[Bug 36040] New: Environment variable to disable wineserver persistence

http://bugs.winehq.org/show_bug.cgi?id=36040

Bug ID: 36040
           Summary: Environment variable to disable wineserver persistence
           Product: Wine
           Version: unspecified
          Hardware: x86
                OS: Linux
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: wineserver
          Assignee: [e-mail]
          Reporter: [e-mail]

Normally, when wine runs a Windows program, and the Windows program exits,
wineserver persists for a couple of seconds to improve the performance of
subsequent programs.  When running a testsuite that waits for all child
processes to exit, this delays each test completion for a couple of seconds. 
Thus, I'd like to have an environment variable that tells wineserver to exit
the instant the Windows program does.

[Bug 36040] Environment variable to disable wineserver persistence

http://bugs.winehq.org/show_bug.cgi?id=36040

Josh Triplett  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |enhancement

[Bug 36040] Environment variable to disable wineserver persistence

http://bugs.winehq.org/show_bug.cgi?id=36040

[e-mail] changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

--- Comment #1 from [e-mail] ---
Before seeing this ticket I have opened a related topic in the forum:
http://forum.winehq.org/viewtopic.php?f=8&t=22645

Currently it is a little tricky to change the default timeout. The WINESERVER
environment variable seems to be not working correctly as it would otherwise
server as a workaround for this problem. But instead of having an environment
variable that controls just a flag if the persistence should be disabled or not
it would make more sense to control directly the timeout (as this would
implicitly include such a flag).

[Bug 36040] Environment variable to disable wineserver persistence

http://bugs.winehq.org/show_bug.cgi?id=36040

--- Comment #2 from Josh Triplett  ---
(In reply to sworddragon2 from comment #1)
> Before seeing this ticket I have opened a related topic in the forum:
> http://forum.winehq.org/viewtopic.php?f=8&t=22645
>
> Currently it is a little tricky to change the default timeout. The
> WINESERVER environment variable seems to be not working correctly as it
> would otherwise server as a workaround for this problem. But instead of
> having an environment variable that controls just a flag if the persistence
> should be disabled or not it would make more sense to control directly the
> timeout (as this would implicitly include such a flag).

A variable to set the timeout, with a specified way to disable persistence
(timeout 0, for instance), would work just fine for me.

[Bug 36040] Environment variable to disable wineserver persistence

http://bugs.winehq.org/show_bug.cgi?id=36040

Sebastian Lackner  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

--- Comment #3 from Sebastian Lackner  ---
@Josh: How does the wineserver block your test from exiting? Do you start the
wineserver explicitly? If yes, you can just pass the -p0 flag. Otherwise it
shouldn't block your tests because wineserver acts as a daemon and immediately
forks into the background.

If you just want to make sure the wineserver is killed after each test
(normally shouldn't be necessary, and will probably be a bit slower than
keeping the wineserver persistent for a couple seconds), why not use something
like:

wine ./unittest.exe
wineserver -k

Of course something like an environment WINE_SERVER_FLAGS would be nice
(especially for users that don't want a permanent wineserver, like in comment
1), but I don't see why it should be impossible to have a similar effect with
the existing commandline options. Can you probably provide a bit more details?

[Bug 36040] Environment variable to disable wineserver persistence

http://bugs.winehq.org/show_bug.cgi?id=36040

--- Comment #4 from Josh Triplett  ---
This is a dejagnu-based testsuite compiling and running test applications,
cross-compiled via mingw, and run via wine and binfmt-misc.  (Thus, I don't
have much control over the invocation of each test, and I'd like to avoid
adding any wine-specific code around test invocation.)  Each test blocks for a
while before completing.  I can watch "ps" when that happens, and observe that
wineserver is the only process running; the moment wineserver exits, the
testsuite moves on to the next test.

[Bug 36040] Environment variable to disable wineserver persistence

https://bugs.winehq.org/show_bug.cgi?id=36040

Ken Sharp  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |download, source

[Bug 36040] Environment variable to disable wineserver persistence

https://bugs.winehq.org/show_bug.cgi?id=36040

Jens Reyer  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

[Bug 36040] Environment variable to disable wineserver persistence

https://bugs.winehq.org/show_bug.cgi?id=36040

Ralf Habacker  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

--- Comment #5 from Ralf Habacker  ---
bug 40811(In reply to sworddragon2 from comment #1)
> Currently it is a little tricky to change the default timeout. The
> WINESERVER environment variable seems to be not working correctly as it
> would otherwise server as a workaround for this problem.

see bug 40811