Sadly he's off on a long weekend. Does the attached patch help solve
the problem?

Mike.

[Fenixk19]

Yes, this patch solved problem. Great thanks. But there is another one.
It hangs on another call of GetOverlappedResult, where windows going on.
It hangs the same way - function always return, because
ovOverlapped.Internal is STATUS_PENDING. In attached file part of new
test problem, to show where it is. I'll send full variant on bugzilla in
minutes. http://bugs.winehq.org/show_bug.cgi?id=19713

Mike Kaplinskiy пишет:
> On Wed, Sep 2, 2009 at 7:31 PM, Fenixk19 wrote:
> > Hello! I've already post the
> > bug(http://bugs.winehq.org/show_bug.cgi?id=19713) on this subject, but i
> > need more help. So I've decided to write here.
> > There is a problem in wine. When I use asynchronous serial port read,
> > data never comes. Event, caused by select comes. But operation status
> > stays pending, and i can't do anything to this serial port anymore. In
> > windows it never get pending, and port can be accessed just after data
> > arrival. Seems to be wineserver problem, but i don't know, where to look
> > at. What code respond for asynchronous serial port in wineserver?
> > Alexander.
> > P.S. Test program attached.
>
> Hi,
>
> Alexandre would be the guy to talk to about wineserver-related things.
> Sadly he's off on a long weekend. Does the attached patch help solve
> the problem?
>
> Mike.

[Wolfgang Walter]

Am Donnerstag, 3. September 2009 schrieb Mike Kaplinskiy:
> On Wed, Sep 2, 2009 at 7:31 PM, Fenixk19 wrote:
> > Hello! I've already post the
> > bug(http://bugs.winehq.org/show_bug.cgi?id=19713) on this subject, but i
> > need more help. So I've decided to write here.
> > There is a problem in wine. When I use asynchronous serial port read,
> > data never comes. Event, caused by select comes. But operation status
> > stays pending, and i can't do anything to this serial port anymore. In
> > windows it never get pending, and port can be accessed just after data
> > arrival. Seems to be wineserver problem, but i don't know, where to look
> > at. What code respond for asynchronous serial port in wineserver?
> > Alexander.
> > P.S. Test program attached.
>
> Hi,
>
> Alexandre would be the guy to talk to about wineserver-related things.
> Sadly he's off on a long weekend. Does the attached patch help solve
> the problem?

Isn't setting commio->iosb->u.Status racy?

commio->iosb may be set from wait_for_event() which is called by an extra 
thread or by io_control().

Couldn't it happen that

wait_on starts the thread
    wait_for_event sets the commio->iosb->u.Status to STATUS_SUCCESS
    wait_on returns to io_control STATUS_PENDING
    io_control overwrites commio->iosb->u.Status with STATUS_PENDING

I think io_control() should set commio->iosb->u.Status to STATUS_PENDING 
before calling wait_on(). After wait_on() io_control() should not set 
commio->iosb->u.Status if wait_on() returns STATUS_PENDING. Here the idea:

------------------
    case IOCTL_SERIAL_WAIT_ON_MASK:
         if (lpOutBuffer && nOutBufferSize == sizeof(DWORD))
         {
             piosb->u.Status = STATUS_PENDING;
             piosb->Information = sz;
             if (!(status = wait_on(hDevice, fd, hEvent, piosb, lpOutBuffer)))
                 sz = sizeof(DWORD);
             else if (status == STATUS_PENDING)
                 return status;
         }
         else
             status = STATUS_INVALID_PARAMETER;
         break;
-------------------

Regards,

[Alexander Varnin]

Thanks, I'll check it, when i have time.

Wolfgang Walter пишет:
> Am Donnerstag, 3. September 2009 schrieb Mike Kaplinskiy:
> > On Wed, Sep 2, 2009 at 7:31 PM, Fenixk19 wrote:
> > > Hello! I've already post the
> > > bug(http://bugs.winehq.org/show_bug.cgi?id=19713) on this subject, but i
> > > need more help. So I've decided to write here.
> > > There is a problem in wine. When I use asynchronous serial port read,
> > > data never comes. Event, caused by select comes. But operation status
> > > stays pending, and i can't do anything to this serial port anymore. In
> > > windows it never get pending, and port can be accessed just after data
> > > arrival. Seems to be wineserver problem, but i don't know, where to look
> > > at. What code respond for asynchronous serial port in wineserver?
> > > Alexander.
> > > P.S. Test program attached.
> >
> > Hi,
> >
> > Alexandre would be the guy to talk to about wineserver-related things.
> > Sadly he's off on a long weekend. Does the attached patch help solve
> > the problem?
>
> Isn't setting commio->iosb->u.Status racy?
>
> commio->iosb may be set from wait_for_event() which is called by an extra 
> thread or by io_control().
>
> Couldn't it happen that
>
> wait_on starts the thread
>    wait_for_event sets the commio->iosb->u.Status to STATUS_SUCCESS
>    wait_on returns to io_control STATUS_PENDING
>    io_control overwrites commio->iosb->u.Status with STATUS_PENDING
>
> I think io_control() should set commio->iosb->u.Status to STATUS_PENDING 
> before calling wait_on(). After wait_on() io_control() should not set 
> commio->iosb->u.Status if wait_on() returns STATUS_PENDING. Here the idea:
> case IOCTL_SERIAL_WAIT_ON_MASK:
>      if (lpOutBuffer && nOutBufferSize == sizeof(DWORD))
>      {
>          piosb->u.Status = STATUS_PENDING;
>          piosb->Information = sz;
>          if (!(status = wait_on(hDevice, fd, hEvent, piosb, lpOutBuffer)))
>              sz = sizeof(DWORD);
>          else if (status == STATUS_PENDING)
>              return status;
>      }
>      else
>          status = STATUS_INVALID_PARAMETER;
>      break;
>
> Regards,