With the new unixcommon library we have things outside of common/,
so make the handling of out-of-tree builds more general.

This can be used to test if a layout if possible.

For example, if we have earlier disabled an output because it was too
large for the framebuffer, we should prefer this output when
allocating new outputs.

Move the code that turn off unused outputs to the end.

Also, add support for checking the output connection state.

This only matters when using clones, which we do not accept yet. This
patch is for completeness and to avoid future surprises.

Return error if no CRTC.

Screen is a Xvnc thing and not relevant for x0vncserver etc.

RANDR is mandatory since Xorg 1.7, which is the oldest version we
support.

In RandR land, there's a lot of return code confusion. Our wrappers
are using the same return codes as RRCrtcSet, RRScreenSizeSet: 1/TRUE
for success. Fixes:

* vncRandRCreateOutputs did not follow this convention

* A lot of code returned -1 upon failure

* vncRandRDisableOutput returned 0 for already disabled outputs

These two code blocks are not specific to Xvnc/vnc.so, but useful for
x0vncserver as well. RandrGlue.h defines the interface on which
unixcommon depends on.

Compatibility with libXrandr, where the name is in dynamically
allocated XRROutputInfo.

Less confusing, and compatible with libXrandr.

Should use 2 spaces like the rest of the project.

TriFan and TriStrip were added in 1.12, so we need to have that code
conditional to support older versions of Xorg.

There were a bunch of parameters that weren't properly added to the
man pages when the features were added. Make sure everything is
properly in sync.

They share a lot of settings, so make sure the documentation is
identical in those cases.

The funcationality was removed in 6e49e95, so let's also clean up
the documentation referring to it.

We had overlooked a few so we didn't track all screen changes properly.

We need to flip the dimensions in the case the monitor is rotated
90° or they won't map correctly to the framebuffer.

No longer tolerate loading the VNC extension but not being able to
initialize it. This avoids a lot of error prone checking to see
if the extension has started fully or not.

The normal Xorg server doesn't check this field, so neither can we
as there are applications relying on this leniancy.

It was moved to the common code in f8e3b34c6, but it was unreliable
because the state could sometimes get out of sync. Push it back in
to Xvnc since it isn't necessarily something all servers will have.

Otherwise we'll wedge the desktop object which expects an approval
or rejection to eventually show up.

It was only used by WinVNC, so push it there instead.

This adds the basic infrastructure and handshake for the QEMU
Extended Key Events extension. No viewer or server makes use of
the extra functionality yet though.

Based on QEMU's behaviour.

This will allow us to use better formats that preserve the entire
alpha channel.

The functions might change so we need to also make sure we grab
the updated value after each call. Clean up the code to use the
same style as the rest of the Xorg code.

One was missing a call to register the fd with the X server, and
one forgot to set it to non-blocking. One result of this was a crash
when hitting the blacklist.

xserver 1.19's OsInit will create a pollfd, followed by checking if fd 2 /
stderr is writable and if it is not, replacing fd 2 with /dev/null.

Since we close stderr in inetd mode to avoid xserver messages being send
to the client as vnc data, the pollfd becomes fd 2, only to be replaced
by /dev/null since a pollfd is not writable.

This commit fixes this by opening /dev/null directly after the close(2),
avoiding that the pollfd becomes fd 2.

Alan Coopersmith: Change to use dup2() for atomic switch of fd

Signed-off-by: Hans de Goede <hdegoede@redhat.com>

ChangeWindowProperty() was removed upstream because it was redundant,
and dixChangeWindowProperty() has been around since 1.5 so we can
safely use that.

Use the new X server API as the reference system and emulate it on
the older API. Avoids a lot of code duplication.

Xvnc didn't handle multiple screens properly as it forgot to place
them without overlap and didn't handle cursor movement between them.

Pointer are global, not screen specific. Hence so is also the cursor
sprite.

The client is not aware of where the screen is placed so it will give
us screen relative coordinates. Convert to and from these and absolute
coordinates before interacting with the input layer.