Adding support for protect-from-close handle flag to CreateProcess
test infrastructure.
Signed-off-by: Eric Pouech <epouech@codeweavers.com>

Signed-off-by: Gabriel Ivăncescu <gabrielopcode@gmail.com>

It's mostly implemented, and issession isn't even exposed.
Signed-off-by: Gabriel Ivăncescu <gabrielopcode@gmail.com>

Signed-off-by: Gabriel Ivăncescu <gabrielopcode@gmail.com>

Signed-off-by: Gabriel Ivăncescu <gabrielopcode@gmail.com>

Signed-off-by: Gabriel Ivăncescu <gabrielopcode@gmail.com>

Signed-off-by: Gabriel Ivăncescu <gabrielopcode@gmail.com>

Signed-off-by: Gabriel Ivăncescu <gabrielopcode@gmail.com>

We need the caller ServiceProvider to obtain the source.
Signed-off-by: Gabriel Ivăncescu <gabrielopcode@gmail.com>

Windows 7 has a lower maximum depth for recursive hooks.

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=56293

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=56291

Decoders might output some fps information and encoders might input fps,
but otherwise is unnecessary and may prevent compatible caps matching.

Better reflect its contents.

There is still a small amount of code in this file not directly related to the
FFP:

 - compile_state_table(), which should eventually go away;

 - states like state_so() and indexbuffer(), which should eventually be moved to
   context_gl.c with the rest of the resource loading/binding code.

The entire rest of state.c is concerned with the GL fixed-function pipeline.

device.c seems the most fitting place for these, since they're associated with
the device and stored in the device. device.c is also one of the few files that
contains client-side code, which these functions are.

Looks like touches are automatically released if the device doesn't send
updates regularly.

This should be more reliable than mouse moves and it exhibits more
issues with the rawinput buffer struct sizes on new WoW64.